Support for Drupal 7 is ending on 5 January 2025—it’s time to migrate to Drupal 10! Learn about the many benefits of Drupal 10 and find migration tools in our resource center.
due to #1779034: Convert theme_field to twig and http://drupal.org/node/1727592 we have to update code to latest HEAD
Comment | File | Size | Author |
---|---|---|---|
#23 | merge.patch | 1.64 MB | fubhy |
#10 | merge_core_binary-1779136-10.patch | 1.25 MB | podarok |
#10 | interdiff.txt | 175 bytes | podarok |
#9 | merge_core_binary-1779136-9.patch | 1.25 MB | 5n00py |
#6 | merge_core-1779136-6.patch | 1.25 MB | 5n00py |
Comments
Comment #1
5n00py CreditAttribution: 5n00py commentedpatch for merging sandbox merge_chx_sandbox branch with core 8.x
Comment #2
5n00py CreditAttribution: 5n00py commentedThis patch need fetched 8.x branch in repo because patch have binary data.
Comment #3
andypost@5n00py++
Applies cleanly, suppose this patch should be merged and all other issues needs to be applied as authored commits, probably we need a separate branch or sandbox to maintain the authorship of all 10 sprinters.
Comment #4
jenlamptonSetting back to active since this will need to be re-merged weekly.
Comment #5
podaroklooks like we need just a separate branch, not necessary for separate sandbox as I see
Comment #6
5n00py CreditAttribution: 5n00py commentedTo adding core repo to git remotes and track only 8.x branch
git remote add --no-tags -t 8.x -m 8.x -f core http://git.drupal.org/project/drupal.git
After this patch can be applied.
Now we have conflicts in core/includes/theme.inc
In new patch all conflitcs merged by hands )
Also this patch should fix #1777976: Installer broken
Comment #7
5n00py CreditAttribution: 5n00py commentedStatus.
Comment #8
podaroklooks like #6 and #1 identical
Comment #9
5n00py CreditAttribution: 5n00py commentedOk, now I include binary data to patch.
Interdiff:
Comment #10
podarok#9 - working like a charm ;)
here is a patch w/o trailing whitespace
looks good
applying clearly to merge_chx_sandbox
Comment #11
podarokmerged!
Comment #12
podarokfixed untill next merge
Comment #13
jenlamptonaccidental double post
Comment #14
jenlamptonHi guys! Thanks so much for your help thus far.
We will be doing the weekly merge on the d8-core branch so we have a safe place to break everything and put it back together. Only if we can get everything working there - and if the whole team agrees we are ready - will we merge these changes back in with the front-end/8.x branches.
@Soulston has agreed to do the merge with HEAD weekly for us, so assigning this issue to him. (Please leave it assigned to him) We have decided on a schedule that works out for the whole team. Please do not merge with head without approval from him :)
Thanks!
Comment #15
podarokok
removed tag
Comment #16
jenlamptonWe should talk about this this week, but we need to start merging back in with head.
There are changes going on in 8.x that we need to account for with our template conversions.
For example, if we convert table.html.twig based on the current markup of 8.x HEAD our sandbox will break. If we convert based on the markup in this sandbox, we'll have to update it again later. I'd prefer not to have to update twice, but that might be preferable to a merge, depending on the current holdups.
What are the current reservations about doing a merge with head?
Comment #17
Fabianx CreditAttribution: Fabianx commentedHere is the status update:
I've rebased front-end onto current core and it worked mostly fine.
There are just little issues to fix (like moved components) or wrong merges.
My current WIP is in issue-1779136 branch in the sandbox.
Comment #18
Fabianx CreditAttribution: Fabianx commentedIt is ready:
Checkout here:
issue-1779136 in pixelmord sandbox.
Only preprocess and added template changes left :-) while preserving almost full history :-).
Comment #19
Fabianx CreditAttribution: Fabianx commentedDeleted "front-end" and pushed rebased branch.
Back to active.
Branch issue-1779136 in pixelmord sandbox will always get the newest updates weekly and front-end will be re-based from time to time after announcements.
Comment #20
adamdicarlo CreditAttribution: adamdicarlo commented#1836096: theme('exposed_filters') is broken, causing admin/content to crash for Seven theme. needs review. theme_exposed_filters() was broken, apparently from the latest merge.
I added a link to the issue to the core theme functions spreadsheet and I yellowed out theme_exposed_filters row and put needs review (good process?).
Comment #21
soulston CreditAttribution: soulston commentedI'm wondering what's the best approach for doing this (specifically the changes to core theme functions), and also if there is any way we can automatically test if we broke anything?
Suggestions?
Comment #22
Fabianx CreditAttribution: Fabianx commentedYup, we will only merge what is affecting us like the recent theme item list change.
Comment #23
fubhy CreditAttribution: fubhy commentedI did a full merge today. Not 100% sure that it's completely accurate but 99.9%. :P
Comment #24
fubhy CreditAttribution: fubhy commentedNevermind, it's broken. I would really be very much in favour of merging D8 HEAD again with the working branch from this sandbox. We have to do it eventually anyways and at least for me it was a major pain when I tried to do it today (and, as you can see, I failed). Maybe I did something wrong tho, I'd love to be proven wrong :P.
Comment #25
Fabianx CreditAttribution: Fabianx commented#24: I don't like merges as they make merging back to HEAD later more difficult.
We can merge single issues via cherry-pick, but besides that I am a huge fan of rebasing even on the cost of destroying front-end and having to re-checkout.
There are 3 issues to cherry-pick from changelog.
I'll get to them later today :).
Best,
Fabian
Comment #26
soulston CreditAttribution: soulston commentedIf the only issue with rebasing is that it would mean having to checkout the front-end branch again I think this might be a better option. Are there any other implications? iirc if everyone rebases (see below), then would we even have this issue?
I realise that we are potentially destroying history (http://lwn.net/Articles/328438/), but is this worthwhile considering this is means to a greater end?
Perhaps we should always perform a rebase before applying a patch marked RTBC and remove most people's commit access to the repo to make sure we always rebase?
The time spent cherry picking could be put to much better use I think and it's probably going to get more involved as work on d8 starts to increase.
I'm not sure if this is the right place to ask this but at what point are we going to start posting patches against core?
Comment #27
joelpittet@FabianX would it help or hinder if I merge some of the smaller changes from core over ahead of your big merge? In terms of rebasing will that throw a bunch of conflicts at you or help let you concentrate on the bigger pieces?
Comment #28
Fabianx CreditAttribution: Fabianx commentedAfter call yesterday: We will rebase after Dec 1st.
joelpittet: You can easily cherry-pick changes from core (if you need them), but I will need to remove all of them during rebasing via git rebase --skip.
Comment #28.0
Fabianx CreditAttribution: Fabianx commentedUpdated issue summary.