3.3.0 Release July 1, 2016

[wiredfool]

Tracking issue.

[radarhere]

So, any remaining priorities for this release?

[wiredfool]

There's been a lot of C level changes in this iteration.

I really want to land the malloc check: pr #1781 .

#1934 is something that I think should land, but I'd like someone else to look at it and agree to the approach.

Something of the palette issue (#1593) should land, I think I'm leaning to the bytearray approach, but comments either way would be useful.

If I have time, I'd like to sweep though all of the numpy/fromarray issues and consolidate/resolve them. The ICC profile stuff is in that category too.

[homm]

From my PR's I want to see in next release at least #1909 and #1970. Anything else is less important. Also, I want #1910 to be addressed, but I don't have a fix.

[wiredfool]

I think we've hit all that.

@homm, what are your thoughts on bugfixes in #1986 and #1991?

The only other one that I'm considering at this point is #1846, which is just waiting on documentation. It's low risk, as it just adds behavior that was previously broken.

Apart from that, freeze except for docs.

[aclark4life]

Thanks all for working on this release, let me know if you need anything! <3

[wiredfool]

Release is tagged and source uploaded.

[wiredfool]

@cgohlke Would appreciate windows binaries when you can.

[wiredfool]

So, postmortem of sorts for this release. This feels like a big release, especially in terms of landing things in C land. Personally, I'm time constrained on reviews especially, so there are still a lot of PRs left.

As I see it, PRs fall into several categories;

• Clearly no.
• Eh, Suspicious without further input
• Those needing tests or documentation
• Those that are clearly 0 effect, and are for style.
• Those that are possibly 0 effect, but take a close reading to make sure.
• Bug fixes with tests
• Small extensions of the api
• Significant, but low risk.
• Significant, high risk
• Those that resemble flies from a distance

I think we've gotten most of the big ones into this release. There's a host in the needing more or possibly 0 effect categories.

[aclark4life]

@cgohlke Probably time, though I'd hate to do that …

[wiredfool]

There are a lot of files. I'm not clear on how many people would likely be impacted.