Upcoming v0.2.2 release

[Erik Bray]

Hi all,

Apparently I've scheduled the release of Astropy v0.2.2 for tomorrow.
We already have 30 closed issues in this release, at least a handful
of which are non-trivial, so it's definitely worth doing a release.

But of the ~50 open issues still remaining for 0.2.2 (most of which
would be delayed 'til the next release) are there any issues anyone
knows of that are major blockers, and should be fixed before the next
release?

Thanks,

Erik


P.S. At some point there also needs to be more discussion of the major
pieces we want for 0.3 are, and a timeframe for getting them all in.
We've already added the modeling package and the cone search so that's
a lot. So the question is, what else do we consider critical features
for 0.3?

[Tom Aldcroft]

On Mon, May 13, 2013 at 2:26 PM, Erik Bray <erik....@gmail.com> wrote:

Hi all,

Apparently I've scheduled the release of Astropy v0.2.2 for tomorrow.
We already have 30 closed issues in this release, at least a handful
of which are non-trivial, so it's definitely worth doing a release.

But of the ~50 open issues still remaining for 0.2.2 (most of which
would be delayed 'til the next release) are there any issues anyone
knows of that are major blockers, and should be fixed before the next
release?

I triaged the open 0.2.2 issues where I have some responsibility and kicked back about 10 issues to either 0.3 or future.  None of the ones I looked at needed to be 0.2.2. 

- Tom

 

Thanks,

Erik


P.S. At some point there also needs to be more discussion of the major
pieces we want for 0.3 are, and a timeframe for getting them all in.
We've already added the modeling package and the cone search so that's
a lot. So the question is, what else do we consider critical features
for 0.3?

--
You received this message because you are subscribed to the Google Groups "astropy-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to astropy-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Thomas Robitaille]

Hi Erik,

On 13 May 2013 20:26, Erik Bray <erik....@gmail.com> wrote:
> Hi all,
>
> Apparently I've scheduled the release of Astropy v0.2.2 for tomorrow.
> We already have 30 closed issues in this release, at least a handful
> of which are non-trivial, so it's definitely worth doing a release.
>
> But of the ~50 open issues still remaining for 0.2.2 (most of which
> would be delayed 'til the next release) are there any issues anyone
> knows of that are major blockers, and should be fixed before the next
> release?

Should we create a 0.2.3 milestone so we can officially triage the
issues that don't need to make it to 0.2.2 but should be in the 0.2.x
branch?

> P.S. At some point there also needs to be more discussion of the major
> pieces we want for 0.3 are, and a timeframe for getting them all in.
> We've already added the modeling package and the cone search so that's
> a lot. So the question is, what else do we consider critical features
> for 0.3?

Good idea - in my view, the most important thing to do for 0.3 at this
stage is sorting out the whole Quantity infrastructure (whether it
inherits from ndarray, etc.) and array-fying astropy.coordinates (in
addition to extensively testing vo and models of course). It'd also be
nice to refactor the Table unified I/O as discussed in some of the
issues.

Cheers,
Tom

[Erik Tollerud]

I just went through the issues that I'm somewhat familiar with. I
traiged some, and have listed below the items that seem important or
easy to get in:

These look like they should go into 0.2.2 because they are basically
done (or just docs/notes):
#1050 (looks ready to merge, but might want to give it until tomorrow
in case others want to look at it)
#843 (ready to merge now if it doesn't break anything in Windows)
#292 (works, but just needs to be made to be adjusted to recognize
temporary directories correctly in OS X)
#988 (just involves listing a "known issue")
#940 (Erik B, are you planning to document this actually as part of
the release process?)


#849 seems rather critical, but I have no idea how much work a fix is
going to be - Mike D or Tom A, is this something we should triage, or
might you be able to get to it in a day or two?


#983 seems moderately important, and it's assigned to me, so I'll try
to get it solved by tomorrow (if not, just go ahead without it).

--
Erik

[Erik Tollerud]

> Good idea - in my view, the most important thing to do for 0.3 at this
> stage is sorting out the whole Quantity infrastructure (whether it
> inherits from ndarray, etc.) and array-fying astropy.coordinates (in
> addition to extensively testing vo and models of course). It'd also be
> nice to refactor the Table unified I/O as discussed in some of the
> issues.

+1 from me on all of this, with the addition of the other significant
API changes for coordinates (i.e., the things that would be backwards
incompatible if we pushed them to the next version).

> Cheers,
> Tom
>
>>
>> --
>> You received this message because you are subscribed to the Google Groups "astropy-dev" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to astropy-dev...@googlegroups.com.
>> For more options, visit https://groups.google.com/groups/opt_out.
>>
>>
>
> --
> You received this message because you are subscribed to the Google Groups "astropy-dev" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to astropy-dev...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

--
Erik

[Erik Bray]

We obviously haven't released v0.2.2 yet. There are a handful of
issues that are *so close* to being fixed that we're trying to finish
up, and in the meantime that's given me a little more time to work on
closing a few additional minor issues. We've also been held up by
some problems with testing (not the tests themselves, but just the
test systems). So if anyone else has any issues for 0.2.2 they think
they can finish up there's still some time, though I still hope to
have the release out by the end of this week.

Thanks,
Erik