Bug#1063274: pydevd: autopkgtest-failing warning with pandas 2.1

[Julian Gilbey]

On Mon, Feb 19, 2024 at 08:03:34AM +0000, Rebecca N. Palmer wrote:
> This has been merged but not uploaded - is there a reason it shouldn't be,
> or have you just not had time?

Hi Rebecca,

Yes; I've upgraded to the latest version in my local repo, which
includes this patch upstream, but it's now causing FTBFS issues in
debugpy, which I haven't managed to track down yet.

:-(

Julian

[Julian Gilbey]

On Mon, Feb 19, 2024 at 10:36:29PM +0000, Rebecca N. Palmer wrote:
> Thank you for caring about not breaking other packages, and yes, that's a
> good reason to not upload that new upstream for now.
>
> Does just the patch (not the new upstream) also break debugpy? (It
> shouldn't be able to, since it only touches test code.)

It's taking time to locate the source of the problem. It's a bit
messy: debugpy vendors pydevd, pydevd in turn vendors bytecode. I've
split out bytecode and pydevd into separate packages. Unfortunately,
the bytecode version in pydevd is behind that in Debian, though that
should not be too much of a problem. The native version of debugpy
(with all other packages being Debian ones) passes all the tests in a
Debian testing lxc container, but with the Debian versions of
bytecode, pydevd and debugpy, it doesn't. So something's gone wrong
somewhere, but it's taking some effort to locate (and hopefully fix!)
the cause.

Best wishes,

Julian

[Rebecca N. Palmer]

Is that a yes to>> Does just the patch (not the new upstream) also break
debugpy?or have you not tried specifically that?

(I'm looking for a quick fix for the autopkgtest fail to unblock the
pandas 2.x transition. I agree that upgrading to a new upstream is a
good idea in the long run.)

> the bytecode version in pydevd is behind that in Debian,

This is currently true (Debian has python3-bytecode 0.15.1, upstream
pydevd 2.10.0 vendors 0.13.0.dev), but if it were a problem I'd expect
it to mean that pydevd/debugpy were *already* buggy, not that my patch
makes them worse.

[Julian Gilbey]

On Tue, Feb 20, 2024 at 09:46:16PM +0000, Rebecca N. Palmer wrote:
> Is that a yes to>> Does just the patch (not the new upstream) also break
> debugpy?or have you not tried specifically that?
>
> (I'm looking for a quick fix for the autopkgtest fail to unblock the pandas
> 2.x transition. I agree that upgrading to a new upstream is a good idea in
> the long run.)

Fair point; I've just uploaded 2.10.0+ds-9.

Best wishes,

Julian

[Julian Gilbey]

On Tue, Feb 20, 2024 at 09:46:16PM +0000, Rebecca N. Palmer wrote:
> Is that a yes to>> Does just the patch (not the new upstream) also break
> debugpy?or have you not tried specifically that?
>
> (I'm looking for a quick fix for the autopkgtest fail to unblock the pandas
> 2.x transition. I agree that upgrading to a new upstream is a good idea in
> the long run.)

Dear Rebecca,

A quick update on this: version -9 failed its autopkgtests on some
archs, so I uploaded version -10 this morning. Hopefully that will
pass and migrate in a couple of days.

On the pandas front, though, I don't know if you've seen the roadmap:
version 2.2 is released, and version 3.0 is in preparation. But 3.0
will depend on PyArrow - the string data type will use PyArrow as a
backend rather than NumPy as at present. This is
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=970021
and it looks like this is going to be a monster to package. It's
probably worth getting a small group of developers together to work on
this soonish - I don't have the capacity to take it on, I'm afraid.

Best wishes,

Julian