Release v4.5.0

[dbarnett]

Process ticket for preparing 4.5 release.

[dbarnett]

I've finished nearly all the tasks I'd scoped to the 4.5 release and should be prepping a release within a week or so.

The last bits I'm focused on are:

• a few more robustness improvements for cal event operations and ics
• a few items I have tentatively in the list while they're being actively discussed/clarified but probably will end up pulling out of the release (initial auth setup, displaying week numbers)

and there were some agendaupdate improvements in flight that @michaelmhoffman mentioned on another issue, but I haven't gotten my head around current status there or figured out whether we should try to squeeze those changes into this 4.5 release.

[michaelmhoffman]

I landed them into master more 3–4 years ago in #553 and #577. So they're already included in the 4.4 release.

[dbarnett]

Alright, I think I've just landed the last of the improvements I was especially itching to get into the 4.5 release and can finish releasing that this week.

I cleaned up the agendaupdate stuff in the issue tracker and a bunch of other loose ends that smelled like they could've been low-hanging fruit to get into this release.

And I went ahead and created a 4.6 release milestone to roll some loose ends into as needed.

[dbarnett]

K, I've cleared out all the loose ends in the 4.5 milestone, forked & moved some leftovers, and I'm ready to prep a release.

FWIW, I won't be surprised if I made one or two trivial whoopsies in this release and will need a 4.5.1 right away after 4.5.0. 😅

[dbarnett]

It's live! https://pypi.org/project/gcalcli/4.5.0/

[dbarnett]

FYI I haven't seen anything outright broken yet but indeed hit a few rough spots already in this update:

1. The setuptools-scm mechanism tripped up AUR packaging, might trip up some other source-based distros (build/install fails from AUR/yay (setuptools-scm can't detect version from source tarball) #793) -- workaround available
2. Date parsing change went too far and assumes YYYY-DD-MM over YYYY-MM-DD on some locales (Datetime parsing shouldn't assume YYYY-DD-MM regardless of locale #792) -- fix coming soon

[dbarnett]

Fixed a handful of such easily-resolved issues for 4.5.1: https://github.com/insanum/gcalcli/milestone/12 (including that awful SSH max recursion error, hopefully!)

Planning to release that this week.

[dbarnett]

Tracking that in #804.