Dropping python 3.7 support

[bsipocz]

Before I drop support, would like to see some input from the bigger teams in astroquery. Would you mind if we drop it?
(most big packages are dropping 3.8 (numpy, scipy, etc) in the upcoming releases, and dropped 3.7 a while ago already, and some of the infrastructure packages we use are dropping support, too. So while we can keep going, it's a tiny bit easier if we could do it without workarounds, or to use nice new features)

@astropy/astroquery-archive-maintainers

[keflavich]

I'm good with that.

[jd-au]

+1 Looks like 3.7 loses security support very soon, so its a good time to drop support.

[andamian]

Yes, that works for 3.7 since it will reach its end of life soon.

We found it easier to justify to our users (and convince them to update their environments) if we follow Python release schedule but we don't rush it until something stops working. We have to be mindful that users sometimes run software on systems that are out of their control and a Python upgrade might require an OS upgrade with sys admin intervention and other complications.

[ceb8]

I'm good with this.

[mkelley]

Sounds good to me.

[bsipocz]

We have to be mindful that users sometimes run software on systems that are out of their control and a Python upgrade might require an OS upgrade with sys admin intervention and other complications.

Yes, that's why I'm not suggesting to follow drop schedule the big libraries follow and endorse. But at some point we need to start dropping as the infrastructure we're using is also slowly doing it. (nothing extra urgent just yet)

[jaymedina]

This seems reasonable. We've already dropped support for 3.7 in Astrocut too (which tesscut/zcut/hapcut wrap around).

[jespinosaar]

Hi all, I think we can drop the support, yes, I have been testing with our modules and Python >= 3.8 and seems to be working.

[bsipocz]

@jespinosaar - it's less about testing, more about whether you know your end users being stuck on an old version?