Thanks Peter for your email. Alexandre
> On Dec 18, 2016, at 10:10 PM, Peter Uhnak <i.uh...@gmail.com> wrote: > > Hi Alex, > > I certainly understand your frustration, I felt it too on Windows to the > point where I stopped using Pharo for couple of weeks out of rage, and then > spend in total at least 40+ hours digging into the VM until I added usable > workaround for Windows. Not to mention how frustrating was the fixing it. > > Obviously on Mac there are still some unresolved pathways, but it will not > magically fix itself. > > The reason for the crash (bad object pinning and moving of canvas memory) has > been known for some time now, so more crash.dmps do not give any more insight. > > If this is to be resolved then one of the two have to happen: > > A) Someone who really understands VM/image memory management / GC / pinning > fixes the issue. > > B) Someone with Mac (which I don't have) digs around BitBlt code (or wherever > it was) and adds a similar workaround. > > Considering the issue has emerged more than a year ago (Spur switch), I don't > think (A) is going to happen any time soon, so I guess the only chance is to > get elbows greasy and fix it yourself (B) (or you make one of your students > suffer :)). > > I didn't have a single Roassal/BitBlt related crash on Windows since my fix > was added, so there should be a way to add a workaround for Mac too. > > (there are of course crashes related to FT, but the story is arguably the > same). > > TL;DR: I don't think it's on a todo list of anyone who actually understands > this, so the only way to fix it is by yourself or find a concrete person that > would be willing to dig into it. > > Peter > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
