Re: Fixing regressions and serious issues
On Fri, 2022-09-16 at 13:36 +0200, Jonas Hahnfeld via Discussions on LilyPond development wrote: > On Wed, 2022-09-14 at 22:35 +0200, Jean Abou Samra wrote: > > Le 14/09/2022 à 22:16, Jonas Hahnfeld a écrit : > > > On Wed, 2022-07-20 at 11:39 +0200, Jonas Hahnfeld via Discussions on > > > LilyPond development wrote: > > > What do we do about this one? Over the past couple of weeks, I tried > > > quite a number of ideas, with no success so far. > > > > Thanks a lot for working on this even if it didn't succeed so far. > > > > Just in case for others: Jonas shared some details about what he tried > > in https://github.com/ivmai/bdwgc/issues/454#issuecomment-1244375504 > > I've tried some more and probably developed an understanding of what's > happening - I will post on the upstream issue later. For our use case, > however, we can "cheat" a bit because we statically link both bdwgc and > libguile; https://gitlab.com/lilypond/lilypond/-/merge_requests/1627 > should work around the crashes (at least they do for me in Wine). Jean thankfully tested the binaries I produced with that change, and they don't seem to completely break while fixing the crashes with the originally reported scores. Note that there are apparently still crashes with debugging scores that Jean produced, calling (gc) on every translator slot... With the fix of the user scores merged though, I've now started to produce 2.23.13, I'll announce it once ready later today. What are the updated feelings on the plan of branching next week? If we agree to stick to that plan, I will bump master to 2.23.80 in preparation of the first release candidate. If not, I'll go with 2.23.14 with the possibility that this version will never see the light of the world... Jonas signature.asc Description: This is a digitally signed message part
Re: Fixing regressions and serious issues
Le 18/09/2022 à 14:46, Jonas Hahnfeld a écrit : On Fri, 2022-09-16 at 13:36 +0200, Jonas Hahnfeld via Discussions on LilyPond development wrote: On Wed, 2022-09-14 at 22:35 +0200, Jean Abou Samra wrote: Le 14/09/2022 à 22:16, Jonas Hahnfeld a écrit : On Wed, 2022-07-20 at 11:39 +0200, Jonas Hahnfeld via Discussions on LilyPond development wrote: What do we do about this one? Over the past couple of weeks, I tried quite a number of ideas, with no success so far. Thanks a lot for working on this even if it didn't succeed so far. Just in case for others: Jonas shared some details about what he tried in https://github.com/ivmai/bdwgc/issues/454#issuecomment-1244375504 I've tried some more and probably developed an understanding of what's happening - I will post on the upstream issue later. For our use case, however, we can "cheat" a bit because we statically link both bdwgc and libguile; https://gitlab.com/lilypond/lilypond/-/merge_requests/1627 should work around the crashes (at least they do for me in Wine). Jean thankfully tested the binaries I produced with that change, and they don't seem to completely break while fixing the crashes with the originally reported scores. I'm sorry for not having been clear on this point: yes, I haven't observed crashes on the original reports; on the other hand, I tried 2.23.10 (supposed to be buggy), and didn't get crashes either, so I dunno :( This is the glory of Heisenbugs. The MR at least fixed the crashes you observed with Wine, so hopefully it should at least reduce the frequency of crashes. At this point, there seems to be a good chance that we are dealing with a bug in BDWGC itself, so I think we can call it good enough if it doesn't pop up for real-world use cases on our end, and allow the 2.24 in that case, even if the underlying issue in BDWGC is not fully understood. The conclusion is that I am happy this change was included in 2.23.13, as we will be able to ask Windows users if these binaries seem to fix the issue for them. By the way, if you happen to still have a binary of what is now 2.23.13 with debugging symbols, that would save me time. I'd like to investigate the crash when loading the edition-engraver, which is 100% reproducible for me, but at a glance doesn't seem related to this GC Heisenbug.
