Changes

LIST OF CHANGES

release 105.3.0 (2025-05-30)
 - Functionality of the Elembio staging monitor is extended. Most important changes:
   - LIMS batch ID is extracted from Elembio run name and saved to tracking database;
   - pairedness and indexing are determined from instrument data rather than hardcoded;
   - definition of failed runs is clarified to exclude runs which are marked by the
     instrument as failed, but produced data for all planned cycles.
 - A daemon for the Elembio staging monitor is implemented.
 - Added a method for retrieving a run record from the tracking database using run
   folder name, instrument name and flowcell ID as query parameters. The method,
   find_with_attributes, is added to the Run ResultSet class
 - Added a method for finding db run records

release 105.2.0 (2025-05-22)
 - npg_daemon_control: added a daemon for Elembio staging monitor 
 - npg_tracking::Schema::Result::Instrument
   - improved documentation,
   - added shortcut methods to inform about instrument's manufacturer.

release 105.1.0 (2025-05-15)
 - Create db records for AVITI instruments

release 105.0.0 (2025-05-15)
 - npg_daemon_control:
   - removed ability to stop/start/ping daemons on remote hosts
   - removed the '--hosts' script option
   - updated docs
 - Remove test affected by host date
 - Restricted Illumina instruments report to Illumina instruments

release 104.0.0 (2025-03-10)
 - Do not send emails to followers if the event relates to a run or a lane
   that was sequenced on a non-Illumina instrument.
 - Following deletion of the instrument_utilisation table from the tracking
   database, dropped exclusion of this table from a script that generates ORM.

release 103.0.0 (2025-02-20)
 - Replace mark docker image with mark golang executable from github,
   on ISG's docker image, for documentation publishing
 - Deleted unused methods from a number of models and views.
   - npg::model::instrument_annotation -
      annotations_by_instrument_over_default_uptime,
      dates_of_annotations_over_default_uptime
   - npg::view::run -
      list_summary, selected_days
   - npg::model_run_status -
       current_run_statuses, latest_current_run_status,
       active_runs_over_last_30_days, epoch
 - Deleted 'Stuck Runs' button from the admin top-level menu,
   the associated template and the view and model methods.
 - Deleted instrument_utilisation table from the tracking database. All
   code that was using this table was removed long time ago. Tha last
   record in this table is from 2016-11-12. The reports about instrument
   utilisation are now generated dynamically by C<bin/illumina_instruments_uptime> -
   a script in this package.
 - Added two new statuses to run_lane_status_dict database table,
   'archival in progress' and 'lane released'. This new statuses will drive
   lane-level archival and user notification system.
 - Replace mark docker image with mark golang executable from github,
   on ISG's docker image, for documentation publishing
 - Deleted two unused methods from npg::model::instrument_annotation,
   annotations_by_instrument_over_default_uptime and
   dates_of_annotations_over_default_uptime
 - Tracking web server:
   - Added a new optional parameter 'manufacturer' to methods returning lists
     of runs and run count - the name of the manufacturer. If this name is
     undefined in the parameters hash, it defaults to 'Illumina'.
   - To avoid code repetition, consolidated queries for listing and counting
     runs in npg::model::run. Previously the same code was repeated in eight
     different places in three models. In one of the models the code was never
     called, so was simply deleted.
   - Simplified the logic of retrieving cached lists and counts of runs.
   - Dropped a custom 'read' method from npg::view::instrument. The code
     there, which computed lists of runs and runs count is not used in the
     template that renders an instrument. Listing of runs is implemented as
     an iframe and therefore rendering is done by the template for listing runs.
   - Dropped 'runs' and 'count_run' methods from npg::model::instrument, these
     methods were only called by the deleted 'read' method from
     npg::view::instrument
   - Simplified the logic of retrieving cached lists and counts of runs.
   - Added a new optional argument 'manufacturer' - the name of the manufacturer -
     to two methods in npg::model::instrument_format.
   - Made changes to views and templates for rendering instruments and instrument
     formats so that the 'manufacturer' parameter can be added to URLs and
     propagated to the models.

release 102.1.0 (2025-01-16)
 - Fixed perlbrew installation by installing libdevel-patchperl-perl in
   GitHub runner.
 - Fixed CI for ubuntu 24.04 by manually installing libmodule-runtime-perl
 - Extend the lims api to consider the sample_uuids attribute (in its plural
   form) while using the ml_warehouse driver mainly for tag0. Also, it
   extends tests for novaseq multiruns to use sample_uuid (in its single
   form).

release 102.0.0 (2024-11-28)
 - Removed 'bin/npg_deletable_dr_runs' since dr storage is decommissioned

release 101.6.0 (2024-10-24)
 - Added sample_uuid and sample_lims methods to st::api::lims
 - Patched st::api::lims::samplesheet to allow for scalar attributes
   ending with 's'. Previously the values of attributes with names
   ending with 's' were automatically converted to lists. In fact,
   this was necessary only for comma-separated strings of email addresses.
   The part of code that deals with email addresses was changed to
   work on attributes with names starting with 'email' rather than
   ending with 's'. Prior to applying this patch the value of the
   new 'sample_lims' attribute was returned as a list when a samplesheet
   was used as a source of LIMS data.
 - Documented handling of multiple values in st::api::lims.
 - Added is_lane method and composition_object attribute to st::api::lims.

release 101.5.1 (2024-10-04)
 - Added .github/dependabot.yml file to auto-update GitHub actions
 - Following a release on 07/09/2024, see https://metacpan.org/dist/App-perlbrew/changes,
   the checksum of the script served by https://install.perlbrew.pl had changed.
   https://install.perlbrew.pl is a redirect to raw
   https://github.com/gugod/App-perlbrew/blob/master/perlbrew-install, so
   the change originates from GitHub and can be trusted. Our CI flow compares
   the checksum of the downloaded script to the expected value. We now store
   an updated expected checksum value, which corresponds to the latest release.
 - GitHub CI - updated deprecated v2 runner action to v3

release 101.5.0 (2024-09-04)
 - Moved the parser for the LIMS reference genome notation from
   npg_tracking::data::reference::find to a new package
   npg_tracking::data::reference::util, thus providing a stand-alone
   implementation of the parser.

   Removed dependency of st::api::lims on npg_tracking::data::reference::find
   since the latter brings in the dependency on npg_tracking::data::reference::list,
   which requires that the reference repository root is defined and exists.
   Instead used the reference genome notation parser from
   npg_tracking::data::reference::util.
 - fixes issues with tests for npg_seq_pipeline 68.5.0 and higher

release 101.4.0 (2024-08-30)
 - Added species_from_reference_genome method to st::api::lims. If the
   reference_genome value is defined, this new method returns the name of the
   species. parse_reference_genome from npg_tracking::data::reference::find
   is used for parsing the reference genome string.

release 101.3.0 (2024-08-05)
 - Provided a direct access to the default section of the study configuration
   file.
 - Enhanced a textual listing of instruments
   Added a column for recently used staging servers so that the loaders can
   easily spot instruments, which write to the same server.
 - Update documentation for tags
 - docs
   - Automatic confluence update for 'docs/api_usage.md',
     'docs/instruments.md', 'docs/run_states.md', 'docs/user_management.md',
     'docs/tag_semantics.md'

release 101.2.0 (2024-05-24)
 - remove dependency on HTML::Tidy
 - update NovaSeqX icon
 - Added a footer to the tracking run web page.

release 101.1.0
 - Deleted npg_tracking::illumina::run::short_info and
   npg_tracking::illumina::run::locations, their functionality is moved to
   npg_tracking::illumina::run::folder.
 - Moved a build method for the run_folder attribute from
   npg_tracking::illumina::run_folder to npg_tracking::illumina::run::folder.
 - Preparation for the deletion of unused instrument_utilisation table from
   the tracking db schema:
   - dropped instrument_utilisation table from the db schema, which is used
     for testing the tracking web application,
   - suppressed generation of the DBIx ORM class for this table and deleted
     the class.
 - Added an option to exclude lanes from a merge (see st::api::lims
   aggregate_libraries method).
 - Extended the basic search functionality of the tracking web application by
   enabling a search for lane annotations.

release 101.0.0
 - npg_tracking::illumina::run::short_info
     - Stopped parsing out run details from the run folder name. Deleted
       attributes, which were populated this way: 'name', 'instrument_string',
       'slot', 'flowcell_id'.
     - Deleted 'short_reference' method.
 - npg_tracking::illumina::run::folder
     - Removed dependency on 'short_reference' method, which used to be
       provided by npg_tracking::illumina::run::short_info.
     - Introduce a check to ensure that the run folder name on staging and
       in the database, when both are available, are the same.

release 100.0.1
 - For aggregation of lims objects by library in st::api::lims, do not
   consider the same library with different tag index in the same lane
   as a error. Chromium single cell ATAC libraries have four copies of
   each sample, each with a different tag.

release 100.0.0
 - To help monitor shadow run folders, include the run folder path into the
   logging messages in the staging monitor classes code.
 - To explain the discovery and handling of shadow run folders, added an
   extensive comment in 'bin/staging_area_monitor'.
 - Deleted npg_tracking::illumina::run::folder::validation role. Most of
   the functionality there depended on parsing run information from the run
   folder name. This parsing will be dropped in the next release.
   Added a simple check for mismatching run folder names to the find_live
   method in Monitor::Staging to replace a similar check in the deleted role.
 - Updated unit tests to ensure that they do not rely on retrieving the
   run ID and other run details from the run folder name.
 - Removed unused test file and run folders.
 - Deleted 'aggregate_xlanes' method from 'st::api::lims'. A more generic
   'aggregate_libraries' method in the same class can be used instead. The
   pipeline has been amended to cope with this breaking change, see
   https://github.com/wtsi-npg/npg_seq_pipeline/pull/815.

release 99.0.0
 - st::api::lims
     - Removed batch_id constructor argument. It was used by the xml driver,
       it is no longer required.
     - Fix a bug in propagating driver details to children to ensure that
       (1) a switch of the driver type does not happen;
       (2) the database handle and the flowcell ID are passed to children
           consistently.
 - Removed provisions for the xml st::api::lims driver from the old samplesheet
   generator npg::samplesheet.
 - To avoid code repetitions, created a parent class for different samplesheet
   generators and moved common code there.
 - Extended the code for the samplesheet daemon to generate samplesheets for
   NovaSeqX instruments.
 - Removed remaining unused test data from the era of xml st::api::lims driver.
 - Updated signature and iRODS documentation link in template for emails
 - Updated NPG tracking url to https protocol and port for links in emails
 - Removed link to illumina/runs qc page in emails reporting status change event
   to subscribers

release 98.1.0
 - Added a new method, latest_revision_for_modification, to
   npg_tracking::Schema::Result::Instrument to retrieve the latest modification
   revision for the instrument.
 - Added 'dragen_software_version' script argument
   for bin/npg_samplesheet_generator_NovaSeqXSeries. If not set and
   id_run is defined, defaults to the record for the latest 'Dragen'
   modification for the NovaSeqX instrument on which the flowcell will be
   sequenced.

release 98.0.0
 - Removed the use of 'xml' lims driver from all tests.
 - Deleted 'xml' lims driver class and all classes which supported this
   functionality, together with all related tests and test data.
 - In release 97.0.0, the default LIMS driver type in samplesheet generation
   was changed to 'ml_warehouse'. Following this change, the code for
   auto-generation of MiSeq default samplesheets is now simplified to exclude
   an explicit creation of LIMS objects.
 - Changes to st::api::lims, which are not related to the 'xml' lims driver
   removal:
     - deleted all previously deprecated methods ('seq_qc_state',
       'associated_lims', 'associated_child_lims', 'associated_child_lims_ia');
     - deleted methods for which we do not have data in ml warehouse, our
       primary source of LIMS data ('request_id', 'project_id', 'project_name');
     - deleted all code for computing inline tag indexes and inferring
       tag sequences and library types from sample description since
       definitions like this have not been in use for the last six years;
     - reimplemented the 'reference_genome' method to exclude a fallback to
       a study reference for tag zero and lane-level objects when samples
       have different references;
     - ensured 'required_insert_size' method works correctly for objects that
       are initialised with the 'rpt_list' attribute, ie have the 'position'
       attribute undefined;
     - added 'aggregate_libraries' method, which is similar to the existing
       'aggregate_xlanes' method, but instead of getting an instruction to
       merge libraries, discovers the libraries which can be merged and returns
       a list of merged entities and, separately, a list of singletons.
 - npg_tracking::illumina::run::long_info role:
     - restricted the 'all_lanes_mergeable' flag to NovaSeq Standard workflow,
       ie dropped an additional case of HiSeq Rapid Run since HiSeq instruments
       are no longer in use;
     - ensured the detection of NovaSeqX and NovaSeq instrument types is
       unambiguous.
 - Updated DBIx classes from the prod database. Changes are due to the MySQL
   database server upgrade to v8.+
 - Finding the runfolder path - stopped errors being raised when a runfolder
   is found in 'analysis' or 'outgoing' in the presence of a duplicate runfolder
   in 'incoming' (the duplicates in 'incoming' are sometimes created by
   instruments well after the run was mirrored and the runfolder moved to
   'analysis'). The duplicate runfolder in 'incoming' is disregarded'.
 - DRAGEN samplesheet generation - limit analysis to human samples only.
 - Fixed incorrect string comparison in the npg_move_runfolder script.
 - Ensure that pull requests to the master branch source dependencies from
   master branches (devel branch was hardcoded).

