You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The changes since last release mainly consist of the C++20 headers being added. I think this is worth making a new release for, maybe even add the C++23 headers along?
If there are sufficient changes on the master or develop branch, and CI passes, I can make a release, or you and @jayvdb can try ans see what Accesses you need. The CONTRIBUTING guide has the instructions that I usually follow.
I have time today and over the weekend. I'll poke around. It looks like Python 3.12 isnt working. setup.py contains from distutils.spawn import find_executable which doesnt exit in py312
Apparently we'd need a newer version of setuptools for that, which would reintroduce deprecation errors due to our still using setup.py. 3.12 came out quite recently, so I don't think it's particularly urgent to get support for it done.
@jayvdb I don't think that is a good idea at this point, as that file has the release dates which we don't know yet and would need countless commits to bump every time it is pushed back. I think we should review the draft release's changelog right now, wait for @tkruse 's availability and input, and then do the rest of the procedures on the same day.
Yes, I know that. That is what I meant by "the rest of the procedures". I was thinking about keeping commits for the same thing to a minimum, but I've also just realized that we can squash and merge.
If you need me to go through the release procedure, you can tag a git commit on master as a release candidate if you both agree on one such commit. Then I can try a release of that in my own time. Else tell me your pypi Ids so I can invite you
Bugfixes have priority over new features, but if a problem was already present in the previous release, it does not need to be fixed for the next one. I don't know whether #250 is old or new.
#250 is probably at least partially old; one of the two aspects of it feels old, and might be a problem with tests rather than the code. But both aspects felt relatively easy to at least isolate, which will conclusively say whether old or new, whether test code or prod code, and may lead to an easy solution. I'd prefer to know the answers to these questions. I'll work on it as soon as the existing Windows PR is merged.
My account on both PyPIs is aliu, though I'd be more comfortable to have jayvdb release.
I don't think the samples use C++14 or C++17 features either. But we can start with updating the sampled code to their latest upstream versions. For example the Chromium sample's upstream now includes a C++17 header. I don't know any popular projects that prominently use C++20 or C++23 though. Maybe we should combine cfg-file with silly-sample and add some new language features to silly-sample?
Activity
tkruse commentedon Jan 29, 2024
If there are sufficient changes on the master or develop branch, and CI passes, I can make a release, or you and @jayvdb can try ans see what Accesses you need. The CONTRIBUTING guide has the instructions that I usually follow.
aaronliu0130 commentedon Feb 2, 2024
@tkruse I have a changelog draft at https://github.com/cpplint/cpplint/releases. What do you think? Should we go forward with it?
jayvdb commentedon Feb 2, 2024
I have time today and over the weekend. I'll poke around. It looks like Python 3.12 isnt working. setup.py contains
from distutils.spawn import find_executablewhich doesnt exit in py312aaronliu0130 commentedon Feb 2, 2024
Apparently we'd need a newer version of setuptools for that, which would reintroduce deprecation errors due to our still using setup.py. 3.12 came out quite recently, so I don't think it's particularly urgent to get support for it done.
jayvdb commentedon Feb 3, 2024
@aaronliu0130 , please put your notes at https://github.com/cpplint/cpplint/releases/tag/untagged-cb79b2ed6027cea1cd0c into https://github.com/cpplint/cpplint/blob/develop/changelog.rst as a PR for review
aaronliu0130 commentedon Feb 3, 2024
@jayvdb I don't think that is a good idea at this point, as that file has the release dates which we don't know yet and would need countless commits to bump every time it is pushed back. I think we should review the draft release's changelog right now, wait for @tkruse 's availability and input, and then do the rest of the procedures on the same day.
jayvdb commentedon Feb 3, 2024
@aaronliu0130 the normal way to solve that is to do:
Then there is a final change at the time of release to set the release date.
