Just like that, yeah. Maybe with @JvmOverloads too, but that's a minor thing

This should not be automatically selected. Instead it should default to production and the API base constructor should take a URL to change the base request URL.

• typos

• Dokkaaaaaaaaaaaaargh

• Little things

• Fix stuff that got asked for

This is why we need a versioned API

How about making "in stack" black (and gray in dark theme ;))?

I'm liking the original Clovis proposal. However:

Superseded by D259

I meant a callback. So you pass a callback into the code to request an object, and it is called with the new object as the only parameter when the cache update is complete, if and only if the object is actually different

Tbh I think we need some system to inform the client when an object has been updated. So that the UI can be updated when an update is performed without closing and reopening something (if you close and reopen a page the old data is displayed because the UI is loaded before the cache is refreshed

All resolved.

Oh, I understand. It's committed to webcl (web client), but its a fix for the webview app, which loads webcl.

In D258#6561, @CLOVIS wrote:

In D258#6553, @Hackintosh5 wrote:

Wait, we have an Android app now? And it uses React?

I think it's about the currently used webapp?

Anyway this doesn't actually fix the issue. Other stuff fails and I won't bother.

• Remove extraneous System.out.println

Wait, we have an Android app now? And it uses React?

Because I'm lazy

Try making a test to edit the server draft (.save().setText().save()) and check the coverage. The method to edit server draft is never called.

isLocalOnly is only set to false in the else clause.
So it's set to false (meaning uploaded to server) only if we edit the server draft. And we only edit the server draft if it's set to false! So it's never changed.

Fixed in D225

In D227#6354, @CLOVIS wrote:

Is this still relevant?

Fix tests

• Don't require active property in JSON as it's not present for Drafts
• Fix draft deletion

Tests pass but an exception is suppressed. I've fixed that exception and I'm fixing the next one, I'll update the diff later. The exceptions are due to errors in main not in the test

No, this is valid

Damn I can't read

Authentication might be difficult.

• Remove extraneous character in comment

• Remove test code

In D223#6352, @CLOVIS wrote:

It looks like the changes here are mainly the same as D247... I know this one was made earlier but D247 is up to date with master and with the new CI system so I think it should be prioritized (unless there's something important here)

Can we get this or something similar in please, I've been trying for months and these deprecation warnings are annoying me

WyldBot, why not fix the library then? I was going to do that but y'all told me the server isn't stable yet. So why aren't we versioning the api server endpoints?

I think you should focus on using wflib rather than rewriting the array.

Wel I have no idea how that works. You can commandeer this and fix it as I don't know what's wrong or how to fix it

So there was never a PGP signature anyway? It only ever existed on your local machine?

Dude. That commit was never signed. I just checked out arc revision 877 which is your original commit. I ran git show HEAD but there's no signature below the commit message.

Remove programming by exception

This comes from D223. I'll deal with it's issues now.

Right, it applies over master now.

Fix minimal-json JAR not being included.

Arc broke. I didn't update the copyrights.

I haven't removed your commits....

Fix minimal-json JAR not being included.

I just searched phabricator the commit is not here, even in unmerged diffs. I'll abandon this and open a new one against master, I suppose

I have no idea when no.
And yes, they must be duplicated.

Fix background invocation

Fix stuff

• Fix bad copyrights
• Don't start API when --norun is specified
• Use rm -f instead of test -f

In D244#6038, @WyldBot wrote:

Add a .gitignore to hide .DS_Store files and all files in the Pods folder

Lint lines are too long. Check my inline comment in AlertChildPageViewController.xib:2

Oh ok.

Needs same changes as D246 and D247

Can you check that the unit tests pass? Gradle skipped them so I guess they do but just to be sure

Again my source has been nuked. Can you just merge the diff?

Hmmm I've nuked my source tree so if it's not there it's not there. Is there a way to just apply the diff?

Something is needed, at least. check is the wrong command; we should use either spotbugsMain spotbugsTest or spotbugsMain.

Tested by manually setting localStorage.blockedUsers = "", now it works.