release 97.0.2
 - pin DBD::mysql to 4.050 to keep temp support for old mysql clients

release 97.0.1
 - update httpd.conf with new values for mpm event

release 97.0.0
 - Change CI workflow to standard file name
 - Change to Perl versions from Perlbrew
 - Add Perl 5.34.1 to the test matrix
 - Switched the default LIMS driver in samplesheet generation from 'xml'
   to 'ml_warehouse'.
 - Added 'lims_driver_type' attribute to npg::samplesheet.
 - Updated the tests for samplesheet generation, which are using the xml
   LIMS driver, to set the driver type explicitly.
 - In samplesheet generation, dropped a special case of MiSeq instruments
   without a batch; last use in production was over 5 years ago.

release 96.0.0
 - Fixed a regression in the npg_move_runfolder script,
   which made it unusable.
 - Rejig run folder tracking function into method
 - Stop using XML LIMS data in a test
 - Import samplesheet daemon
 - Ensure we have a loader to generate ORM classes
 - Flagged planned onboard analysis
 - locate_runfolder: simplify, cope with complex globs, link page for multiple
   locations
 - Allow NovaSeqX samplesheets to embed instrument and slot

release 95.0.0
 - Remove singularity recipe and related files
 - Import mlwh drivers
 - Reimplemented waiting for the onboard analysis
 - Added a samplesheet generator for NovaSeqX

release 94.0.1
 - Fixed bugs in the staging area daemon script:
   - an incorrect method name
   - use of previously deleted method

release 94.0.0
 - Added checks for on-board analysis in progress
 - Improved logging for a long wait case
 - Simplified run complete and mirrored detection
 - Simplified the staging area daemon.
   Bring the code in accordance with the current practice:
   - consider only one staging area,
   - narrow down the glob pattern for areas below the staging
     area to 'IL*'.
 - Removed roles which defined attributes that were used
   in one class only.
 - Changed the way the connection to the database is established
   by the staging monitor. Set up the connection explicitly in
   the script rather than inside one of the classess.

release 93.0.0
 - Introduced a new instrument format database record for NovaSeqX,
   updated test data and one of the tests. 
 - Fixed visual representation for 'cBot 2' model:
     the instrument now appears in the visual listing,
     the instrument does not appear in the list of sequencing instruments
     when a new run is created.
 - Added basic provisions for NovaSeqX run tracking:
     dragen_analysis_path attribute, platform_NovaSeqX method,
     correct identification of run completion.
 - Added basic provisions for analysis of NovaSeqX data, ie extracting
     relevant information from RunInfo.xml.
 - Added visual representation for NovaSeqX.
 - Removed the GAII model from the 'Instrument Types' filter on the runs
   listing page.

release 92.8.1
 - update version of github actions

release 92.8.0
 - change CI runner from Ubuntu 18.04 to ubuntu-latest
 - add ad-hoc api usage example for run annotation and tagging

release 92.7.0
 - add primer bed path to gbs_plex find.

release 92.6.0
 - modified is_i5opposite for NovaSeq v1.5 single-end runs

release 92.5.0
 - Fixed instrument name check.
 - Fixed the problem in the script that generates the Illumina instruments
   uptime report. A few month ago the lab stopped recording the washing
   cycle. As a consequence, statuses of some instruments do not get updated
   for many months. The instruments with no status within the report period
   were excluded from the report.
 - Simplify directory view for Tileviz data.

release 92.4.0
 - Staging area daemon - when the folder name is not in the database,
   prior to updating the database check that for the given run id:
      1. the current run status is "run pending",
      2. the instrument name is consistent between the database record
         and the RunInfo.xml file.
 - Switched the default st::api::lims driver from xml to samplesheet.
   Fixed tests, which implicitly relied on the xml driver being the default.
   Some tests used the default driver type. For those tests either test
   samplesheets were generated and used as a source of LIMS data or the
   driver type was explicitly set as 'xml'.
 - In samplesheet generation (npg::samplesheet) explicitly set the default
   st::api::lims driver to xml. This driver was previously used implicitly.
 - Apache https config:
     redirected all non-SSL requests to a port with enabled SSL
     removed provisions for self-compiled httpd server.

release 92.3.1
 - A WHILE loop in run_list.tt2 iterates over a number of pages (40 runs
   per page) in a visual listing of runs. The number of pages is now
   (January 2022) sufficiently high to trigger an error in Template Toolkit,
   which, by default, allows 1000 iterations only. The value of
   $Template::Directive::WHILE_MAX variable is, therefore, set to 2000.

release 92.3.0
 - New link in sigularity image to pass npg_tracking configuration file.
 - Removed ability of the tracking server to provide directory views of
   visible to the server /nfs partitions since nfs staging servers are
   no longer used.
 - Removed co-location functionality in npg_tracking views. Co-location checks
   will not make sense once npg_tracking runs in containers in head nodes.
   Staging areas are not local to head node/reachable from within the
   container.
 - Removed old team names from run add template.
 - Removed unused daemon runners for staging daemons and the Jenkins server.
 - Made NovaSeq instruments skip wash statuses after completing a run, this
   does not affect statuses set manually through the UI.
 - Added lab location to instrument pages. Instruments are also now 
   colour-coded on the graphical instrument list page.

release 92.2.0
 - Added singularity definition for the tracking web server.
 - Tracking web pages:
     added folder name column to run listing pages,
     added external instrument name to the instrument page.
 - Patched and refined a script for adding roles to users.

release 92.1.0
 - Change LDAP URL to read-only URL
 - Added a SOP for creating/decommitioning instruments.
 - Removed the samplesheet daemon definition npg_tracking::daemon::samplesheet
   and the auto-generator for samplesheets npg::daemon::samplesheet::auto.
   Both are no longer in use, the functionality was moved to
   https://github.com/wtsi-npg/npg_ml_warehouse, see
   https://github.com/wtsi-npg/npg_ml_warehouse/commit/0b98fc4defb8338914f8d11448df4b30cd75528c

release 92.0.0
 - Create 'primer_panel' alias for the 'gbs_plex_name' methods in
   st::api::lims. The new accessor will be used in contexts that are outside
   of processing GBS runs.
 - Remove the API wrapping around tracking XML API.
   This API has not been used in production for a long time. The part of it
   that links run id and batch id (npg::api::run) was used in tests. Tests
   are amended either to use samplesheets is a source of LIMS data or, where
   they are still using XML LIMS driver for st::api::lims, batch id is used
   explicitly.
   st::api::lims::xml previously had dependency on npg::api::run to allow          
   for retrieval of batch_id from run XML API in cases where it was not            
   supplied explicitly. This dependency is removed without changing the            
   public interface of the class. Formally speaking, the batch_id                  
   attribute remains optional for creating an instance of the class.               
   However, failure to supply it will result in an error when trying to            
   access batch XML API of the LIMS system.
 - Remove code for sending run status events to LIMS, this functionality is
   no longer required.
 - Remove st::api::event class, which is no longer used.
 - Stop generating XML output from the tracking server. Remove all templates
   for generating XML output, all associated code in views, tests and test
   data.

release 91.15.0
 - Remove status daemon. This daemon is no longer run since the pipeline
   updates run and lane statuses by writing directly to the database. 
   Remove the script, npg_status_watcher, for detecting run status JSON
   files in run folders and updating database run and lane statuses,
   which was run by the status daemon.
 - Removed live tests for XML and warehouse drivers. Live data can change
   at any time without notice, causing tests to fail.

release 91.14.0
 - Remove live tests for st::api::lims XML driver and other XML-based APIs.
   The CI tests are mostly run on hosts with no access to either production
   or development hosts. 
 - NPG_TEST_TRACKING_CONFIGPATH environment variable can be used to set a
   custom location of the Clearpress configuration file for unit tests.
   The default location of the file, data/config.ini, is used if the
   variable is unset. 
 - Move README file to README.md, improve Markdown format. Add sections
   describing environment variables and tags.
 - Move db schema dump, which is only used for tests, from the data to t/data
   directory.
 - Block access to tsv files via httpd_sfweb.conf

release 91.13.0
 - Extend npg_status2file script to optionally save statuses to the tracking
   database and renamed the script to npg_status_save.
 - Move from Travis CI to GitHub Actions.
 - metaref_repository attribute is added to the reference repository API

release 91.12.0
 - Change instrument and instrument format pages to only show current
   instruments and instrument formats
 - Remove superfluous columns in display tables for lists of instruments
   and instrument formats
 - Remove unused views (templates) for an instrument manufacturer
 - Remove manufacturer model and view classes, create manufacturer_name
   method in the instrument and instrument_format model classes

release 91.11.0
 - a partial reimplementation of the serialization framework for components
   and compositions in order to speed-up frequently used methods 
 - add geno files to disabled access list in http_sfweb.conf
 - cookie decryption - use an up-to-date Crypt::CBC API and algorithm
 - reduce the sleep type of the status daemon from 60 to 1 sec in order to
   reduce the time it takes to loop over the run folders on lustre 

release 91.10.0
 - script npg_samplesheet4MiSeq is removed this utility has been moved
   to a different git package; the samplesheet daemon stays in this
   package till all MiSeq instruments are redirected off old staging servers 

release 91.9.0
 - Change label 'New Instrument Mod' to 'New Instrument Upgrade' on the Admin page
 - Remove 'New Instrument Status' and 'New Run Status' options from Admin page
 - Sort the dropdown menu of instruments on the 'New Run' page
 - Staging servers usage page retired

release 91.8.0
 - BotSeq renamed Duplex-Seq
 - added new functions to long_info for BotSeq analysis
 - added support for NovaSeq v1.5 reagents, i5 index sequenced in opposite direction

release 91.7.0
 - remove the template for an unused tracking server url - run/summary -
   and all underlying code to remove obsolete SQL queries used by that code

release 91.6.0
 - enforce custom type NpgTrackingReadableFile to check that the
   given path is not only readable, but is actually a file
 - discontinue 'npgdonotmove' and 'npg_do_not_move' file and
   directory names being used as flags to stop moving the run folder
   from analysis to outgoing

release 91.5.0
 - ignore new revision field optionally added to primer panel name in LIMs.

