Sig, Fair enough. I find 1..3 to unstable, and the thrashing sounds about right. For now I am trying to carve a home in 1.4.
Bill ________________________________________ From: [email protected] [[email protected]] on behalf of Igor Stasenko [[email protected]] Sent: Friday, February 24, 2012 1:37 PM To: [email protected] Subject: Re: [Pharo-project] [Bug Tracker] Issues tagged 1.3 On 24 February 2012 16:11, Schwab,Wilhelm K <[email protected]> wrote: > Guys, > > I'm hoping Sig will see this an immediately know what is at play. I don't > even know if there is a fix, but there appears to be in 1.4. If backporting > that to 1.3 is feasible, it should be done. I *think* I know some keywords > that might pull his relevant reply (something like "you are a victim of") > from the archive. Can't do it right now. > i remember it. And no, i don't see why these changes are needed to be backported to 1.3. It was : http://code.google.com/p/pharo/issues/detail?id=5167&q=finalization&colspec=ID%20Type%20Status%20Summary%20Milestone%20Difficulty and fix which were integrated was about improving a situation with runaway processes and interrupt handling. Now, that bug entry is still not closed, becaue we have to deal with so-called "finalization trashing", whether it is real issue or not. As to me, this is not an issue, because finalization "trashing" is a consequence but not the cause. And since it is better to fix the cause instead, that what i actually did. And then during integration i made a mistake in one method (an overlooking), causing debugger to not open, but it is already reverted. All this is related to 1.4 bleeding edge, not to 1.3. And i personally don't see good reason porting it to 1.3. It is not critical. We just improved this part a bit. So, the new pharo version is naturally better than older one.. that's why we got versions and development process. :) If we would start backporting everything, then we don't need versions anymore. We then will just have single image and fight with all dragons at once. > BTW, I took Stef's reply and used it to good effect - might have even > figured out why I couldn't build images. All part of testing. If you look, > you'll find code of mine in the image. Have you noted some of my responses > to help others - probably not. But I would rather not play "I haven't seen" > games. > > Bill > > ________________________________________ > From: [email protected] > [[email protected]] on behalf of Sven Van > Caekenberghe [[email protected]] > Sent: Friday, February 24, 2012 6:02 AM > To: [email protected] > Subject: Re: [Pharo-project] [Bug Tracker] Issues tagged 1.3 > > On 24 Feb 2012, at 11:12, Schwab,Wilhelm K wrote: > >> There is a tracker entry, but please understand it's the middle of the night >> here and I'm in and out. I think Sig will know the details more off the >> top, otherwise I'll search later. Would you rather I say nothing about a >> known bug? > > I can't find anything with Schwab, Bill or Wilhelm in any issue ever in > http://code.google.com/p/pharo/issues/list (there are a handful that are more > than half a year old, magically you even received help). I read this mailing > list all the time, I can't remember seeing a concrete, reproduceable bug > report. Please, correct me if I am wrong. > > Some, like Stef, even took the trouble of trying to load some of the stuff > that was causing you trouble, only to report that it worked for them. > > I and many others ask for help and report problems here all the time. It is > incredible how people spring up all the time to help others with both simple > and deep problems. It is often so that the one reporting a problem did > something wrong himself, I am guilty here as well, but even then people took > the trouble to look and give some of their valuable time. > > The way you ask the question makes a huge difference: provide details, make > it reproduceable, don't include negative stuff. > > I'll make you a promise: I'll give you some of my personal time to try > whatever you report next as 'BAD bug, makes 1.3 unstable, rendering the image > useless', provided there is a description of how I should do it. > > Sven > > > > > > -- Best regards, Igor Stasenko.