We can not release without a updated
Changelog.rst. That is how this project distributes the release notes to users & package maintainers.This is documented at https://github.com/cpplint/cpplint/blob/develop/CONTRIBUTING.rst#releasing
aaronliu0130 commentedon Feb 3, 2024
Yes, I know that. That is what I meant by "the rest of the procedures". I was thinking about keeping commits for the same thing to a minimum, but I've also just realized that we can squash and merge.
tkruse commentedon Feb 18, 2024
So is there anything blocking you both from doing a new release yourselves?
[-]Make new release?[/-][+]Make new release 1.7.0?[/+]tkruse commentedon Feb 18, 2024
If you need me to go through the release procedure, you can tag a git commit on master as a release candidate if you both agree on one such commit. Then I can try a release of that in my own time. Else tell me your pypi Ids so I can invite you
jayvdb commentedon Feb 18, 2024
My account is https://pypi.org/user/John.Mark.Vandenberg/
tkruse commentedon Feb 18, 2024
Invite sent, I can be available today for more help. Though I don't know the current state of the main branches, a lot has changed recently
jayvdb commentedon Feb 18, 2024
Would you be able to break a deadlock at #247
tkruse commentedon Feb 18, 2024
I added my opinion. Can you also add an account on https://test.pypi.org/manage/project/cpplint/collaboration/ (test pypi for testing the release process in a safe space)?
5 remaining items
tkruse commentedon Feb 18, 2024
Bugfixes have priority over new features, but if a problem was already present in the previous release, it does not need to be fixed for the next one. I don't know whether #250 is old or new.
jayvdb commentedon Feb 18, 2024
#250 is probably at least partially old; one of the two aspects of it feels old, and might be a problem with tests rather than the code. But both aspects felt relatively easy to at least isolate, which will conclusively say whether old or new, whether test code or prod code, and may lead to an easy solution. I'd prefer to know the answers to these questions. I'll work on it as soon as the existing Windows PR is merged.
aaronliu0130 commentedon Feb 18, 2024
My account on both PyPIs is
aliu, though I'd be more comfortable to have jayvdb release.I don't think the samples use C++14 or C++17 features either. But we can start with updating the sampled code to their latest upstream versions. For example the Chromium sample's upstream now includes a C++17 header. I don't know any popular projects that prominently use C++20 or C++23 though. Maybe we should combine cfg-file with silly-sample and add some new language features to silly-sample?
[-]Make new release 1.7.0?[/-][+]Release 1.7.0[/+]aaronliu0130 commentedon Mar 31, 2024
Since we're removing Python 2 support and we have so many new changes, maybe we should just call this 2.0?
aaronliu0130 commentedon Mar 31, 2024
Also, shouldn't the "changelog" in "changelog.rst"'s name be in all-caps?
tkruse commentedon Aug 2, 2024
I would also be ok to advance to 2.0. Also I may have a bit more time the coming weeks if my support is required.
aaronliu0130 commentedon Aug 2, 2024
(I've registered my account as
aliuon both instances of pypi)aaronliu0130 commentedon Sep 17, 2024
@tkruse It would be really helpful if you gave me PyPI permissions and added another maintainer, maybe @geoffviola.
anonrig commentedon Sep 28, 2024
Lack of C++20 features in 1.6.1 forces us to disable a rule in cpplint at Node.js. What is the timeframe for a new version? nodejs/node#55155
[-]Release 1.7.0[/-][+]Release 2.0[/+]aaronliu0130 commentedon Oct 3, 2024
@anonrig I need to get someone with permissions at PyPI (realistically either @tkruse or @jayvdb) to release a new version. It could happen anytime.
jayvdb commentedon Oct 5, 2024
Just noting there are not changes upstream in https://github.com/google/styleguide/tree/gh-pages/cpplint for two years, so no need to sync anything.
I've created #292 to capture any release related changes needed before we pull the pin in the next few days.
jayvdb commentedon Oct 6, 2024
It is done
https://pypi.org/project/cpplint/2.0.0/
https://github.com/cpplint/cpplint/releases/tag/2.0.0