release 91.4.0
 - disable downloading *.fa (consensus) files from tracking server
   views of staging folders
 - add executable for npg samplesheet4MiSeq wrapper
 - st::api::lims - two new accessors, 'sample_is_control'
   and 'sample_control_type'

release 91.3.0
 - pp_archive_path runfolder accessor - a path to an archive directory
   for third party portable pipelines

release 91.2.0
 - add support for finding primer_panel bed files

release 91.1.0
 - staging monitor changes for SP flowcells
 - Update URL for production Sequencescape LIMS
 - method for creating a lane-level st::api::lims object from
   any other st::api::lims object where the resulting object has
   the same driver as the original one
 - add support for finding primer_panel bed files 

release 91.0.0
 - modified is_i5opposite when there is no reverse read
 - removed scripts for monitoring instruments via ftp; related
   modules and methods, their tests and test data are removed
 - code for staging monitor reorganised so that the Monitor::Runfolder
   class has methods for updating run, run_lane records and run tags
   and the Monitor::Runfolder::Staging class mainly deals with staging
   folder inspection and changes; some methods renamed to remove
   ambiguity about their functionality
 - some of the functionality of the instrument monitor is incorporated
   into the staging monitor script code, namely, (1) setting 'multiplex',
   'paired_read' and 'single_read' tags and (2) deletion of superfluous
   lanes; this functionality is now available for NovaSeq instruments,
   while previously it was missing for this instrument type
 - redundant 'rta' run tag is not longer set
 - redundant Mirror.completed file is no longer created in the run
   folder
 - functionality for checking instruments' wash status and setting
   'was required' instrument tag is lost; this shoud be done outside
   of staging monitor
 - npg_tracking::util::pipeline_config:
     move bqsr and haplotype caller to a tertiary section which can be
     set specifically by reference or by using a default study config;
     restrict access to attributes which should not be set by the caller
     (local_bin, product_config);
     enable setting product release configuration file path via the
     new product_conf_file_path attribute
 - genome reference hierarchy in the current samplesheet is not compatible with
   the latest version of MiSeq Illumina software; setting Workflow value to
   GenerateFastQ prevents the instrument's software from using this value,
   hence we can stop setting genome reference path altogether
 - make impossible to download cbcl file via tracking web pages
 - runs_on_batch method of npg::run::model is refactored not to cache
   the return value; proper input checking is added
 - error on run creation if the batch id associated with the run is already
   associated with one of active runs, i.e. a run that has no current status
   or its current status is not either 'run cancelled' or 'run stopped early

release 90.3.0
 - npg_tracking::glossary::moniker -
     added a method for file name parsing to retrieve id_run,
     position, tag_index, suffix and extension;
     private attributes are prevented from being serialized under
     MooseX::Storage framework
 - tracking server instrument visualization - remove visual cue
   for connection delay since ftp-ing to instruments is discontinued

release 90.2.1
 - staging area daemon to save the flowcell barcode to the
   database if it does not have this value already available 

release 90.2.0
 - reference finder for compositions not to return
   a PhiX reference when at least one other reference
   is found
 - staging area daemon will validate database flowcell barcode
   against the value in files in the run folder; in case of mismatch
   the run will not track

release 90.1.0
 - no_archive attribute for run folder API

release 90.0.0
 - cleanup and simplification of runfolder subpaths logic
 - add uses_patterned_flowcell attribute to long_info

release 89.1.0
 - Where semantic names are not possible, standard file and directory
   names are based on sha256 digest of the composition JSON
   (previously the md5 digest was used). This makes possible
   conversion from the name back to the composition using the ml
   warehouse records, which might be useful in case of clients'
   queries.
 - Unused run folder path accessors removed (reports_path,
   lane_archive_path, lane_archive_paths, lane_qc_path,
   lane_qc_paths).

release 89.0.1
 - live test patch to reflect a change in spiked PhiX tag index

release 89.0.0
 - the samplesheet driver's parser extended to cope with data from
   multiple runs
 - samplesheet generator extended to cache data for multiple runs
 - status monitor will now work only with RFC 3339 compliant
   timestamps defined in WTSI::DNAP::Utilities::Timestamp
 - pipeline configuration module is moved to this package from
   npg_seq_pipeline package in order for the product configuration
   be accessible from other packages 
 - staging area daemon modified so processing for each folder is wrapped as
   try/catch block

release 88.2.1
 - patch to accommodate different URL info passed to the CGI
   module by different Apache httpd versions

release 88.2
 - scripts to list run folders that are candidates for deletion in
   the Data Recovery area
 - status monitor: improved logging, which now shows timestamps
 - alleviate differences in cloud hosts time zone configuration by
   storing UTC offset as a part of the timestamp in serialized
   run and lane status objects

release 88.1.1
 - test data fixes following ml warehouse schema changes, see
   https://github.com/wtsi-npg/ml_warehouse/pull/104 for details

release 88.1
 - status monitor running with inotify disabled should be able to
   recognise when a cached file is changed
 - 2 sec sleep introduced in the beginning of the job that saves
   statuses to a file to ensure no two statuses of the same run/lane
   have the same timestamp so that the iscurrent db flag could be
   assigned inambigiously when the statuses are saved to a database

release 88.0
 - tracking database:
     add column `incompatible_tag` to the tag table;
     populate `incompatible_tag` column with values previously
       hardcoded in a DBIx class;
     add tags for instrument workflow;
     simplify code in DBIx classes for setting and unsetting tags;
     drop requirement for a valid user id when unsetting a tag;
     drop a check for tag existence when unsetting a tag
 - staging monitor to set/fix tracking db record about instrument side and
   workflow type by setting/resetting appropriate run tags
 - Change staging area monitor so for NovaSeq runs, both CopyComplete + RTAComplete
   or RTAComplete + max waiting time must be true before a run can be considered
   copy complete.

release 87.4
 - changes to support geno_refset repository
 - moved method for creating full file names from npg_tracking::product
   to the moniker role
 - extension of the composition/component serialization role to ensure that the
   serialised objects can be deserialized
 - tracking server run page links to run folder locations:
     'Latest Analysis' and 'Latest IVC' links removed since the target files
     are no longer generated;
     'Tileviz' link is updated to point to a location in the new-style runfolder
 - 'locate_runfolder' cgi script:
     code to handle redundant links and paths is removed;
     if the target tileviz directory is not available, fall back to the one
     expected for the old-style run folder
 - redundant filename method of the illumina component object removed
 - status monitor:
     two modes - with and without inotify;
     option to set the prefix from command line so that multiple monitors
       could potentially run on the same host, each monitor dealing with
       its own staging area;
     avoid repeatedly saving the same statuses
 - when searching snv_path use bait_name=Exome for (cD|R)NA library types
 - tracking database: 
     add a unique db constraint for all types of status descriptions;
     make temporal_index column in run_status_dict table non-nullable and
     add a unique constraint on it;
     simplify DBIx binding for statuses;
     add a method for temporal comparison of run statuses

release 87.3
 - for a run on ESA, redirect to an ESA url

release 87.2
 - added method to find file with MT genes inside transcriptome repository
 - change attributes in transcriptome-find to use lazy and builder rather than
   lazy_build to conform with Moose's best practices
 - sequencing instrument checker:
     treat error reading a recipe file as non-fatal;
     adjust warning message generation to use methods supported by the poller
     object instance
 - a new factory method, create_tag_zero_object, to create st::api::lims object
   for tag zero
 - a new factory method, aggregate_xlanes, to create a list  of st::api::lims
    objects for merges across lanes

release 87.1
 - bug fix in JavaScript for run creation - the reflector cannot
   be used any longer, allow client side JS to request XML feeds
   directly from LIMs server

release 87.0
 - new role - file and directory naming factory
 - added bin/npg_move_runfolder script
 - make NovaSeq instruments represented on web pages
 - removing reflector allow client side JS to request XML feeds directly from LIMs server
 - use lims_url config file entry to set LIMs server location
 - gseq cluster hoop jumping for test 10-gbs_plex.t
 - use 'fasta' as default aligner when searching for references
 - long_info role:
     when parsing RunInfo.xml look for FlowcellLayout rather than version,
     remove unused methods and attributes,
     use run parameters file to retrieve additional information

release 86.10.1
 - following changes to npg_tracking::illumina::run::short_info role,
   patch staging monitor, which started to fail if run id could
   not be retrieved
 - other changes to staging monitor:
     improve reporting
     change the order when validating run folders, will try
       to retrieve run id first, then validate the name
     propagate db schema handle
     use db schema accessor that is used by npg_tracking::illumina::runfolder
       and its roles
     remove custom code where existing APIs can be used

release 86.10
 - the bug in paged run views where a particular instrument was selected
     is fixed
 - the frame height for displaying runs on an instrument page is increased
 - displayed instrument statuses in the instrument view are limited
     to most recent (created within a year from the current date)
 - npg_tracking::illumina::runfolder is simplified at a cost of losing
   ability to pass to the constructor an arbitrary path
 - npg_tracking::illumina::run::short_info role - added database
   look-up aided ability to deal with arbitrary run folder names
 - use icons from the latest set of brand images
 - display institute's logo and favicon only if pages are served
     from institute's servers
 - removed unnecessary aliasing of urls
 - deleted unused (dif_files_path and pb_cal_path) runfolder accessors
 - method to find file with globin genes inside transcriptome repo added
 - changes to support gbs_plex repository

release 86.9
 - to account for ephemeral staging areas on OpenStack, extend staging
   directory path match expression in Apache httpd config file
 - use bash shell when executing system calls from Perl
 - copied srpipe::runfolder module (from SVN data_handling package)
   as npg_tracking::illumina::runfolder to this package, removed tests
   that depended on test folder hierarchy in the data_handling package

release 86.8
 - remove redundant script for reporting sequencing failures to RT
 - remove uptime info from instrument page - function no longer available
 - correct handling of transition from 'down for service' to 'down for repair'

release 86.7
 - stop building on Travis under Perl 5.16,
   see https://rt.cpan.org/Public/Bug/Display.html?id=123310
 - reference find role: when parsing a genome reference string
   return the analysis, along with the organism, strain
   and transcriptome version, if present
 - refactor transcriptome::find to improve code reuse and add
   new find file/path/index attributes and functionalities
 - when inflating rpt key, cast id_run, position and tag index
   (if any) values to integers and raise an error if conversion
   gives a warning
 - unused tracking server features removed:
     simultaneous status update for multiple runs;
     simultaneous status update for multiple instruments;
     simultaneous mod update for multiple instruments;
     XML feeds for individual run statuses with listing of
       all runs at this status;
     flowcell verification;
     complex computation of width of padding id_run in run name
       for display purposes

release 86.6
 - unused db tables, which were renamed in the previous release,
     are dropped; exclusions in teh DBIx class loader script removed
 - Jenkins daemon - a bug in locating prod ssl server certificates fixed
 - added a convenience factory class to transform rpt list strings
   into a composition

release 86.5
 - small fix in locate_runfolder for rna seqc links
 - remove unused classes for image generation
 - be compatible with ClearPress version 475.3.3 and above
 - set explicit dist to precise in travis yml
 - add dual configuration for apache mpm worker/event
 - Jenkins' daemon with HTTPS as default configuration
 - drop or rename orphaned tables from db
   - drop
     cost_code
     cost_group
     sensor_data_instrument
     sensor_data
     sensor_instrument
     sensor
     mail_run_project_followers
   - rename
     event_type_service to todrop_event_type_service
     ext_service to todrop_ext_service
     event_type_subscriber to todrop_event_type_subscriber

release 86.4
 - reference find role: remove unused 'subset' attribute and
   previously deprecated 'single_ref_found' method
 - increate http tmeout for event notification
 - run tracking server on two ports - one accepting http traffic, another
   one https only; the same set of URLs on both ports
 - tracking web GUI link to request LIMs page now changed to point to
   a more useful lane page; request id will not be available if XML
   LIMs feeds are discontinued
 - link to auth script residing in the same virtual host
 - sensor-related code removed
 - mkfastq option added to make a samplesheet suitable for 10x mkfastq

release 86.3
 - event notification - do LIMs notification first; if it fails and is run
     multiple times as a consequence, users are not spammed

release 86.2
 - event notification:
     use faster ml warehouse st::api::lims driver (ml_warehouse_fc_cache)
     find templates directory relative to the running script bin
     fix npg tracking server url in reports' text
     filter out undefined email addresses
     bug fixes
 - remove models for db tables that are due to be dropped
 - update apache conf so server run with 1 thread/child w/mpm_worker_module
   Config::IniFiles used by ClearPress::util does not seem to be thread safe
 - replace image generation/manipulation routines by a stub
   returning a simple static image (these routines are used
   in npg_qc in web pages that are no longer used).
 - st::api::lims helper over boolean values for a pool fixed to work
   correctly for compositions

release 86.1
 - update models, droping cost code and cost group
 - updating templates to remove broken R&D watermark based on cost_code
 - deleting fixtures and tests for cost code and cost group
 - under modperl information about request url is not available in
   env. variables; now getting it directy from the CGI module
 - when possible, any subsequent requests should go to the server
   from which the original page was served (makes sense in general
   and helps to set up proper dev environment)
 - web pages - drop buttons linking to non-existing pages

release 86.0
 - httpd config files for httpd v.2.4.23 and mod-perl v.2
 - simple uptime report for Illumina sequencing instruments
 - reinmplementation of events reporting
 - timing of emails to followers is changed from statuses
   'run complete' and 'run archived' to 'qc review pending' and 'qc complete'.
 - ClearPress library is pinned to v. 473.0.5

release 85.6
 - code and tests changes to reduce number of warnings under Perl 5.22.2
 - use mysql 5.7.13 in travis (installing from deb)
 - Run web page
   - removed image browser link
   - Add EBI submissions link
 - Build.PL fix to ensure that images under htdocs dir are copied to blib
   when building.
 - updated links to NPG QC since the pages moved to SeqQC

release 85.5
 - removed all code that relates to instrument utilisation
 - removed an option to create database entries via XML POST.
 - removed custom error view
 - deleted special privisions for the 'pipeline' user
 - removed unused xml views for run lists
 - Install all files, including cgi-related, within the install action.
 - Remove instrument utilisation views for web pages.
 - Deprecate 'request approval' instrument status, enable status change
   from either of down statuses directly to 'wash required'.
 - Remove code and tests for deprecated instrument statuses.
 - Create flags to mark users and usergroups as deprecated,
   deprecate 'approvers' group.
 - User and usergroup lists are alphabetically sorted and include only
   current entries.
 - Fixed instrument status menu for batch updates to display only
   current statuses.
 - Composition-enabled LIMs object and reference finder.
 - Subset support added to the rpt key based composition factory.
 - Objects's attributes based composition factory sets only those components
   that are either defined or required. Previously an undef value might
   have been used for a component attribute even if the value did not have
   to be set, i.e. the code was not future compatible with planned removal
   of 'Maybe' qualifies from the tag_index and subset definition.

