On Wed, May 04, 2016 at 11:19:56PM +0100, Guillaume Munch wrote: > Le 04/05/2016 21:39, Scott Kostyshak a écrit : > > > > 4. Insert Cross-Reference Causes Crash. See #9992. Unfortunately, no > > one can reproduce except the OP, Joel. Joel has been extremely > > helpful in debugging and is able to compile LyX but only with Qt > > 5.5.1. Unless we make further progress in debugging, we will either > > do nothing, or consider Guillaume's proposal to disable > > QT_NO_ACCESSIBILITY during compilation. Disabling that might actually > > be good for other reasons (haven't we seen a few bugs related to > > it?). I don't know what the consequences of that is though. > > This was not a proposal, just a summary of what I read. This is a flag > in the qt compilation and we do not compile our own qt, do we?
No we do not. I was thinking that a possibility would be to set the environment variable inside LyX. > In addition there are indeed bugs related to accessibility, but enabling > accessibility is a pre-requisite for recording progress in this area. Good point. Let's do nothing on #9992 unless we can make further progress debugging then. > > 6. Icon set corrupted when moving window from a HiDPI screen to a > > low DPI screen. See #10113. I'm planning to remove the 2.2.0 > > milestone. I don't think we should even attempt a fix for this > > situation at this point in the release phase. > > Yes, the milestone was set by the user. OK done. > > 9. Disabled OK button in Doc Settings if negative value (allowed in > > 2.1.x). See #10095 and [3]. I posted a patch. It is just a matter of > > seeing if there is enough support for it. > > > > I thought there was already enough support. But for what it is worth I > am in favour. Done (I committed). > > What am I missing? > > > > #10068: Files distributed with lyx should not contain parbreak and latexpar > http://mid.gmane.org/nfu741$c3j$1...@ger.gmane.org > > Regarding inclusion in 2.2.0: I did not insist on it, but in my opinion > this is welcome and low-risk. (I would not suggest the same for the > documentation, which is going to require more work checking the > converted files.) > > More importantly, even if this is not interesting or funny, I would like > to have feedback about this soon, whatever the branches, because if > somebody changes a file in the meanwhile, then somebody could have to > re-do the conversion. OK go ahead for master. Scott
signature.asc
Description: PGP signature
