-
Notifications
You must be signed in to change notification settings - Fork 428
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
12 3.19.4 bits #3952
12 3.19.4 bits #3952
Conversation
Sorry about this @mbargull
…plicit-inclusion-only when given as a list
… or dirty are set .. same thing with host prefix. The code and concepts behind --dirty, --keep-old-work and also perhaps build/no_move_top_level_workdir_loops seem a bit muddled and overlapping. --dirty should *only* deal with finding a new croot, and have no part in whether to keep stuff around or not. That should be left to a new option, like: --keep=work,build-prefix,host-prefix,test-prefix,test-tmp where --keep-old-work is deprecated but enables all of these. Also, 'keep' might not be the best word, call it what we do, --retain-via-rename=work,..
…ors from the last command do not exit the interactive shell (very annoying when testing package build failures). We then must 'set -e' just after sourcing the env script.
Reverse failed patches so they do not leave behind newly created files for bits that manged to apply OK (causing subsequent patches to potentially fail). Rewrite apply_patch to behave more sanely, esp. in the face of fully-correct, binary patches. .. also, try the same patch modifications we try on Windows on the other OSes. There is no reason not to.
These files are left behind by macOS when it needs to ratain its extended attributes through round-trips to 'foreign' file-systems. And do we need two different bits of code globbing for meta.yamls? No, we do not.
Changes :: Uses lief 0.10.1 now (no code changes were needed for this update) Fixes _map_file_to_package for finding '.dylib' and '.so' without suffixes beyond that Added our FakeDist to run-prefix, improve library_nature, do not fail if ccache exists Fixes detection of where packages comes from. We must avoid using 'canonical' channel names (such as 'default', needing instead 'pkgs/main') when determining which channels packages come from. To do this, we use a new function, linked_data_no_multichannels(). Let's hope we never end up comparing these with ones gotten via linked_data() Add local_output_folder to get_build_index()'s locally cached things. Ensure we pass 'channel_urls' as a list and add the 'local' channel to that. Disambiguate 'prefix' DSO files in the build/host throughout ovelink detection. Change ERROR :: {} not in prefix_owners to a more helpful warning and add a comment though it might be better to just remove this warning and let the final info detail any issues (for they will show up as overlinking errors). Notes :: This is still not the speed boost that this stuff needs, so I am hoping the update to 0.10.1 helps with that. Looking at the history, it should, somewhat. Case-insensitive lookup for Windows (hopefully only for files in the sysroot - which is being referred to as the whitelist in the messages ATM unfortunately - still remains to be done (and will probably get squashed into this commit).
@jjhelmus preview, I will also check some community PRs once you (and ci) are happy with all this! |
Thank you for your pull request and welcome to our community. We could not parse the GitHub identity of the following contributors: distro-bot@anaconda.com.
|
Thank you for your pull request and welcome to our community. We could not parse the GitHub identity of the following contributors: distro-bot@anaconda.com.
|
.. New link location: None for example
too liberal and we may want to do it only for Windows exes/dlls.
1ed181d
to
5cb5d01
Compare
Thank you for your pull request and welcome to our community. We could not parse the GitHub identity of the following contributors: distro-bot@anaconda.com.
|
2b4aa0c
to
119e434
Compare
…lease The downside is that some packages may not get picked up. I need to check that though. It cannot remain this broken. The problem seems to be that depending on how the channel was identified to conda when it made the Dist is how it will remain and one created from 'defaults' is not equal to one created from 'pkgs/main'. During conda-build post, our channels seem to come from 'pkgs/main' but at other times they come from 'defaults'.
I think this is good to go now. If you have time and agree, I'll look over community PRs after some food. |
Hi there, thank you for your contribution! This pull request has been automatically locked because it has not had recent activity after being closed. Please open a new issue or pull request if needed. Thanks! |
No description provided.