On Fri, 2018-03-16 at 13:51 +0100, Kalev Lember wrote: > On 03/16/2018 01:08 PM, Adam Williamson wrote: > > On Fri, 2018-03-16 at 11:52 +0100, Kalev Lember wrote: > > > > > > I'd appreciate a lot if people could give it a quick spin and see if > > > they can find anything that looks like a regression compared to the > > > stable F28 images, and then we can further discuss this at the blocker > > > review meeting on Monday. > > > > The new version of Cantarell is, by its own authors admission, a > > regression for all practical purposes: > > > > <madigens[m]> adamw: thanks for the feedback :) cantarell's spacing > > does indeed look a bit funky... in gtk3 and firefox. that's due to a > > missing technical detail called "subpixel positioning". Spacing looks > > more consistent in Qt apps and Chrome. there also isn't any kerning > > yet, so things like AVA, To, etc. look gap-y > > > > So, it looks fine...except in GTK+ 3 (i.e. basically all of > > Workstation) and Firefox (Workstation's default browser). > > > > That seems like a problem. :P (Never mind that it just chucks out > > Cyrillic coverage, which the previous version had). Can we consider > > sticking with the old one at least until it's clear if this will be > > cleaned up in time for Final? > > Fair enough, that's probably a prudent thing to do. I've split it out > from the rest of the 3.28.0 update now: > > https://bodhi.fedoraproject.org/updates/abattis-cantarell-fonts-0.101-1.fc28 Thanks. I mean, I'd suggest not sending it to F28 at all (i.e. unpushing that update; keeping the update around will probably mean it gets karma and then gets auto-pushed when the Beta freeze is lifted) till madigens has had time to clean things up; it clearly looks worse than 0.0.25 for now, so why not just keep that one in F28 until the new series is at least roughly on a level with it? It's not like pulling it in is really *dangerous*, it's just slightly inconvenient to openQA, so if it turns out that it *does* get improved in time for Final, we can always just pull it in then. We can keep the new series in Rawhide, of course, seems reasonable to assume madigens will have got it looking nicer by the time we get to F29. Anyway, that's just my vote. Here's some of the links I posted to openQA screenshots that kinda flag up the issues in 101: https://openqa.fedoraproject.org/tests/204833#step/disk_custom_ext3/11 https://openqa.stg.fedoraproject.org/tests/255085#step/disk_custom_ext3/11 (set the "Candidate needle:" dropdown to None to see the clean screenshot) The former is with Cantarell 0.0.25, the latter with 0.101. In some ways 101 is a bit of an improvement already ('AVAILABLE SPACE' and 'TOTAL SPACE' look a bit better to me, for e.g.) but overall it seems clearly worse. The spacing between glyphs in words is pretty inconsistent...you can see it clearly in text like "BIOS Boot", "Encrypt" and "Modify". The spacing between words is often not much bigger than the spacing between glyphs in words, which has the effect of making the words seem to run together - this is clearest in "New Fedora Rawhide Installation" and the block of italic text at bottom right. Another example I found this morning is: https://openqa.stg.fedoraproject.org/tests/255343#step/_graphical_wait_login/10 Especially the text "Connect Your Online Accounts" has all kinds of spacing issues there, there appears to be almost no space between "Online" and "Accounts" in particular, whereas there's a lot of space between the "Y" and "o" of "Your", and between the "o", "n", "n" and "e" in "Connect", for instance. Here's how that screen looks with 0.0.25: https://openqa.fedoraproject.org/tests/202838#step/_graphical_wait_login/9 -- Adam Williamson Fedora QA Community Monkey IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net http://www.happyassassin.net _______________________________________________ test mailing list -- test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to test-leave@xxxxxxxxxxxxxxxxxxxxxxx