release 85.4
 - Run/instrument tracking; added full support for HiSeq 4000 and limited support
   for NextSeq and MiniSeq
 - Add HiSeq 4000 to graphical instrument list
 - Limit menu for changing run status, when not analyst, to those on instrument
 - Composition framework - the composition object becomes immutable as long
   as its components are immutable.
     - Functionality for an incremental assembly of a composition moved to
       a generic composition factory.
     - Empty compositions are not allowed.
     - Composition constructor returns a sorted composition.
     - Public sort function is removed.
 - Composition - find is implemented as a binary search.

release 85.3
 - Add deployment to GitHub Releases
 - In order to have correctly versioned files in a distribution,
     changed the builder base class.
 - Travis conf update to pre-load perl-dnap-utilities from github.
 - Samplesheet daemon:
     fixed logger;
     changed reference repository root so that it can be run on a staging host.
 - Changes to handle a different number of tiles in each lane
 - Increased TAG_INDEX_MAX to 9999 to handle pools with >999 samples

release 85.2
 - Added 'sample_accession_numbers' to lims shim
 - Deprecated old accessors for qc outcomes.

release 85.1
 - Added 'purpose' to lims shim
 - utf8 option for test sqlite db

release 85.0
 - lims shim
   - endeavours to ensure consistency of primary info (e.g. id_run, position)
     with drivers
   - will pass through arguments to driver constructor if it takes them

release 84.12
 - composition and component serialization to rpt

release 84.11
 - Operations (inflation and deflation) on rpt strings and their hash
   representation and lists of rpt values (based on npg_qc::autoqc::role::rpt_key)
 - Factory to create composition from a string representation of an rpt list
 - New filename method for the illumina component
 - Alter "run complete" email alert text to make clear analysis and qc time

release 84.10
 - create multi-attribute methods automatically
 - added method st::api::lims::any_sample_consent_withdrawn
 - removed potential warnings in event notification RT#495946

release 84.9
 - transcriptome find updated for RNA-SeQC directory and its gtf file
 - generic serializable sequence composition framework and illumina
   sequencing specific component
 - lims shim now works out driver type from a given driver
 - sequencing failure RT emails cope with GCLP (no batch_id)
 - add predicates for glossary roles

release 84.8
 - jenkins daemon definition: remove hardcoded path to prod user home directory
 - refactor snv finder to stop warnings when ref genome is undefined RT#481506
 - make function for parsing reference gemone public since it is used outside
   of the module where it is defined
 - use RTAComplete file in fallback local (instead of ftp) determination of run
   progress (which is consistent with ftp process and required for HiSeqX)

release 84.7
 - accommodate LIMs-defined dual indexes
 - Illumina samplesheet generation: do not truncate single index when shorter first
     part of dual index is present; this was written with spiked Phix in mind,
     however, spikes are not reflected in LIMs for MiSeq runs

release 84.6
 - when changing ownership of run folder, set gid bit on every level
 - reporting croaks without full stacktraces for folder and folder/location
 - daemon control:
     do not use deprecated load_class method
     read a list of production users from a configuration file

release 84.5
 - staging monitor extension:
     move qc complete runfolders from analysis to outgoing
     propagate group ownership inside the runfolder
 - staging/instrument monitor:
     remove superfluous run lanes
     assume all runs are RTA
 - perform runfolder name validation against the tracking db run record
   rather than against xml feeds; compare to expected runfolder name
   if this name is not in the database yet
 - add sample_supplier_name, sample_cohort and sample_donor_id to
   st::api::lims interface

release 84.4
 - to avoid the need to create deep directory hierarchy containing www-live
   for a live server, set dev env variable value for the Apache server
   explicitly

release 81.3
 - accessor methods for frequently used configuration file entries
 - status monitor script and staging area daemon definition - allow for
   farm-specific variations of path prefixes for staging areas
 - staging area and status daemons use common log directory /nfs/{gs|sf}XX/log

release 81.2
 - web server configuration generalisation to allow for running on
   different clusters
 - cross-cluster staging links for both tracking and seqqc links
 - make sample name compalsory in samplesheet generation
 - provide path substitution mechanism to cope with running code on a file
   server where the abs_path may differ from the cannonical network path.

release 81.1
 - extended samplesheet contains only the Data section, Illumina
   on-instrument reference paths removed
 - factor out parsing database configuration files into a separate
   module
 - when using MySQL database in testing, do not use database configuration
   files implicitly from the user's home directory; force the caller to
   supply the location of the file
 - remove default repository root from npg::samplesheet
 - no longer stop runs without batch id or with 13-digit entries in the
   batch id field from moving from run mirrored to analysis pending (
   allows automatic analysis of GCLP and QC runs reespectively).
 - optionally change group ownership of certain analysis directories on
   moving runfolder form incoming to analysis

release 81.0
 - remove default config file
 - allow environment variable NPG_REPOSITORY_ROOT to set repository root
 - move config of staging areas and repository into a config file
 - remove vestigial code for dealing with manual qc state
 - remove good_bad column in run_lane table
 - remove manual_qc_status table
 - npg daemon 'local' host option avoids ssh

release 80.11
 - removed staging area sf48

release 80.10
 - default to in-memory option for SQLite in tests
 - added a method to transparently return live or dev LIMs base URL
 - reference genome values - trim white space

release 80.9
 - do not try to infer phix reference from library/sample name
 - correct regular expression for modern MiSeq reagent kits - previously
   part of the id starting from teh dash was not returned by the flowcel_id accessor
   in the short_info role
 - samplesheet generation: library id shoudl be defined for all samples

release 80.8
 - fix memory leak in the samplesheet driver for st::api::lims -
   pass data hash by reference

release 80.7
 - change reference repository host

release 80.6
 - reenable and clean up cBot monitoring
 - try to avoid recursion in st::api::lims
 - cope with no reference genome info in XML
 - provisions for ml_warehouse driver
 - change host for development tracking server
 - when building the package, change version in existing directories only

release 80.5
 - remove redundant status update method from DBIx run_lane table binding
 - remove unused methods from npg::api::xx objects
 - remove auto-change of run statuses RT#420151
 - status monitor enhancements RT#420193:
     cope with runfolder move from outgoing to analysis
     on detecting LatestSummary creation, check for existing status files
       in the status folder

release 80.4
 - list of staging areas updated
 - reset monitor to use local disk if FTP to instrument fails

release 80.3
 - report sample_name, library_id etc at pool level if non-spike children
   give same value (to better cope with single plex pools)

release 80.2
 - cope with old path being reported for a directory when it is moved

release 80.1
 - glob expression fix for running on Ubuntu lucid with perl 5.14.2

release 80.0
 - DBIx run and run-lane status update: ensure duplicate records are not created
     allow to create a record for a status that is already current if the date
     of the new record is earlier that the current status date
 - status monitor bug fixes
 - daemon for status monitor watcher
 - propagate child/plex library level study data "alignment_in_bam" and
   "separate_y_chromosome" to parent lane level as is done with
   "contains_nonconsented_human" data

release 77.9
 - DBIx run and run-lane status update refactoring to accommodate passing
     status date and make setting these two status types symmetrical
 - initial code for monitoring status files
 - bug fix in inferring recalibrated path from the runfolder path - previously
     the code returned basecalls directory path before getting to the recalibrated path
 - following psd dev url change, update regular expression in the reflector url filter

release 77.8
 - npg dev url is set to the url of the dev server that is actually used for API calls
 - psd next release dev url replaced with training url, RT#412417
 - wrapper object for the status and a script to write status to a file
 - dual index refinements: ensure max length of "Index" used by MiSeq is the split
 - dual indexes' size, and use custom default_expected_tag in preference to Index{,2}
   when using samplesheet as source of LIMS info
 - ensure samplesheet testing works without WTSI filesystem
 - instrument location tests skipped if host name is too long
     (TODO - increase db field size)
 - travis ci configuration file added
 - mysql db test configuration - password removed so that this configuration
     can be used on travis ci
 - implicit dependency on DBD::mysql driver made explicit to ensure all
     dependencies are built on travis ci server

release 77.7
 - run in progress tracking:
     look for lane sub-directories in the BaseCalls directory if none found in the Intensities directory;
     check for existence of a RTAComplete.txt file rather that a ImageAnalysis_Netcopy_complete_Read#.txt file

release 77.6
 - added code to generate HiSeqX runfolder names
 - updated list of directories to search for cycle sub-directories

release 77.5
 - added new version value for HiSeqX RunInfo.xml

release 77.4
 - changed list of top level directories for HiSeqX

release 77.3
 - samplesheet creation now supports dual indexes
 - sensor.pm now uses npg::api::request to allow for caching xml feeds when
   retrieving sensor data
 - web app and database adjustments for HiSeqX
 - callbacks for uptime graphs removed
 - removed the uptimes table from the textual listing of intruments page -
     made page loading very slow
 - removed upgrade schema scripts and redesign proposal file from MANIFEST
 - merged instrument checker scripts for all existing sequencing instrument types
 - removed instrument checker scripts for invividual seq instrument types
 - all instrument checkers to use local libraries if available

release 77.2
 - ensure module and package version starts with a digit
 - npg_tracking::illumina::run::folder and its tests updated; the following accessors removed:
       data_path (unused from outside),
       tag_qseqs_path (redundant),
       score_path (redundant),
       qseq_location_path (redundant);
     tests specific for old illumina mashines (IL) removed;
     unused analysis_path accessor reimplemented to point to the top-level custom analysis folder
       paths from analysis_path removed;
     definitions of most paths placed together
     public write accessors for bam_basecall and dif_files paths since they are used from the pipeline,
       old private accessors retained for backwards compatibility

release 77.1
 - transcriptome changed so that a default is not used if not supplied
 - transcriptome version directory changed so tests changed accordingly
 - reference find changed to handle transcriptome version parsing
 - unused templates removed
 - samplesheet regenerated when the existing file does not come from the currently
   pending run
 - samplesheet daemon moved to a staging server where samplesheets are stored
 - copyright for all modules and script belongs to GRL - the copyright
    notice edited where it was incorrect
 - pass Illumina samplesheet study name field components through a filter to replace
   commas
 - use git describe directly to avoid need for gitver script

release 77.0
 - the build file and module - code improvements, POD and PBP for the module
 - scripts and modules:
     version variable added where needed,
     RSC keywords removed,
     superfluous empty lines, 'Readonly' module import and other features removed,
     where needed, 'use Readonly' added
 - tests - remains of RSC keywords removed
 - the distribution test does not perform pod tests (separate tests available),
   checks that module version matches distribution version
 - allow staging daemon to check progress by looking for .bcl.gz files (as
   well as .bcl files)

release 76.5
 - code action on Build only if install target specified
 - global removal of [LastChanged]Version from POD; change VERSION svn declaration to ='0';
 - added modules to handle locating of transcriptome files required by tophat2
 - updated reference module with additions needed by transcriptome
 - added new test for transcriptome
 - updated reference test to check that transcriptome directory structure exists
 - added files from a transcriptomics run for use in test

release 76.4
 - replace uri_escape() by uri_escape_utf8()
 - some unused code removed
 - some small test scripts merged into related larger test scripts
 - VERSION variable and headers with SVN properties removed from test files
 - all test scripts are given .t extension
 - unnecessary 'use' statements removed from tests
 - server-wide definition for modperl custom perl libraries location in order
   to provide consistent perl environment for all scripts at server start-up
   time

release 76.3
 - use staging area for staging daemon logs

release 76.2
 - lims samplesheet driver bug fix - spiked phix tag index should be defined for plexes
 - method to return the name of the env. var for the cached samplesheet file
 - misleading xml lims driver POD entry fixed

release 76.1
 - use scratch110 for references as scratch109 is hanging
 - extend timeout for Jenkins auto-logout

