Refinements to timed release

• Reworked publication date entry: radio button for today or future; mm / dd / yyyy textbox entry or datepicker; enforces maximum delay; extra validation and error messages
• Publication date and license in separate sections
• License acceptance is a checkbox rather than an alert/message box popup.
• Reviewer download link redirects to landing page if dataset is now public
• Generated reviewer link no longer has expiration and is automatically available for the dataset author when logged into the UI
• Revised large file prompt for email address that is a modal-style dialog w/ AJAX submission
• Logging on stage at medium "info" level
• Showing both original submission date and this version submission date in landing page.
• Fixed download count incrementing when using reviewer link
• Bug fix for missing "processing" step in displayed submission states
• Minor text changes such as tagline "An easy-to-use data publication service"
• Facets renamed in 'Limit your Search' and 'sort by relevance' section.
• Added further explanation to Usage Notes section.

Sprint release: 03/28/2017

See related tag v0.1.35 for embargo features for the same sprint (listed below)

• (Embargo) Embargoed material streams through the Dash UI when requested by the owner or someone with a secret link

• (Embargo) Determination if a download will be returned by Merritt synchronously or asynchronously

• (Embargo) When a dataset has been recently submitted and is not available from Merritt yet, a friendly warning is displayed instead of an error page

• (Embargo) Fixed "wrong arguments" error in the "Explore Data" link

• (Embargo) The submitted datasets in the "My Datasets" page show a new "embargoed" status in addition to the "published" status; publication date reflects embargo as needed

• (Embargo) Fixes to the embargoed, asynchronous (large file) download workflow

• (Embargo) User Testing Scenarios document for Embargo workflow added in confluence.

• (Embargo) Fixed issues with "/" in secret_id generation for Secret Links.

• The dataset title shows as the landing page title

• The Dash service manager gets notified of newly submitted datasets

• Merritt & Dash version numbers are now tracked separately

• Quotation marks and similar are no longer escaped in search results

logging changes, fixed typos, Merritt version tracking, keywords display fixes -- sprint ending March 10, 2017

• Fixed minor typographical errors
• Added dash logo to public directory so it can be used by InCommon to display in list of service providers
• Hiding routing errors (404s) in Rails log files to keep log file size much smaller
• Added a Merritt version number that is separate from the Dash version number because they may not align with each other and are used in different circumstances
• Embargo date being submitted to Merritt in mrt-embargo.txt
• Keywords text display no longer cuts off moderately long keywords
• Filenames display with ellipses in PDF file to fit column width
• Orcid formatting changes and icon changes
• Fix for map showing twice when breaking across pages (disabled Any3d setting)
• (embargo, in progress) added form for capturing embargo date and saving to database
• (embargo, in progress) added form for creating a secret, sharing link with copy/paste for sending to reviewers
• (embargo, in progress) sharing URL that allows someone with the secret identifier to download a dataset
• (embargo, in progress) adding email form for capturing asynchronous downloads
• (embargo, in progress) using API calls for determining download type (async or synchronous) and doing asynchronous downloads from Merritt
• (embargo) streaming downloads where credentials are required through the Dash UI

Sprint release: 02/28/2017 PDF tweaks, bug fixes, embargo features (embargo branch)

• (embargo branch) Migrated embargo table out of DataCite engine and into Stash Engine
• (embargo branch) Able to choose to publish today or on a specific date
• (embargo branch) Download, data paper and files sections hidden if under embargo unless logged in and owner of dataset
• Tweaks to PDF rendering
  • Smaller fonts in header and more room between it and body
  • DOI links are blue, underlined and clickable in PDF even where not linked in landing page
  • Removed "Type" section
  • Listing only up to 100 files with ... to indicate list is not complete
  • Text changes for download and data paper buttons
• Added to and revised install and config docs
• Fixed file packaging bug that attempted zipping file not updated in that version.
• Fixed version cloning and status problems for a new version of a dataset and worked around problems with Amoeba cloning gem
• doi URLs use doi.org rather than dx.doi.org as the domain
• Worked around Sauce Labs problems with testing of the upload page
• Adding ORCID icon to review and landing pages
• (embargo) Merritt embargo file added to the package in embargo branch of gem

Data paper, importing PLoS subjects, automated testing, refactor of merritt-specific code

