10:39 <kusma> zmike: Looks like the wide-lines stuff is technically speaking dEQP bugs, and the equations we use should be allowed.

10:39 <kusma> I'm splitting out that patch into its own MR, because it's a bit more involved, it seems.

10:39 <zmike> huh

10:39 <zmike> neat

10:40 <kusma> I'm actually wondering if the reason we fail tests is that we support unusually large lines, and this way of interpolating along the lines gets very strange with extremely wide lines.

10:40 <kusma> unusually wide lines, I mean.

10:43 <zmike> easy enough to test

10:44 <kusma> Nah, seems not to be it :/

10:45 <kusma> The CTS seems to only test the "preferred" interpolation, but not the alternatively allowed one. I was kinda hoping that the difference wouldn't be too large if the lines aren't too wide, but no...

10:52 <zmike> conformance just keeps getting further away

10:54 <kusma> https://i.giphy.com/media/3ohhwontcexZ7expok/source.gif

10:56 <zmike> I always think of zink as being this guy https://gfycat.com/courageousdangerousbagworm#

10:57 <kusma> 😆

11:58 <daniels> looooll

20:31 <zmike> hch12907: 🙏🙏🙏

20:34 <hch12907> oo boy, it was not fun to maintain a patchset that touches almost every single file.

20:34 <hch12907> rebase conflicts every other day!

20:34 <zmike> willywonka.meme

20:35 <zmike> the key is you just rebase at the end before merge

20:35 <zmike> damn first try

20:35 <zmike> I'm on fire today

20:40 <daniels> who's on fire? :P

20:40 <zmike> the guy clicking the button obviously

21:08 <zmike> kusma: can you think of a good reason not to clamp PIPE_CAP_MAX_VIEWPORTS to 16?

21:08 <zmike> this is the required number for extensions

21:10 <zmike> (also anyone else who has strong feelings about it)

22:21 <daniels> zmike: can you please hold reassigning your MR (the fdno cache server is back now) until !11309 lands?

22:30 <zmike> I'm clocked out