-
Notifications
You must be signed in to change notification settings - Fork 215
Move to migrated platform terminal #1233
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
base: main
Are you sure you want to change the base?
Conversation
@ruspl-afed as discussed in the IDE call yesterday, this is the PR for adjustments in CDT to migrate to the new terminal view, would be great if you can help with this as well. |
Please have a look here |
The feature is still there! So this should not be the problem, it currently fails for some org.bouncycastle stuff, maybe I should try first a PR that uses ibuilds repo to not change too much things at once.... |
The bouncycastle version issue seems unrelated to this PR. I have been experimenting on #1241. It would be preferable to use a separate PR for releng changes, but I won't have time to look at this again today. Please feel free to work on a solution yourselves. |
@laeubi I plan to continue looking at your changes today, please let me know about your plans. |
I think needs investigation first and if that works I'll rebase my PR to see what else fails (maybe). |
looking into it. may be just an infra failure |
at least license check converged after restart (was ClearlyDefined failure) |
I think |
|
See here: #1241 the test seem to timeout currently :-\ |
@laeubi it looks like https://github.com/eclipse-cdt/cdt/blob/main/core/org.eclipse.cdt.ui.tests/src/org/eclipse/cdt/ui/testplugin/EditorTestHelper.java#L376 was broken by the recent platform UI change: eclipse-platform/eclipse.platform.ui@0e4942b |
Renaming a private (!) field of an internal class can merely be seen as breaking change .. I'm not sure why CDT do the same bad hacks as Platform did, but it has to do the same adaption than here then: |
CDT needs to update its hack to follow internal changes in Platform eclipse-platform/eclipse.platform.ui@0e4942b#diff-54c81a16536eba9489797a9d665c195f0908163e2c7b14ad055d78cc62579fdaL193
CDT needs to update its hack to follow internal changes in Platform eclipse-platform/eclipse.platform.ui@0e4942b#diff-54c81a16536eba9489797a9d665c195f0908163e2c7b14ad055d78cc62579fdaL193
CDT needs to update its hack to follow internal changes in Platform eclipse-platform/eclipse.platform.ui@0e4942b#diff-54c81a16536eba9489797a9d665c195f0908163e2c7b14ad055d78cc62579fdaL193
CDT needs to update its hack to follow internal changes in Platform eclipse-platform/eclipse.platform.ui@0e4942b#diff-54c81a16536eba9489797a9d665c195f0908163e2c7b14ad055d78cc62579fdaL193
@laeubi now CDT uses the later Platform I build to simplify Terminal-related changes |
84d70a7
to
49ebebd
Compare
Rebase done, lets see if it works out better :-) I think I should probably squash all commits into one if everything works. |
Please have a look at product manifest of |
It was actually the feature using a wrong version here... |
Some more feature cleanups ... |
Sorry for the wrong pointer, I hope that at least direction was right :) |
Yes no problem, now the build actually starts and resolves it gives better errors to work on... code cleaniness check seem needs adjustments at least I see some exceptional handling from @jonahgraham last time something was moved to platform... so we are getting closer! |
for code cleanliness failure the message is
|
b529ac7
to
ca2fd97
Compare
I'm thinking how we can converge on this. May be we can split this PR and first transfer CDT-DSF to a new Platform Terminal? |
Some thing might be able to be extracted, but I think there is not much benefit and will possibly result in conflicts in the build. I think I'm almost done with that and then I can try to came up with trying it in smaller chunks, e.g. currently the actual deletion of stuff is the far biggest change here, so one might simply let linger around the deleted stuff and then have one PR that actually delete it (but nothing else changes). I also bumped the major version for the both CDT connectors from 4 > 5, since they are using new API in their own API it seems not useful to not do it. One might think about after doing the migration to also clean up these remaining connects and maybe even change their namespace as well to prevent confusion with the old ones. The also their version can be reset and maybe not export everything as API... e.g. But I would leave this aspects to the CDT maintainer to decide! |
My congrats with the first fully green build @laeubi ! There are two kind of things that I cannot merge without visa from CDT PL @jonahgraham
What I think I can merge is switch of CDT itself to the new Platform Terminal |
I think that is the cleanest solution here, as it depends on new API that is leaked into the bundles own API. An alternative would be the rename / reversion of the whole thing. In the end I think no one should actually API wise depend on a connector anyways, but it is API at the moment sadly.
If one depends on anything from the old world, it is still there in the old releases an/or branches, it does not seem useful to retain the sources here if they have evolved/moved to another project I think. Also I think usual users/extenders of CDT should never notice that particular thing:
|
8d13390
to
7c24da0
Compare
I now squashed everything, so this is ready for review, will try to extract some aspects into separate PRs now. |
GDB Migration: |
remote migration: |
7c24da0
to
d49db2d
Compare
Migration of remaining connectors to new API including major version bump: |
Using of the new bundle names from platform in feature files: |
d49db2d
to
90d2f50
Compare
Thank you for supporting this PR in actual state @laeubi ! |
@ruspl-afed are you able to merge this? You have done the work to review so far. thanks @laeubi for this PR. |
@jonahgraham I think I should rebase the PR and have already creates smaller chunks that should be merged first:
so I will convert this to draft, @ruspl-afed has had some concerns but these are maybe better discussed on the smaller PRs. |
90d2f50
to
61ad699
Compare
FYI @jonahgraham