release 76.0
 - added defaults for Perl indentation in Emacs for this repository
 - definition of the repository root for single nucleotide variation files
 - a new module defining rules for finding single nucleotide variation files
 - changes to the database schema and code to ensure compliance with sql
   strict mode, relevant changes to tests and test data
 - if batch_id is not supplied at run creation, enforce zero in the model,
     wider tests to be performed if NULL needs to be used
 - id_run_pair defaults to null; examining and setting this field removed
     from the code, the concept of run-pair has not been used for a long time
 - use ${JENKINS_HOME}/tmp instead of /tmp for jenkins tmpdir
 - add perlcritic policy to prohibit tabs, convert tab to two white space, correct
   indentation issues
 - old and unused apache configuration files removed
 - IL instruments removed from instrument utilisation menu (we do not have them any more)
 - instrument uptime menu link removed - target graph is broken
 - consolidated all instrument menu 70* tests in one 20 test
 - removed code repetition from bin/ga_II_checker
 - removed deprecated last_30_days method from npg::model::instrument_utilisation
 - fixed the code where in test conditions operations on undefined variables
   were performed
 - removed excessive verbosity in most of the tests
 - documentation in README, INSTALL and REDESIGN_PROPOSAL updated

release 75.6
 - extended inline index to read 2
 - hardcoded link for enigmatic authentication (not mod perl url) page so that the page
   is accessible from the server on sfweb, where enigmatic is not deployed;
   uncomment old code to restore the relative to the server root link for testing
 - apache2.2 compatible configuration files to run the tracking server under
   default apache version for Ubuntu precise; virtual host definitions in
   stand-alone files allow for dynamic inclusion of virtual hosts
 - fixed occasional authentication problem (inability of a particular thread to load
   additional modules due to the fact that on a second invocation of npg script in the
   same server thread @INC array is not extended since BEGIN block is not run) when
   running under apache2.2 by extending the server @INC array at server start up time
 - bam, cram, etc files access is prohibited from the runfolder directories listings,
   however, these files are visible in listings
 - more convenient sorting order of directories and files in the runfolder directories
   listings
 - directory index CSS changed to get a margin on the left; brighter color scheme
   introduced
 - Build.PL file extended to ensure that the shebang line of CGI scripts is
   set to the Perl interpreter that is used for deployment
 - explicit setting of PATH is removed from CGI scripts that do not use shell commands
 - SVN-dependant VERSION removed from CGI scripts
 - where necessary, PERL5LIB of CGI scripts is extended
 - links to staging area server are picked up from config "staging_url" parameter,
   and so are relative (to current host and port) if this is empty or non-existent
 - switch to samplesheet driver if NPG_CACHED_SAMPLESHEET_FILE is set
 - spiked_phix_tag_index accessor implementation for the samplesheet
   driver to ensure that spiked_phix_tag_index is defined for a pool

release 75.5
 - on Ubuntu precise daemon .pid files in /tmp directory get deleted after a week,
   making daemons non-responsive; change modification time of the .pid
   file when pinging the daemon, which is done at least once a day

release 75.4
 - change of default reference repository location

release 75.3
 - default UTF8 output for samplesheets
 - add number of lanes to run list views
 - new farm host for a samplesheet daemon

release 75.2
 - host-agnostic jenkins daemon
 - version of npg::api::request module updated to help tracking requests

release 75.1
 - patch for samplesheet parser
 - switch live npg api base URL to a different host in order to separate
     human- and api-driven requests
 - removed deprecated methods from npg::api::run_lane
 - removed unused save2cache method from npg::api::util
 - simplified logic for saving to cache in npg::api::request;
     removed unused save2cache method
 - staging area indexes defined in single place

release 75.0
 - authentication through npg own enigmatic.cgi residing relative to
   where the request comes from
 - updated expected result for a live reference retrieval test following
   a change in the reference genome study field
 - add new staging areas 52 to 55
 - don't bail out on finding no runfolders when FTP'ing to instrument for update info
 - sample sheet generator extended to include npg specific fields and deal with multiple lanes
 - use URI escaping for samplesheet contents instead of squashing whitespace and flipping , for |

release 74.14
 - minor bug fix to lims sample_description

release 74.13
 - data_access_group method to lims (study) to determine data access control requirements
 - add new staging area (50)

release 74.12
 - to ensure host-specific path is used when calling a script, use bare script name in staging
   daemon definition

release 74.11
 - samplesheet lims driver moved to st::api::lims namespace, contains parser only
 - daemon utility: execute daemons in a bash login shell in order to pick up
   target host environment; remove setting PERL5LIB and perl version
   from definitions of daemons for samplesheet and staging
 - removed Makefile.PL for web deployment - this functionality is now in Build.PL

release 74.10
 - update live URL for getting/posting lims data as requested in RT#336085
 - code to find data repository directory for tag set files

release 74.9
 - Remove TT interpolation to create Perl code (hack prone and caused some pages to fall over)
 - update links to NPG QC web pages

release 74.8
 - rationale for using taint mode added to the apache conf file
 - remove unused, mispelt and misleading code from st::api::lims
 - switch to scratch110 for references while scratch109 is down

release 74.7
 - deal with EAN13 barcode - put in batch_id field for now and allow creation of
   samplesheet using warehouse lims driver
 - separate_y_chromosone_data available in lims
 - samplesheet driver for lims object
 - general tidy-up of the xml lims driver
 - logic for sample/study_publishable_name methods moved tost::api::lims wrapper
 - reduce spurious warnings from Clearpress related code and simplify
 - JS and CSS internal to server (browser https security no longer breaks JS)
 - links to Sequencescape dev or production as appropriate

release 74.6
 - add new staging area (51)

release 74.5
- order run_status_dict by temporal_index to display in Runs horizontal panel.
- restrict statuses in menu for status editing to the statuses after the current status for all users
  other than those in the new analyst group, who can see all statuses.

release 74.4
 - memory leak fix in st::api::lims constructor - move metaprogramming out of the BUILD method

release 74.3
 - children method of the st::api::lims backed up by a lazy-built attribute;
   once children objects are created, their driver peers are cleared
 - new npg url is propagated through the package
 - use updated DBIx::Class::Schema::Loader
 - don't define new constructor now DBIC is Moosified: use BUILD

release 74.2
 - xml specific lims module, st::api::lims, is split into a generic lims wrapper,
   st::api::lims, and xml parser, st::api::lims::xml
 - a new parser/generator for Illumina-style samplesheets,
   npg_tracking::illumina::run::lims::samplesheet
 - bait finder update to cope with white space around bait name RT#334881
 - fixed test failures for Monitor::RunFolder::Staging by making the timeout
   the object is using variable and setting it to a small value in the test;
   this allows for testing a fresh clone

release 74.1
 - improvements and patches to tests
 - apache httpd conf file for running tracking server on centrally supported
   ubuntu precise vm
 - build script extended to deploy wtsi_local folder, which has apache conf files
 - in tests, explicitly clear existing db connection before resetting the tables and loading fixtures;
   otherwise, an very long wait (as set by GLOBAL wait_timeout ) occurs, see RT#336398
   for some reason this wait only occurs on MySQL servers v 5.5.30 and above
 - correct skip message for a test that accesses live reference repository

release 74.0
 - removed unnecessary separate log from apache configuration file

release 73.9
 - cgi scripts that are linked from the main tracking web app are moved to this
     project together with an apache configuration file for servers on sfweb nodes
 - apache configuration file for sfweb nodes updated:
     bed files are not visible
     all tracking-related services sre run on port 9000, port 9080 is not used
     the cgi-bin url component is mapped to the cgi-bin directory underb the npg
       root /software/solexa/npg, where all npg cgi scripts are deployed to
     npg own modules are taken from the default deployment location,
       /software/solexa/npg/lib/perl5
     htdocs directory is mapped to the deployed /software/solexa/npg/htdocs

release 73.8
 - make dependency on IO::All::FTP explicit

release 73.7
 - propagate caller environment to scrits running under daemons
   as a part of a move to perl 5.14.2
 - warnings under per l5.14.2 tidied up
 - unnecessary '##no critic' statements removed

release 73.6
 - test for npg_daemon_control fixed copes with presence of daemon plugins
     that are not defined in this package
 - npg_testing::db module:
      copes with old-style naming convention for yml db fixtures
      new subroutine for running tests against existing test database

release 73.5
 - live environment might be set explicitly or implicitly (default) RT#325903
   for the purpose of log file location
 - st::api::lims class variable for inline index end
 - deamon monitor script, daemon parent class and tracking related daemons moved to
   this project from instrument_handling
 - disable percritic policies for having VERSION defined and having Rcs Keywords
   since these are not available in git
 - extend perlcritic tests to all namespaces except Monitor and npg_tracking::Schema
 - notification of perlcritic test failute: include policy name
 - mysql client does not read configuration in default places to avoid reading wrong configuration
 - use password for the root user on a local test database
 - webapp: reanable a link to HTML based RTA reports for MiSeq

release 73.4
- build procedure for the web server
- test that resets modification time of test files is directed away from source distribution
- top-level .gitignore file to stop git reacting to back-up files and files generated in a normal build
- reference finder to return no reference if the sample is not suitable for alignment RT#313168
- calling abs_path for non-existing files breaks under ./Build test with latest CPAN modules
    reference finder refactored to ensure that abs_path is called on existing directories
- tests improvements

release 73.3
- change MiSeq samplesheet writing location to sf49

release 73.2
- get build version from git tag or version with Keith's gitver script
- inject module version into npg::view when copying it to blib

release 73.1
- rename NPG_REFERENCE_REPOSITORY type to NPG_TRACKING_REFERENCE_REPOSITORY
  to allow for the new and original npg_common modules to co-exist
- npg_tracking::data::reference::list - do not export repository root since
  this is only needed for testing

release 73.0
- ref&bait finders and runfolder modules - copies created under npg_tracking RT#306995
  dependencies withing this svn project refactored to use new modules
- Build.PL amended to install a version of CGI compatible with the post requests in tests
- sample_description method added to the lims interface
- ensure correct number of lanes in NPG (tracking GUI and the database) for HiSeq2500 RT#306985

release 72.8
- live and dev sections removed from the config file with db credentials
- new instrument status 'wash in progress' RT#305630
- tag from sample description for John Collins libraries RT#301185

release 72.7
- npg_common::roles::run, npg_common::roles::run::lane and npg_common::roles::run::lane::tag moved to npg_tracking::glossary; dependent code refactored
- dependency on npg_common::roles::log removed

release 72.6
- ensure email templates are specified on the script level
- staging monitor - cope with dev env variable being not set

release 72.5
- MiSeq samplesheet generation: list on-instrument genome reference files RT#301274
- infer the tag sequence from a sample description - sample level only RT#301185
- removed explicit dependency on /software/solexa
- removed explicit dependency on npg_common::config
- extended Build.PL to deploy 'data' element - contains email templates
- moved npg_common::config::roles::db_connect to npg_tracking::util::db_connect
  removed dependency of this role on npg_common::config
  enabled parsing config files without top-level domain section (like live, dev)
  config file is expected to be in .npg directory, which should be in the home
  directory or, if $HOME is not defined, in the current directory
- backed up npg_tracking::Schema by npg_tracking::util::db_connect

release 72.4
- Makefile.PL file replaced with Build.PL that can build all dependencies from scratch
- db fixtures updated (NULLs removed)
- possible to use local mysql test database without having socket specified in the
  mysql configuration file

release 72.3
- a fix to instrument uptime as exposed in xml feeds RT#301282: npg uptime numbers correct?

