Delete pybundles

[dstufft]

This removes the bundle functionality that was deprecated in 1.4. Originally there was some talk about waiting to remove this until we split it out into it's own command and possibly have some extension support. I don't think it makes sense to wait because none of us are motivated to actually maintain it and I don't see why splitting it out into a second repository is going to change that. Additionally even if we do split it out we (or someone else) can pull it from the VCS history so keeping it around in the branch isn't winning us much. Finally we don't need extension support because there isn't much difference between typing pip bundle vs pip-bundle and anyone can provide a pip-bundle command without any help from us.

Closes #68
Closes #236
Closes #110
Closes #164
Closes #438
Closes #450
Closes #534
Closes #794
Closes #977
Closes #978
Closes #1046

[dstufft]

Ok, so @pypa/pip-developers what do we think?

Obviously I vote yes, let's kill this, we deprecated it, none of us want to maintain it, keeping it in is just letting it slowly wither and die instead of killing it outright.

[pfmoore]

I'm happy for it to go

[qwcode]

to be clear, I think the idea of bundles has merit for repeatability (which we don't have right now easily; consider peep's existence) and the simplicity of moving around one file for deployment or sharing across environments.

but what we have isn't being maintained, and could use a total refactor in light of wheels. like you said, I had brought up before holding on this until we (or I) had moved all this out into an extension (for the sake of current users), but I haven't done it, so I can't really ask you to hold up, if you want to move on.

[pfmoore]

Thinking about this a bit further, do we have people using bundles? I remain happy to delete the feature if no-one is using it. But if it is being used, do we want to deliberately break their use of pip with no replacement functionality?

[dstufft]

There are (were?) a handful of people, they saw the deprecation messages and raised an issue, one or two of them even said that they were going to help out maintaining it and I don't believe we ever saw a PR from them. The deprecation warning has been going since 1.4 and there hasn't been more than a handful even mention it. Given that there hasn't been much outcry, and nobody apparently cared enough to actually help maintain it I don't think keeping it around is that useful.

It shouldn't be very difficult to build a replacement using wheels and tarfile - I just have no interest in doing so.

[dholth]

Get rid of it.

On Thu, May 8, 2014, at 05:19 PM, Donald Stufft wrote:

There are (were?) a handful of people, they saw the deprecation
messages and raised an issue, one or two of them even said that they
were going to help out maintaining it and I don't believe we ever
saw a PR from them. The deprecation warning has been going since 1.4
and there hasn't been more than a handful even mention it. Given
that there hasn't been much outcry, and nobody apparently cared
enough to actually help maintain it I don't think keeping it around
is that useful.

It shouldn't be very difficult to build a replacement using wheels
and tarfile - I just have no interest in doing so.

—

Reply to this email directly or [1]view it on GitHub.
[208018__eyJzY29wZSI6Ik5ld3NpZXM6QmVhY29uIiwiZXhwaXJlcyI6MTcxNTIwMzE5MS
wiZGF0YSI6eyJpZCI6MzE3NzE0OTR9fQ==--11ad975502ce63c32c9c634a69fcd0338f8
3ae4e.gif]

References

1. Delete pybundles #1806 (comment)

[qwcode]

Get rid of it.

from the wild blue yonder....

[Ivoz]

@ThePyPA should tweet this fact!