Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Release 2020 03 10 #1010

Merged
merged 27 commits into from
Mar 13, 2020
Merged

Release 2020 03 10 #1010

merged 27 commits into from
Mar 13, 2020

Conversation

fisx
Copy link
Contributor

@fisx fisx commented Mar 10, 2020

New features

Bug fixes

Significant internal changes

Internal Changes

jschaul and others added 27 commits February 18, 2020 19:44
* update nginx to latest stable 1.16.1
* alpine 3.11
* Cleanup formatting and syntax in package.yaml.

* Remove illegal ghc options.

Before this change:

```
[...]
sodium-crypto-sign      > Warning: 'ghc-options: -prof' is not necessary and will lead to problems when
sodium-crypto-sign      > used on a library. Use the configure flag --enable-library-profiling and/or
sodium-crypto-sign      > --enable-profiling.
[...]
```
* Make 'EmailDomain' effectively abstract.

* Remove feature flags and CPP from core libs.

This is a noisy commit since I ran ormolu right away, but you only
need to know that the packages have more dependencies now even if the
flags are turned off, and that CPP is not enabled any more in all
modules that have been changed.  this is the only change.

This is a good change independently of ormolu, because turning off the
flags wasn't tested anywhere and didn't quite work anyway.
* Create schaffolding for new service for federation.

* Minimum federation API and instances.

* Fixup
* Rename cassandra-schema.txt to cassandra-schema.cql

This will ensure github highlights the code.
Co-authored-by: Arian van Putten <arian.vanputten@gmail.com>
(It's safe to change these, since the machine-readable meaning doesn't change.)
- don't choke in check mode, but report all errors.
- in check mode in case of error, give hints on how to fix this.
* federator: some churn

* move Handle to common-types

* introduce Qualified to types-common

* move EmailDomain to Data.Domain

* define opaque user and conversation IDs

* clarify terminology inside/outside the codebase

* Add Data.IdMapping
* Compile nginx with libzauth using nix/niv

Noteworthy:

both vts and moreheaders are packaged in nixpkgs, so no need
to fetch those git submodules... but perhaps there are version
conflicts. we should check! :)

I also use niv for version management of nixpkgs. instead of hacky own
thing.

* Git ignore rust build artifacts

* Actually add rust build artifacts to gitignore.

It's in the top-level gitignore of wire-server, but not in the one in
libzauth

This is all a bit confusing. We should unify everything in one top-level
gitignore. That makes things a lot less complicated. However,
this is an impurity to tackle in another commit
* Make nginz switch for integration tests reachable from outside the Makefile.

* Cleanup gitignore files a little.
* Test different viewing users (not just viewed).

* Permission check for email visibility in same team.

* Change visible_if_on_team behavior back to "no perms check".

* Cleanup.
* galley: change input API types of minimal endpoints

* make everything compile again

* add FUTUREWORK notes
Before, for building libzauth we were only considering the .gitignore
in the ./lib/libzauth directory, (ignoring the toplevel one) which
didn't contain a Cargo.lock ignore (and shouldn't! lockfiles should be
comitted so builds are reproducible!).  Then we deleted the local
gitignore
which tripped up the nix build as the gitignore didn't exist anymore.
We now use the toplevel .gitignore, but remove the global Cargo.lock
ignore as Cargo.lock's are important to be comitted (and were already!).
For some reason we do not have a Cargo.lock for ./libs/libzauth/libzauth
but only ./libs/libzauth/libzauth-c so I kept it like that.  libzauth-c
is the thing we directly build, and then pulls in libzauth as a
dependency, so  locking at just libzauth-c is sufficient and is also the
lockfile we had in source control already anyway.
This does not change any dependencies from 2.2.

I broke 2.* a while ago by editing the files, and thus learned
that even editing typos in comments is a bad idea.  the easiest
way out is to abandon the old snapshot and clone it into a new
one with the same content.
Still not pretty, we should try to find a better solution.

* ormolu.sh: more robust queries

* ormolu.sh: print version and enabled language extensions

* check that tools are installed
@fisx
Copy link
Contributor Author

fisx commented Mar 10, 2020

Note that the #995 commit is still listed in this branch here on github, even though it's already in master since the last release.

Copy link
Member

@akshaymankar akshaymankar left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Lots of ormolu noise, but looks fine.

@fisx fisx merged commit c50c3b6 into master Mar 13, 2020
@fisx fisx deleted the release_2020-03-10 branch March 13, 2020 12:13
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants