Support for Django 5.0 and python 3.12 + Fix CI

[pfouque]

• Add Django 3.2, 4.0, 4.1, 4.2, 5.0 support

• Remove support for deprecated Django versions

• Add Python 3.8, 3.9, 3.10, 3.11, 3.12 support

• Remove support for deprecated Python versions

• Fix CI

[apmathento]

Hello! How long will it take aprox? Thank you! :)

[Pietro395]

Hello! How long will it take aprox? Thank you! :)

If your comment refers to me I think I'll try it between tomorrow and the day after tomorrow

[coddingtonbear]

Thanks for helping out by testing this -- I really, really appreciate the effort! Assuming you don't discover any interesting problems, and you or somebody else checks on a few things I've noted below, I can commit to pulling a release together in the next few days.

There are just a few things I think might need some attention from you or anybody else who might have a little time:

1. Verifying the processincomingmessage e-mail ingestion management command. I think that would be just as easy as cat-ing an e-mail message to python manage.py processingomingmessage SOMEMAILBOXNAME where SOMEMAILBOXNAME is literally any string (it'll create a mailbox with that name if it doesn't already exist) and your e-mail message is raw e-mail text (see https://github.com/coddingtonbear/django-mailbox/tree/master/django_mailbox/tests/messages for some options, but, really, anything will do).
2. Verifying the Office365 integration. You've said you're planning to look into this already, @Pietro395.
3. I assume there are no backward-incompatible changes around this, but if there are, somebody let me know so I can be sure to increment the version number's major digit as necessary.

I mentioned above that once the above items are looked into I can pull together a release, but I do have to warn you all that I have some time constraints around this in that I'll be unavailable for several weeks starting next week; so if we're hoping to have a release before mid-January, I'm afraid there's a little bit of a crunch. If some folks can confirm the things noted above before midnight GMT on Friday night (8 December), I'm confident I can get the release sorted out, though.

[Pietro395]

1. Verifying the `processincomingmessage` e-mail ingestion management command.  I think that would be just as easy as `cat`-ing an e-mail message to `python manage.py processingomingmessage SOMEMAILBOXNAME` where `SOMEMAILBOXNAME` is literally any string (it'll create a mailbox with that name if it doesn't already exist) and your e-mail message is raw e-mail text (see https://github.com/coddingtonbear/django-mailbox/tree/master/django_mailbox/tests/messages for some options, but, really, anything will do).

(env) [intacloud@intacloud-test django-struttura]$ cat /home/intacloud/django-mailbox/django_mailbox/tests/messages/message_with_attachment.eml | python3 manage.py processincomingmessage HELLOWORLD2 INFO 2023-12-05 21:57:54,103 Message received from Adam Coddington test@adamcoddington.net

Looks good to me!!

[Pietro395]

2. Verifying the Office365 integration.  You've said you're planning to look into this already, @Pietro395.

I tested the integration with Office365 by authorizing a new tenant and trying to receive emails. I haven't encountered any problems, everything seems to work fine

(env) [intacloud@intacloud-test django-struttura]$ python3 manage.py getmail
INFO 2023-12-06 17:37:12,397 Gathering messages for HELLOWORLD
INFO 2023-12-06 17:37:12,397 Gathering messages for HELLOWORLD2
INFO 2023-12-06 17:37:12,397 Gathering messages for donotreply
INFO 2023-12-06 17:37:12,580 Gathering messages for admin_office365
INFO 2023-12-06 17:37:13,612 Received Test for Office365 (from ['pietro@xxxx.it'])

@coddingtonbear I believe you can create the release :rocket:

[coddingtonbear]

Released as part of 4.9.0.

I have to say, though, that I couldn't get the tests running in the short period of time I had to look into this -- apparently many of the testing dependencies are out-of-date.

I wonder if I could trouble any of you to see if you might be able to get the tests running... because it sure feels pretty weird to me to push up code without running tests, but I understand that getting a release together is important to some of you so I've decided to, I guess, cross my fingers that nothing unexpected is broken?

It should be just as easy as installing test dependencies in test_requirements.txt and running pytest, if I remember correctly, but it has honestly been a very long time since I've run tests on this particular project.