release 72.2
- illumina sequencing failures ticket figures out what to do if no id_run option is given RT#161876

release 72.1
- bug fix for child scripts of instrument poller not pickign up the correct perl executable

release 72.0
- npg::samplesheet - OnlyGenerateFASTQ is no longer valid in settings
- decription key is taken from the relevant portion of the config. file
  rather than from the live section
- changes to run under perl 5.14.2 and updated Moose, in particular, coping with variations in Moose error messages
- staging monitor bug fix for getting file modification time
- iscurrent flag for both instrument and run statuses
- four new statuses for instrument, 'planned repair', 'down for repair',
  'planned service', 'down for service', replacing 'down' and 'planned maintenance'
- instrument status rules explicitly defined and propagated
- instrument status auto-assignment refactored to simplify further additions
- wash required instrument status auto-assigned on the completion/cancellation
  of the run rather than on the start of the run in both DBIx binding and Clearpress web app
- DBIx binding definition of the idle instrument brought in line with the
  web app definition, ie on reaching run mirrored
- npg_testing namespace moved to this package
- InstrumentPoller cron job to set 'wash required' status if it's due according to a schedule
- cbots change status wash performed to up automatically
- DBIx instrument creation creates an initial was required status for sequencing instruments
- test to check that scripts compile

release 71.4
- Monitor::RunFolder::Staging - check_tiles falls back to checking bcls if there are no cif files
release 71.3
- Monitor::Roles::Cycle - use Intensities/BaseCalls/L001/C* when cifs are not copied and Intensities/L001/C* fails for getting the latest cycle

release 71.2
- st::api::lims
    change associated_lims to descendants
    change associated_child_lims to children
    add methods for returning lists of ids for studies, samples, projects and libraries
    add a method for returning a list of library types

release 71.1
- change contains_unconsented_human to more accurate contains_nonconsented_human
- add contains_nonconsented_xahuman method for determining if a study has non consented X and autosome human data

release 71.0
- Sample_ID not SampleID in generated samplesheet for MiSeqs
- Fix cBot fqdn suffix - change  from dynamic.sanger.ac.uk to internal.sanger.ac.uk
- add tileviz link off run page

release 70.4
- cope with V2 MiSeq reagent kits in samplesheet generation

release 70.3
- typo fix - space between batch id and .xml extension in javascript

release 70.2
- npg tracking test server config and README removed; see the configuration in sflogin svn project

release 70.1
- sample consent_withdrawn method added
- svn::executable ON property set for scripts in cgi-bin pending tracking server deployment on sfweb

release 70.0
- bait_name attribute added to st::api::lims object RT#254705
- change pass colour to match that in SeqQC i.e. from green to blue RT#262812
- tweak to samplesheet class so that id_run is not required if run object is provided

release 69.6
- MiSeq sample sheet generation tweak: bung PhiX in as a reference when we cannot figure one out so that the MiSeq will run

release 69.5
- time issues (implicit use of UTC time) in DBIx binding resolved, the method use explicitly local time  RT#259861

release 69.4
- heatmaps for npg-qc: tile layouts added for HiSeq and MiSeq
- sample sheet refinements and bugfixes - remove repeated 'GenomeFolder', end lines with CRLF rather than LF, const number of commas per line
- code for daemon or cron to generate sample sheets for pending MiSeq runs

release 69.3
- loaders can now duplicate a run. 'Run Duplicate' button appears only if this run's current status is run cancelled. RT#257773
- minor change to MiSeq samplesheet
- sensor test refactored
- sensor code fails if encounts an unknown sensor - fixed #RT259833

release 69.2
- additional configuration files, scripts and instruction for stand-slone web app in a sandbox
- tests fixed so that they run in an Ubuntu sandbox with ClearPress v422 (Jan 2012)
- loader levels dropped

release 69.1
- parsing of unusual usernames like mark@illumina.com in sanger_sso authentication module #RT258529

release 69.0
- Added support for storing temperatures from netbotz
- Independent implementation of Sanger LDAP access and Sanger SSO cookie decription
- npg_common::mailer moved to npg::util::mailer to be used by npg::sensor and npg::email
- npg_common::image package and related modules moved to npg::util::image namespace; removed Moose dependency in npg::util::image::colourmap

release 68.2
- Unused npg tracking db tables deleted: entity2reference_sequence, reference_sequence, project;
- Unused columns deleted from the run_lane npg tracking db: auto_analyse, id_project;
- Run table changes: column 'is_dev' is dropped, a new column 'team' is created to associate teams with a run. Where id_dev column had 1, the 'team' column has 'RAD', otherwise the value is 'joint';
- DBIx binding, ClearPress models, views and templates and npg::api updated accordingly
- DBIx binding: create and search cbot runs removed
- new run page redesigned; instrument type detection by javascript on this page improved
- team selection drop-down list on a new page list

release 68.1
- create run page javascript bug fix

release 68.0
- rebalanced the code in st&npg::api::base and ::util onjects; st::api::util object removed
- deprecated and unused code removed from st and npg apis, tests adjusted accordingly
- old test data for st package removed
- tests refactored to exclude deprecated calls and make use of test useragents transparent
- cgi script for e-mailing annotations removed

release 67.1
- Cope with tracking runs with more than 3 reads and drop support for older runfolder Netcopy_READ<digit> and only use Netcopy_Read<digit> files for run complete and mirror complete
- Run, New Run, and Duplicate Run pages - MiSeq specific changes RT#254706
- permissions for manual qc group to change run status and make annotation; were missing in some places
- run-lane and instrument annotation comments fixed; were invisible

release 67.0
- st::api::lims - unused warn_fallback and fallback attributes removed, no fallbacks left when retrieving values
                - unused library_asset_type method removed
                - internally, _library and _request attributes removed
- image link fix for MiSeq RT#252322
- run priority to reflect lims' lane priorities RT#248112
- permissions for manual qc group of users to create annotations and add lane tags
- run, runs, instrument pages GUI de-cluttering and re-styling
- runs page - page reload respects existing selection;
    javascript deleted, reloads channeled through a full page reload
- instrument page restyled, run list ffor an instrument displayed as iframe
- ical, atom and excel feeds removed
- rfid input box moved to under the login widget; rfid javascript moved to a stand-alone script

release 66.0
- "qc in progress" state to help avoid mulitple qc people overlapping their work
- stop displaying batch update option to users who do not have the authorisation.
- MiSeq - remotely check both MiSeqTemp & MiSeqAnalysis directories
- skip cif check if no cifs seen at all in check_tiles
- npg_tracking::Schema::Results::Run - change current_run_status to current_run_status_description so original name can return object in future

release 65.1
- cgi-bin/npg script that can work both stand-alone and on sanger intweb; needs apache conf file similar to the one in t/test_scripts
- highlight colour made lighter to make reading easier
- fix longstanding javascript code bug which would screw up the underlying links (apparent when right clicked) to runs of differnt statuses/ different instrument types

release 65.0
- st::api modules
    passing around st_cache attribute is removed
    some additional methods deprecated
- npg tracking web app:
    npg::util - st_cache and useragent accessors removed
    deprecated module npg::graph removed
    own decorator object is used instead of SangerWeb; no dependency on SangerPath
    app version number is taken from npg::view instead of npg::controller
    /nfs/WWWdev/SHARED_docs/lib/core dependency retained in npg::view object for SSO and LDAP access, but the code runs OK if any of these modules is unaccessible
- documentation and files for running stand-alone apache tracking service are in t/test_server
- dependency on SangerPath and SangerWeb removed from tests
- npg::email tests that were sending live updates fixed
- MiSeq - alter reporting of runfolder and add a checker (interrogate instrument) script

release 64.0
- new staging areas sf46,47,48,49 added to the disk usage view
- json feeds removed
- npg project model and views removed
- foreign key constraint for the project_id column of the run_lane table dropped
- trackign db DBIx binding regenerated
- project_id is not rendered in XML feeds
- npg basic&advanced search: cas project removed, test extended to include st_cache table
- run list html page instrument selector changes: GAI removed, MiSeq added; MiSeq added to instrument db test fixtures
- instrument utilisation loader script refactored and moved from ./scripts to ./bin
- old unused scripts from ./scripts removed
- icons directory added to htdocs to render icons when a stand-alone server is run

release 63.4
- npg::email and illumina qc failures script for creating RT tickets use st::api::lims now; RT#241274
- link to SeqQC on analysis in progress RT#245662
- deprecated module st::api::run removed

release 63.3
- a date-related bug in the 10-model-run_status.t test fixed
- add instrument <-> designation many-to-many relationship in DBIx class
- add tag <-> run many-to-many relationship in DBIx class
- deleted some old test data
- deleted some old, unused scrips

release 63.2
- st::api::lims->contains_unconsented_human defined for a pool and tag 0 - looking at plexes

release 63.1
- st::api::lims - keys for insert size hash are always defined #RT241906
- st::api::lims - default fallback value is false

release-63.0

- npg::api::project object removed
- npg::api::run module:
    ->batch accessor is deprecated
    ->lims accessor created insted
- npg::api:run_lane module:
    ->lims, ->is_pool, ->is_library, ->is_control methods added;
    previously deprecated ->stproject_id, ->stproject_name, ->sample, ->project methods removed;
    ->id_project field removed;
    ->asset, ->library, ->pool, ->control, ->entity, ->st_lane methods made deprecated;
    ->is_spiked_phix, ->asset_id, ->contains_unconsented_human, ->manual_qc methods are refactored to use npg::api:run_lane->lims method;
- st::api::request module:
    previously deprecated ->libraries method removed
    ->project method marked as deprecated
- st::api::sample module:
    previously deprecated ->project ->workflow_samples methods removed
- st::api::run module marked as deprecated
- st::api::lane module  as a whole is deprecated
    previously deprecated ->sample method removed
    all methods marked as deprecated
- st::api::batch:
    methods ->batches, ->batches_released_cluster, ->lane_count, ->lanes marked as deprecated
- st::api::asset module:
    previously deprecated methods ->workflow_sample, ->libraries, ->projects, ->project removed
    methods ->required_fragment_size, ->studies, ->reference_sequence, ->organism, ->study, ->study_name, ->sample marked as deprecated
- st::api::tag module is deprecated, all methods marked as deprecated
- refactoring of some tests to use the web cache where possible
- update schema scripts moved to scripts/upgrade_schema directory
- some unused scripts removed from the scripts directory: runs_for_casproject, api-demo-project, project_finder, api-demo-stproject, report_q_values_to_sequencescapeprojects

- updated image thumbnail URL to work (only) for HiSeq machines
- v prelim MiSeq samplesheet generation

release-62.0
- cache npg api run object in st api lims when id_run given
- return spiked phix tag_index for a lane as well as a plex
- st::api::lims - lane_id and lane_priority attributes added
- batch xml parsing in st::api::lane tightened up, RT #238508

release-61.3
- st::api::lims - read library name from batch xml, RT#234177
- st::api::lims - new fallback accessor; if fallback is disabled, no falling back to get study id or library name
- set user agent in the tracking ajax reflector to a string starting with 'NPG_INTRERACTIVE' to allow Sequencescape to recognise that this is for an interactive (user waiting) xml request
- fixed test data to reflect the fact that in the test we cannot work out the instrument (see release-61.3 notes), so the created run NHTL page should not contain data about any particular instrument

release-61.2
- to avoid use of uninitialized value message in log when checking instrument name
- when creating a run, if we can't work out the instrument, then option Please Select: displayed

release-61.1
- run lane status now being used to record where the lane is through analysis, and to update the run status if all lanes have reached a certain applicable status (i.e. all lanes at analysis complete - run status set to analysis complete, qc review pending)
- RT #231143: st::api::lims->required_insert_size - no fall back to old ways if insert size definition present in batch XML
- st::api::lims->required_insert_size bug fix for pools

release-61.0
- RT #229481 extension to st::api::lims to get multiple lib., sample, and study name with an option
  not to include spiked control
- removal of batch model and view, as now data obtained via sequencescape ajax request directly
- bugfix - ensure that loaders can't verify their own work

release-60.5
- display against a lane if it is from a cost code which determines if it is R&D
  - as lanes are added, tick (prod) displayed and then as it retrieves the project
    info by ajax changes to a cross

release-60.4
- a test for st::api::lims to ensure that all delegated methods are callable without an error

release-60.3
- bug fix: remove sac sponsor from study delegation in st::api::lims following its removal from the study object

release-60.2
- provisions for keeping specific project cost codes in the tracking db
- improvements and additions to st::api::lims module
- redundant and unused code deleted from st::api::study and at::api::project, their tests simplified

release-60.1
- bugfix: instruments should be ordered by name

release-60.0
- npg::api::request module instead of npg_common::request
- npg::api::request module saves to cache if certain env variable is set
- st and npg api use the new npg::api::request module
- refactoring of st and npg api and st::api::base so that the request object is created in st::api::util only
- module for single LIMS data access point, st::api::lims, and tests
- a new line 45 test
- staging area daemon updates DB with folder_name and folder_path_glob
- 3 stage verify for runs (flowcell, reagents R1, reagents R2)
- record loader of R2 reagents
- Display verify columns on the run list pages, sortable
- MiSeq added to instruments (image, display, database)
- Display of instruments refactored to ensure that the instruments are
  - displayed in required order
  - only displayed if they are required and have current instruments
- Annotating multiple runs has the Y/N Run OK options

release-59.1
- do not store sample object in pool description if sample_id is a bad sample id(4)

release-59.0
- button for verifying run
- bulk run annotation entry - replaces bulk status update
- increment priority of run, limited ability to do this to manual_qc and approvers
- tags on run lanes now displayed on main run page, and manual_qc can add tags through this page
- st::api::study now has methods to retrieve the Accession number, title and publishable name

release-58.1
- add method publishable_name in sample

release-58.0
- login with rfid on instruments for submitting forms
- annotate lanes from run page, enabling multiple lanes to get same annotation, and also annotate instrument at the same time as the lane
- enable button update of run from qc review pending to archival pending if the user is in manual qc and if all non-control lanes return back from Sequencescape as pass/fail

release-57.0
- Fairly thorough purge of st::api from Clearpress based npg::model and npg::view and templates
- Remove table lock on run creation.
- No longer link to MPSA from run page....