• Code from the stash_engine and stash_datacite was refactored into stash-merritt gem which isolates Merritt-repository submission specifics from our main code.
• Most tests with Sauce Labs browser testing are working from travis.ci to run automated browser tests when code is checked in.
• First pass of the "data paper" functionality released which generates a PDF similar to the landing page and linked from the landing page.
• Added manual script (at stash_datacite/lib/script/plos_keywords.rb) to import PLoS Subject Area Thesaurus (https://github.com/PLOS/plos-thesaurus). It is flattened and imported to database of DataCite keywords for autocompletion.
• DOI text shows as a full URL rather than doi:xxxxx/xxxxx format.
• Fixed upload size in "help" vs "getting started" page in order to show the same size (total size).
• Fixed bug in Merritt packaging that could add wrong files in V2 of dataset and onward (updates).
• Fixed bug with old versions of datasets that was brought in when state history stopped being saved and new datasets would have a completed state on initial creation of v2 and later.
• Fixed bug occuring when trying to edit old datasets where the version wasn't created until submission meaning a missing version caused errors.

Release Notes, January 31, 2017

• Identifiers get autolinked linked when they show in related identifiers in review or on landing page.
• Uploading files works on three files in parallel
• Separate module for Merritt submissions (not yet integrated)
• Help and Home links removed from top menu. The logo goes home and help is in the bottom navigation.
• Progress bar is higher contrast.
• Author names are entered correctly when using browser autofill rather than generating multiple records.
• Map only zooms with scroll wheel after a previous click inside the map area.
• Added code to do a manual tab separated value export of related identifiers for validation.
• Increased timeouts for large file uploads.
• PDF Creation for a data paper is partially working (but not exposed to the user yet).

Release notes, January 17, 2017: Testing, documentation, bug fixes

• Added documentation for an install of dash with all required local engines and with MySQL and Solr setup (under documentation directory in this Dashv2 repository)
• Created new example configuration repository at https://github.com/CDLUC3/dash2-config-example.git which contains example commented configuration files and a sample record for the database/Solr
• Created Capybara integration tests to test application at the browser level
• Moving packaging and submission into separate module (in progress)
• Added dynamically generated site map to point crawlers at the public datasets on each domain
• Changed formatting for markdown on help and about pages so the bullets and tables look more like the rest of the application; differentiated heading levels in help page so heading levels are more distinct
• Email message changed so that it displays DOI correctly and indicates that the link may take time to become active
• Reduced logging level so log files in stage/production do not show informational messages, but just warning or higher
• Using --deployment switch in Capistrano scripts which installs gems in the bundle rather than in system/user Ruby install, reduces number of gems on servers
• Fixed database connections so they don't leak when used in background thread
• Changed page titles to contain "Publish and Preserve" rather than Dash2, and footer text wording changed
• Fixed bug where multiple versions of a dataset all used the same geolocation records, so changing a later version also affected the earlier versions

Minor UI and Workflow fixes to application Jan 3, 2017

v0.1.22 (Jan 3, 2016)

• Integration testing working for many pages in the application with Capybara browser testing.
• Zip packaging handled by background task, so not a long wait after a submission.
• Branches and tags added to config repo and deployment changed to use config version consistent with our code.
• 3 different maximum sizes ("total" size, submission size, file size) which are configured per tenant.
• Unknown/404 pages showing a custom not found page.
• Less verbose logging in stage/production.
• Mobile menu working in discovery part of web application.
• Titles of pages include "publish and preserve' and footer modified.
• Things that look like URLs in text of landing page auto-converted into links and shorter URLs.
• Graph icon shows as graph instead of random box.
• Maximum size for submission only counts added/replaced files in a version.
• Added Google Analytics.
• ResourceTypes with free text that come from outside imports are retained even though no way to set in the UI.
• Misc code cleanup and simplifications.

File Submission, testing, ajax loading support

Release notes

v0.1.15 to v0.1.21 (Nov 18 to Dec 9, 2016)

• UC Merced logo -- Added extra space added on top
• Maximum upload size is displayed based on configuration in all places
• Ajax loading errors fixed, caused by jquery-ui gem released with missing dependencies, updated to later version of gem
• Added manual script to import file list from an Atom feed when we must do manual deposit
• Renaming resource_type field to resource_type_general which is what this controlled list actually represents
• Adding resource_type_general (free text) and updating all code so it fills in both values and uses the new fields
• Up to 2 GB submissions
  • (In private puppet repo) Updating Apache config to allow longer http requests before the balancer times out
  • Updating Capistrano deploy to allow longer-running background tasks to complete before they are killed
  • Updated configuration to allow larger submissions
• Installed rspec testing framework and set up directories
• Minor cleanup of stash-sword gem and file upload code
• Use short campus names when possible -- both tenant display and autocompletion when available
• Fixing "Participating Partners" drop-down to stay on top of buttons/text on the page
• Simplify code that sets the version number for an identifier
• Fix DOI formatting/links in email