[pfouque]

Thanks a lot @coddingtonbear ( and @Pietro395 for your support) I did run the tests using tox and it was all good. I even fixed the CI, but I made a small mistake with the name of the master branch in the workflow file. If we replace 'main' by 'master' it should run on the main branch (in addition to the pull requests). Now we can consider the project is back on track, and a bit of triaging in the PR/issues would be nice. Maybe we could have some labels to identify the outdated ones before closing them.

[Pietro395]

Thanks a lot @coddingtonbear ( and @Pietro395 for your support) I did run the tests using tox and it was all good. I even fixed the CI, but I made a small mistake with the name of the master branch in the workflow file. If we replace 'main' by 'master' it should run on the main branch (in addition to the pull requests). Now we can consider the project is back on track, and a bit of triaging in the PR/issues would be nice. Maybe we could have some labels to identify the outdated ones before closing them.

Thanks for your contribution @pfouque and for your time @coddingtonbear 😊, I also think we should tag the old issues to start from scratch

[coddingtonbear]

Hey there @pfouque & @Pietro395 -- You've already noticed that I've added you both as maintainers on this project. I did that because you have both already put forth more more than enough effort to demonstrate that I can trust you both to make good choices around things. Don't feel like I'm putting an obligation on you, though -- of all people, I can understand that you probably have other priorities, but I do thank you for any help you can provide!

As far as guidelines, I think in general the following sound reasonable to me:

• I trust you both to merge things that are obviously safe and unlikely to break anything -- ideally after it has been reviewed by one maintainer (other than whoever wrote the original PR, in the case of one of us making a PR), but please try your best to not break the tests. They don't offer us a huge amount of safety on this particular project, but they're better than nothing, and ideally we'd have more rather than fewer.
• Ideally PRs adding features should be weighed against the long-term maintenance overhead that they might add. In my experience, one of the main things that kills projects like this one is adoption of complicated, niche features that few people use and understand the intent of well enough to understand what other changes might break it. If you decide to merge a new feature, please think about how you (or somebody else) might support that feature years down the road -- very often the original contributors disappear after the initial PR; so you can't rely on them to support it.
• For the moment, I can handle the releases for us, but eventually, I think it's probably reasonable to wire things up to github actions (assuming we end up having enough releases for that to make sense). Just let me know when you think you've added a feature worth releasing and (for semver reasons), whether that release includes any breaking changes (no matter how minor), a new feature, or just a fix for an existing feature.

We should probably set up a channel for just normal conversations about things, though. There's already a public gitter channel for this project, but maybe we should exchange contact info so we can figure out a place for us to chat about things privately? I already have your e-mail, @Pietro395, but maybe you could e-mail me, @pfouque so I can start an e-mail conversation among the three of us to figure out where we'd like that chat to be?

Cheers, and thank you so much for all of your help so far!

[pfouque]

Hi @coddingtonbear, Thanks a lot for that; it's warmly welcomed!

maybe you could e-mail me, @pfouque so I can start an e-mail conversation among the three of us to figure out where we'd like that chat to be?

I suppose you have access to it now that I accepted the invite. TBH, I'm not sure to follow gitter as a new canal, but I'm fine with email or we could have a issue-only repository to hold the private discussion all in Github.

For the moment, I can handle the releases for us

👍 The last release is already a great achievement! And IMO the scope of the project is "small" enough so that it should not require any monthly release plan.

please try your best to not break the tests.

I fixed the CI, especially for that! ;)

Ideally PRs adding features should be weighed against the long-term maintenance overhead that they might add

(I'm applying the same rule to my corporate project contributions! 😜)

very often the original contributors disappear after the initial PR

About that, we have plenty of stalled issues and PRs...making the project look "unmaintained". Since the project is now back on track (with currently supported versions of Django),it should be easier to welcome new contributors. I would consider triaging as a priority, especially the items without any PR, evidence, or useful information. Even for those who have all of this can be tedious to diagnose whether the problem has been resolved in the meantime or not. The fact is that most of the issues are on a now deprecated version, so I would vote for a pretty aggressive cleaning after some basic investigation (contacting a 3-year-old contributor has little chance of success).