release-56.1
- failures tickets e-mails to get the study of the st lane directly
- when parsing batch xml for st-lane, getting a study from the request is optional to allow the warehouse loader to run in reasonable time

release-56.0
- A fix to instrument view to make visible the fifth digit of the batch number
- st::api - move functionality out of batch to lane for creating the lane information, and add in obtaining the studies,
  preferably via the request, else via the sample->study

release-55.0
- copying_problem flag set/unset at run complete status as well as run in progress
- delay also includes any missing cycles from L1
- stuck_runs page, visible to annotators, loaders and admin
- alter instrument graphic to indicate no recent contact or copying_problem with current runs
- manaipulate hyb buffer tag sequence to be of same length as pool sample tag sequences

release-54.3
- HiSeqs and GAs to go(auto) down correctly if their current status is planned maintenance
- initial model-based code for listing potentially stuck runs

release-54.2
- following a change in live data, 45 test fixed
- tests for phix reference that use a deprecated method removed from 45 test
- some tests that compare DateTime object to a string data fixed by stringifying the object explicitly
- added blocking_runs method and  mapping of hiseq slots to blocking runs for the instrument model
- possible to create a new hiseq run if there is a run-complete run in a slot

release-54.1
- spiked tag part of batch xml parsing extended to allow for reference finder to deal with it
- spike tags to be added to the lanes tags if the lane is a pool

release-54.0
- is_spiked_phix method added for st lane and run lane
- fix the gantt charts that are broken
- all events are emailed out separately from the web application
- Sequencescape is notified of a run status change outside of the web application

release-53.0
- correct criteria for determing run/mirroring complete
- add new staging areas to usage_list template
- minor bugfixes and updates

release-52.0
- a new table and DBIC changes - manual_qc_status
- st::api - add alignments_in_bam method to study
- Colors for two hiseq slots should be different if run status different in instrument images.
- Refactor and revamp of instrument and staging area monitoring
- Run reports - add filtering by machine type

release-51.0
- Use additional Illumina files to test for latest_cycle, run_complete,
  mirroring_complete. These should work for both HiSeqs and GAIIs.
- More efficient testing for completion of file transfers in staging_area_monitor
- Log event_notifications to disk
- Update instrument_status where necessary when a run_status changes

release-50.0
 - Enable full functionality of monitors so that they can be moved to live.
 - Fixed memory leak in npg::api::run

release-49.0
 - add a new checker script for HiSeqs
 - fix and add tests for HiSeq runfolder names

release-48.0
 - preparation for extra nfs/samba storage area for hi-seqs
 - new monitor scripts added to watch instruments (cbots, ga-iis), staging
   areas and notification events
 - ancestor_samples added in st api asset
 - gantt charts fixed and display for both HK(GAIIx) and HiSeq
 - change useragent used in npg::api and st::api so the version is marked correctly and shows its inheritance from libwww-perl

release-47.0
 - completely drop goldcrest functionality and records
 - use DateTime inflation for DBIC
 - a new npg::model::instrument::current_runs methos to return multiple current runs for the instrument
 - npg::model::instrument::current_run method refactored to use npg::model::instrument::current_runs
 - npg::model::instrument::fc_slots2current_runs method to return mapping of flowcell slots to current runs for HiSeq instruments
 - two new tags to identify a particular flowcell slot
 - create a new run based on given flowcell id
 - npg::model::run::create method handles assigning flowcell slot tags
 - template extended with fc radio buttons and current runs listing
 - npg::view::run::create to pass the flowcell radio button selection to a model
 - textual and graphical views of instruments updated to include HiSeq instruments
 - hight restriction in dd sections of styled lists is removed

release-46.0
 - purge old deprecated st::api classes and methods
 - dbix binding for the npg tracking db can read configuration files
 - catch the retrieving batch error and return when sequencescape server down in Javascript for mod_perl
 - reduce retrying times of sequencescape batch call in model run to avoid database connection time out when changing run status, max_retries and retry_delay to 1
 - change dev environment to look at Sequencescape next release (psd-dev:6800)
 - instrument utilisation ready for multiple instrument types
 - events now have notification sent in preparation for removing Sequencescape interaction and emailing from core NPG Clearpress model

release-45.1
 - use port 6800 for Sequencescape in dev environment (their "next-release")

release-45.0
 - remove st calls when generating run and run_lane XML
 - alter npg::api to retrieve manual_qc now not in run XML

release-44.0-webcache
 - npg and st libraries changed to use npg_common::request to make web requests

release-44.0
 - remove st calls when generating run HTML
 - add reflector CGI to allow for AJAX calls
 - add AJAX calls on HTML run pages to show Sequencescape info
 - only fill current cycle number in run annotation form when run is in progress
 - cache st batch object with information about batch_id missing or not
 - pool_description method is added to the st lane object; returns a mapping of tag indices to tag object references and lib tube and sample ids

release-43.2
 - bug fix, cache loader_info in hash with information about full date or not

release-43.1
 - bug fix, only new run with scs2.8 having 5-digits run name

release-43.0
 - add run_folder method in npg api run
 - don't display current_cycle as 0 when no current_cycle given for run annotation
 - run annotation email with current_cycle and currently_run_ok included
 - truncate event description if too long
 - croak when trying to read a non-existing run
 - croak in npg api when an error occurred message returned from npg
 - add project_cost_code method in st::api::project
 - change run name to five digits if the instrument with SCS 2.8 or above
 - get expected tag sequence from batch for each lane
 - add run read table
 - added an api for cbot XML - npg::cbot::api - and added some methods to the DBIC class for feeding cbot data to database
 - create a daemon to periodically poll all cBots for instrument status

release-42.1
 - Post Sequencescape release Wednesday, 12 May 2010 fixes.
  + RT170288 - las reports $oAsset->projects not working
  + assets had project_id in past but do not have study_id following release (expected in the project -> study change : go via sample)
  + work around for SAC sponsor being in study instead of project where it will go....

release-42.0
 - add buttons for each isnturment status to annotate them
 - cycle count and current_ok fields added to run annotation
 - new st api study calss, the same class as old project class
 - deprecated some methods in st api project and transfer the calls to the corresponding study if the sequencescape api version newer than 0.5
 - fix summer-time error in annotation creation time-stamps
 - move planned maintenance instrument to down on run cancelled as well as run complete - RT167177
 - Add latest_contact column (DATETIME type) to instrument table
 - rename link QSea to NPG-SeqQC
 - add many to many usergroups and make username ubique to allow use as DBIC authentication infrastructure store

release-41.0
 - interpret taxon id as an integer if it's a float with one decimal number that is 0
 - add flowcell_id field to run model
 - wash required after every run
 - DBIC Schema for tracking
 - new asset parent and child parsing
 - add flowcell_id to run fields
 - don't return attribute if value is null/undef

release-40.0
 - add a having_control_lane method in api run
 - methods to get organism-related info about a sample
 - deprecated st::api::sample::reference_sequence
 - added required_fragment_size method to an asset object
 - add utility script to create tickets for failed lanes
 - add email infrastructure to email run complete etc to contact found via Sequencescape
 - stop sending error emails with no recipients
 - separate out different instrument types on graphical list page (cBot preparation)

release-39.0
 - add backward compatibility methods to make life easier for api users followinf Sequencescape V4 release

release-38.0
 - st:api altered for Sequencescape V4
 - instrument icons representing their status indicate paired or single reads, whether the machine is R&D

release-37.0
 - instruments' utilisation info for 90 days
 - re-styled menu for instruments
 - 'Instrument Formats' pag unlinked from the top level menu and linked to the Instruments menu
 - npg::graph object made deprecated
 - npg::model::instrument_utilisation->last_30_days marked as deprecated
 - st::api::pool - cope with new Sequencescape release - Pool id in Sequencescape Pipelines refers to an item in Sequencescape Projects as of 13/08/2009
 - minor st:api refinements

release-36.0
 - link to RTA status page (if appropriate)
 - user list view redirects to read page if a user is logged in, else just displays a request to log in through SSO. Doesn't error any more.
 - npg::api::run_lane has contains_unconsented_human method, which will return a boolean based on what the sequencescape project can tell about the sample
 - links to NPG run pages and NPG QC run pages in event emails
 - planned maintenance - mark machine as down immediately if not in a run.
 - run tags and is_paired_read info now in run list json
 - is_paired_read key/value now in run read json

release-35.0
 - gantt style charts now have hover over the run/down blocks, and a combined chart
 - email the loader of a run when a status or annotation is added
 - display a default message if a Clearpress default method used, but not supported (currently only set for batch list)
 - hyperlink directly to NPG-QC
 - Reveal link to click through to analysis prelim if that is available

release-34.0
 - add_tags and remove_tags function are added in npg api run
 - give user pipeline permission to add tags
 - gantt style graphs displayed showing
   - timeline of up/down on instrument for last 90 days with instrument annotations marked on
   - timeline of runs running on the instrument (run pending to run complete) fo last 90 days
 - Goldcrest graph links removed from GUI
 - Project page link removed from GUI
 - Links on run page which go out to the staging area, dependent on availabilty of runfolder on staging, and availability of Illumina Analysis
 - Instrument graphics Show Run instead of R to clarify that this is going to be a paired run
 - Repository info moved out of Lanes table on Run view to single point on page, since all for a run go into the same directory now
 - When annotating a run, radiobutton to annotate instrument as well with the same annotation option given
 - Error emails now show all environment variables

release-33.0
 - alter terminology to reflect the new distinction between a paired run and a paired read run
 - in npg api run, is_paired function will complain that it is deprecated and return the same as the new is_paired_run function
 - A new is_paired_read function is added

release-32.5
 - search/<term>.json view
 - npg::api::run_lane has an entity method which returns either the library or control object
 - new script: report_q_values_to_sequencescapeprojects

release-32.0
 - bug-fixes as errors came to light
   - LDAP - noticeably pipeline - if LDAP down, stopped moving statuses on due to email trying to be friendly and getting realname. Now, for pipeline, doesn't go to LDAP, and for other users, traps LDAP error, returning username so should continue
   - bug fix in st::api::base when looping through the fields looking for an element, was returning if not found, instead of nexting
   - search/<term> not available, so inserted
   - read_attachment url not working. corrected
 - st::api::request and post statuses back to request_id, not item_id
 - st::api::project - fixed ref sequence bug due to change in xml
 - code preparation for when samples <=> workflow_samples becomes available in SeqScape

release-31.0
 - error handling now has more output in email to the errors group
 - instrument_utilisation now counts instruments as utilised if no chance they could be loaded again
 - instrument_list_json and instrument_read_json templates (uses default ClearPress list_json and read_json methods)
 - tidier pod
 - fixed several bugs in web application
 - st::api methods to do with checking project names are deprecated

release-30.0
 - improved error handling, emailing new-seq-pipe on non-authenticated errors, and how the error page looks to the user
 - internal changes to reflect sequencescape's changes in terminology, and xml, for sample/workflow_sample/library
 - st::api::sample re-written to model sample and NOT workflow_sample
 - report qc_state in run_read page. Remove older qc references from other pages.
 - report disk-space on new mirroring hosts.

release-29.0
 - Synchronise with the release of SequenceScape v3.0, make sure that correct method names are used.
 - Remove good_bad test data, tests and templates.

release-28.5
 - patch release to tidy up some views and fix bugs
 - old instrument_utilisation and uptime graphs removed from the instrument page
 - usage changed to utilisation in utilisation as %age of uptime graph
 - utilisation/uptime production accounts for hot spares in the denominator
 - bug fix where colour for imstrument remains green if idle between R1 and R2
 - run folder name displayed on run read for cutting and pasting purposes - coloured red if run pending
 - repository points to first end of paired run
 - batch to runs script

release-28.0
 - Added instrument designation labels
 - Script to generate csv report of runs loaded per month
 - Instrument utilisation table and loader script plus model and views
 - switched repository reporting to fuse/runs/
 - Internal changes, including renaming references to "sample" to "library"
 - More multiplex pool handling, event xml emission checks
 - analysis_report script to produce monthly stats on runs loaded within a month. Also links with npg_qc api modules for Analysis info
 - Extended st::api support for more-correct library vs. sample distinction
 - Removed read_associated & ical support from projects (due to removed SequenceScape functionality)
 - Dropped analysis, analysis_lane, analysis_lane_qcal tables, supporting modules, tests & test data
 - Removed last remnants of run_lane_status
 - Added basic run xml service for SequenceScape run-pair queries: /run.xml?id_run=x,y,z
 - Removed 'lane/library/individual aggregate' todo - functionality now in dfind

release-27.0
 - Remove QC input on reads from interface.
 - Added run_status method to API to get all the run status of a run.
 - Streamed project list_xml response & api support.
 - Script to compare project.xml to a repository ls to see which expected projects are not in it.
 - run->create_tags renamed to run->update_tags to be more restful.
 - Test reworked to use t::request - much cleaner!
 - is_good now returns either the stored value of good/bad or the value for unknown. No attempt is made to try to determine whether a lane is good or bad. Eventually, this should become deprecated.
 - Removal of any data and methods which come from analysis_lane (or derived from using analysis_lane).

release-26.1
 - Link to NPG-QC summary on page for run
 - Scrapped automatic QC hints
 - Various minor bugfixes
 - Trapped pressing return/enter in batch-entry box during run creation
 - Streamed response for run_status_dict_read.xml
 - Support fetching instruments for runs via API

release-26.0
 - Run reports link to IVC plots
 - Run reports link to Error plots
 - Default tile-layout for run-creation based on instrument-format
 - Reworked some tests to use fixtures
 - Basic-search now includes instrument annotations and statuses
 - Revised instrument colour palette; added new 'request approval' colour.
 - Run creation now identifies sequencer based on X-Sequencer HTTP request header, supported by mod_headers & mod_rewrite in the seq-farm forward proxy
 - Run creation is now less tightly coupled to SequenceScape batch services

release-25.0
 - API updates for improved denormalised data services
 - Instrument-type/platform attached to runs
 - Database-connection / garbage collection problems fixed
 - Paginated run lists for instrument- and run- views
 - Only members of R+D group can set 'is_dev' on runs
 - Updated GA2 images
 - Improved service handling in st::api::base

release-22.0
 - New instrument annotations
 - Improved, tab-based instrument view
 - Good/bad manual QC events are now posted back to ST system
 - Ability to automatically track tile layout (similar to automatic actual/expected cycles)

release-21.0
 - Added Cache::Memcached support for Sample Tracking services which
   should further decouple NPG's dependency on the ST services.
 - st::api::project can now fetch samples. $oProject->samples();

release-20.0
 - Many updates to improve code quality (perlcritic)
 - Updates to account for (positive) changes in the ClearPress data
   model, e.g. differences in handling NULL/undef, zero and
   empty-string.
 - A large amount of effort in improving the test suite
   o More tests now run under fixtures meaning they're more reliable
   o The test suite is more complete
   o Errors are trapped and handled in many more places
 - Additional handling in the API, in-particular the ability to fetch
   all runs

release-19.0
 - NPG fully deployed under mod_perl resulting in significant speed improvements
 - "New Run" page now includes the current instrument status
 - All instruments individual uptime now shown on textual instruments page
 - Additional, faster & lighter run service for use in ST
 - Column revisions in "Sample Summary" report for Toby: SE/R1/R2 plus run pair id
 - Smarter run-pairing now ignores cancelled & discarded runs

release-18.0
 - Support for fetch runs by sequencing-project-id: .../run?project_id=<...> (no interface yet)
 - Fix for a memory / database-connection leak preventing mod_perl deployment
 - New st::api::run representation of runs from Sample Tracking, containing yield & billing data
 - Instrument uptime reports - all instrument uptime percentage per day for last month, individual instrument uptime for last 3 and 1 month
 - Addition of auto_qc field on run lanes in JSON services, reflecting billability of a lane drawn from ST system
 - npg::api now using mod-perl service for speed improvement
 - test coverage 91.22% (1915 tests)

release-17.0
 - approval for instruments to go up from being fixed - new 'approvers' group and hardwiring of rules to enforce fixed order of instrument statuses
 - when a run is in progress, but maintenance is planned for instrument, 'planned maintenance' status can be set, and instrument will move to down automatically on status change of run to run complete
 - qc pass/fail now changed to qc complete. automatic move of run to this status when all lanes have been qc reviewed as good/bad
 - enforcing of annotation (reminder box) for run when all lanes qc reviewed, and for a lane when it is checked bad
 - run read now has tags, status history, annotations and goldcrest graphs on tabs at the bottom. Annotations now shows all annotations for the run, run pair, run lanes and run pair lanes
 - tests now check JSON feeds are parseable
 - st::api and npg::api have built in retries
 - st::api::batch now does more lane data caching
 - st::api::events now sends id_run and lane position back to st
 - improved test coverage - 91.63% (1847 tests)

release-16.0
 - instrument batch mod update page - shows all current mods for each instrument and allows batch update of a mod
 - admin page - admin options dependent on group memberships (current engineers can add new mods and new inst_statuses, admin can do much more)
 - templates modified to not show certain functions on pages dependent on group memberships, instead of relying on return to say unauthorised
 - added standlone server for developers/external
 - improved test coverage - 91.26% (1567 tests)

release-15.0
 - Sticky widgets to ensure that select info doesn't get lost when changing batch on run add
 - instrument utilisation chart by hour as well as by day
 - batch run status change
 - API access to ST annotations (st::api::annotation)
 - Superior test framework including dummy data and some refactored tests
 - Support for rolling over full repository project folders to new projects
 - Batch update of instrument mods

release-14.0
 - Optimisation of speed in run_add loading
 - Allow run_status updates for both ends of a pair simultaneously
 - Cumulative quality plots for run_lanes
 - Instrument modification tracking
 - Create run now allows loaders to create a run
 - Refactored some javascript out into npg.js
 - Bug-fix : instrument utilisation broke after move to mysql5
 - Reworked JSON basic-search responses to include JSON representations of matched entities
 - Added sac_sponsor to run_lane_row_xml.tt2 (which feeds into project_read_xml.tt2)
 - $oUser->is_member_of now returns true if member is admin, so admin people are automatically members of all groups
 - Extended project_read_xml to include additional <stproject ... /> data
 - annotators group added
 - authorisation added as follows (admin permissions have not changed)
   create runs - loaders
   add tags to runs or lanes, run annotations - annotators
   run status changes - loaders and engineers
   change instrument status and instrument modifications - engineers
 - Good/bad determination documented on NPD wiki

release-13.0
 - instrument-wash-required now triggered after two runs as well as one month after last wash
 - instrument-down indicator is now much more obvious
 - bug-fix : tags can now be removed, and refactored saving of tag frequency
 - bug-fix : replying to annotation email has a do not write below line, which will ignore that part of the email
   if reply by email, but doesn't delete the rest of the emails body
 - bug-fix : cancel button on run_status and run_annotation entry forms now works (by hiding the form, so it can be
   revealed without the need to reload the ajax request). Refactored the javascript to remove duplication
 - searches now search tags, both basic by default, and advanced by options
 - json and xml responses for advanced search (input cgi params in the url)
 - json response for list_samplesummary, and refactored the xml response to use attributes rather than tags
 - new run statuses ('run stopped early', 'qc review pending', 'qc pass', 'qc fail', 'data discarded') to synchronise with archive-by-default in sanger-pipeline v5

release-12.0
 - run_lane data included in project_read.xml REST/API interface
 - current_run included in instrument_read.xml REST/API interface
 - loader names and loading date info added to xml and json templates for run_list and run_read
 - tagging of runs and run_lanes can now be done (these export for runs in the xml and json feeds)
 - run_lanes can be checkboxed as good or bad on run page
 - sample tracking for sample updated on change of run status now includes id_run
 - titles of pages show run name for run read page

release-11.0
 - progress bar on instrument-state images (only for runs in progress)
 - advanced search now can allow searching for CAS projectnames, and where CAS projectnames and/or sample names are like entered text
 - automatic instrument-wash-required indication
 - engineer usergroup for instrument-status-change event mails
 - user from sample tracking for sample on run mailed when run status updated to run complete and run archived
 - sample tracking for sample updated on change of run status
 - advanced search now keeps selections for query, using ajax if enabled

release-10.0
 - Basic search hit highlighting
 - Tracking of 'actual_cycle_count' per run
 - Caching of sample tracking sample & project names for searchability
 - JSON-format views for run_read & run_list
 - Bugfix for event mails to external email addresses
 - Floated hyperlink buttons for actions and run statuses, so that they don't break on whitespace
 - Advanced search created. Creates single SQL query. Not perfect, but can retrieve various information based selected criteria. Always retrieves run id as first column, which is then hyperlink to the run page
   only retrieves information from npglive database for now, cannot bridge to CAS or ST

release-9.0
 - Sample-Tracking Projects listed & linked at the top of individual (CAS) 'Projects' view.
 - loader_info method added to run - fetches loader and date of loading (set to 'run pending')
 - runs_loaded method added to user - fetches instrument, id_run and date loaded for all runs loaded by user. table display on user page
 - script and module (with tests) to parse a piped in email from srpipe@sanger.ac.uk to create an annotation for a run
 - goldcrest data on runs improved to - no show of PF clusters (as incorrect to final tally), graphs show all cycles so far, and are hidden on tabs
   so loading of run page is quicker

release-8.0
 - Batch instrument_status change
 - Graphs on run/x and instrument/ILx to show last 10 (or all if not had 10 run) cycles phasing, prephasing, Signal mean and Noise,
   plus table of PF Cluster count. These pages auto refresh after 60 minutes to enable
   to be left on screen next to instrument to show progress (created npg::model::goldcrest)
 - No good/bad run indicator shown if it cannot be calculated for a lane
 - Good/bad run indicator now also visible on run_lane_list
 - Graphical barchart of instrument activity over last 30 days
 - instrument utilisation all calculated within SQL query
 - Basic text-search facility

release-7.0
 - Addition of sample data to email alerts
 - Addition of loaders usergroup, and additional column to show level to user2usergroup table
 - Show loaders level within user display against group loaders, within group loaders display,
   and against user in the run display (display is a medal (or lack of one if not assigned a level
   yet) Bronze, Silver or Gold)
 - Ability to promote the loaders level added to user display, if requester is not public, the
   user displayed, or they are already at Gold level
 - Session handling to remember status-type (id_run_status_dict) choices on run lists & instrument views
 - 'Run Summary' and 'Sample Summary' now only list runs & samples with recent 'run mirrored' statuses
 - R&D flag on runs
 - 'Latest Analysis' link on run view now links to the first read of paired-end analyses
 - Updated to support streamed page responses
 - Fetch run-lanes indirectly (as well as directly) associated with a CAS project
 - Sample Tracking API now supports fetching batches for projects. See scripts/api-demo-stproject
 - Instrument list now displays utilisation for last 30 days
 - 86%+ test coverage

release-6.1
 - modified sample summary to show different data (PF and Total Mbases) and improved look with scrolly css
 - refactored obtaining number of days parameter into selected_days method in npg::view::run
 - Good PF cluster density reduced to >17000

release-6.0
 - new sample summary, with assessment of good/bad samples
 - assessment of good/bad sample method in run_lane model
 - download sample summary to excel spreadsheet
 - improved some ability for testing html by checking expected tags in order, but is whitespace agnostic and allows for additional tags to come in, and ignores comments
 - improved reporting of error when croaking
 - added yearmonthday method to npg::util in order to return a string in format yyyymmdd for todays date, mainly for worksheet naming

release-5.0
 - integration with Sample Tracking web services

release-4.0
 - run ids allocated sequentially rather than with auto_increment to avoid sequence holes made by failed transactions
 - instrument_statuses added for service messages
 - improved run summary (reduced redundancy)
 - iCal support for exposing run_statuses
 - various template changes for interface improvement (incl. some javascript validation)

release-3.0
 - Atom feed for run_statuses
 - General event handling including email subscriptions for usergroups via event_type_subscriber table
 - Repository path added to run_lane and project lists
 - Excel-format dumping of sample information over a date-range
 - Library-analysis support - analysis targets & types are associated upon library creation
 - Generation of XML sample-sheets based on the SampleSheet.xsd schema from the run folder but extended to support the new analysis types & targets
 - Web API & example scripts
 * Recipe.xml loading

release-2.0
 - Some schema changes to provide more systematic names (e.g. lane => run_lane).
 - RunLog processing (scripts/process_runlog) via run_event, run_event_type and run_event_attr.
 - Basic graphing for time-courses of run events.
 - Project & repository handling for libraries and run_lanes
 - auto_analyse moved from run into run_lane
 - tags allowed in libraries

release-1.0
 - First release! Initial schema: manufacturer, instrument(format), library, sample, flowcell(format), run, lane.