Re: Fixing regressions and serious issues
On Sun, 2022-09-18 at 18:17 +0200, Jean Abou Samra wrote: > By the way, if you happen to still have a binary of what is now > 2.23.13 with debugging symbols, that would save me time. Yes, I always try to keep around the latest build, for accessing logs and debug symbols. I uploaded the lilypond.exe before stripping here: https://cloud.hahnjo.de/index.php/s/ZB8mG82TNpLwi3b Let me know if you also need the Dlls. > I'd like to investigate the crash when loading the edition-engraver, > which is 100% reproducible for me, but at a glance doesn't seem related > to this GC Heisenbug. I saw your comment, but you need an account to even read the openlilylib forum, so I didn't bother... signature.asc Description: This is a digitally signed message part
Re: Fixing regressions and serious issues
Hi Jonas, I run the OpenLilyLib Discourse forum and have taken over from Urs managing OLL (where more work will be happening soon!). Is it too onerous to register with a username and password? It's totally free. One cannot make forums totally open because of spammers, although I may consider making it readable to the public, but given it's a group of about 20 so far there does not seem much point. If you register, your personal data is kept totally private and not disseminated or used in any way apart from providing a login to the service. You are very welcome to join. Andrew On 19/09/2022 2:35 am, Jonas Hahnfeld via Discussions on LilyPond development wrote: I saw your comment, but you need an account to even read the openlilylib forum, so I didn't bother...
LilyPond 2.23.13
We are happy to announce the release of LilyPond 2.23.13. This is termed a development release, but these are usually reliable. However, if you require stability, we recommend using version 2.22.2, the current stable release. Please refer to the Installing section in the Learning Manual for instructions how to set up the provided binaries: https://lilypond.org/doc/v2.23/Documentation/learning/installing signature.asc Description: This is a digitally signed message part
Re: Fixing regressions and serious issues
Le 18/09/2022 à 18:45, Andrew Bernard a écrit : Hi Jonas, I run the OpenLilyLib Discourse forum and have taken over from Urs managing OLL (where more work will be happening soon!). Interesting, would you mind sharing about the work you plan? (Preferably in a separate thread) Is it too onerous to register with a username and password? It's totally free. One cannot make forums totally open because of spammers, although I may consider making it readable to the public, but given it's a group of about 20 so far there does not seem much point. Requiring an account to post is understandable, but I'd make any forum public for reading by default unless there are reasons not to. Subscribing to read posts is not a big barrier, but people are driven by curiosity, and if you need to take administrative steps before reading a post it can turn you away. If you register, your personal data is kept totally private and not disseminated or used in any way apart from providing a login to the service. You are very welcome to join.
Re: Fixing regressions and serious issues
On 18/09/2022 20:38, Jean Abou Samra wrote: Requiring an account to post is understandable, but I'd make any forum public for reading by default unless there are reasons not to. Subscribing to read posts is not a big barrier, but people are driven by curiosity, and if you need to take administrative steps before reading a post it can turn you away. From my point of view, the FEWER accounts I have the happier I am. Having to create an account is a massive turn-off that, if I don't have a burning desire for what's on offer, I just won't bother. (And what happens if I create an account, forget all about it, and then try and create a new account six months later? The friction is greatly increased because I can't create a new account, but can't get into the old account either ...) The other thing is, how much personal information does it want? The best internet security is just not to be on the net - if a site does not have your information, it can't lose it! Even when I *want* something, I regularly abandon any attempt half-way through because I don't want to provide the information they're demanding. Cheers, Wol
Re: LilyPond 2.23.13
Le 18/09/2022 à 18:49, Jonas Hahnfeld via LilyPond user discussion a écrit : We are happy to announce the release of LilyPond 2.23.13. This is termed a development release, but these are usually reliable. However, if you require stability, we recommend using version 2.22.2, the current stable release. Please refer to the Installing section in the Learning Manual for instructions how to set up the provided binaries: https://lilypond.org/doc/v2.23/Documentation/learning/installing A little call for testing: There has been a bug for a while in the 2.23 series, happening only on Windows, which would result in sporadic crashes with return code -1073741819. See https://gitlab.com/lilypond/lilypond/-/issues/6361 if you are interested in further information. In this release, the frequency at which the bug occurs seems much reduced, although it can still be triggered under very specific circumstances with Scheme code if you really try hard. If you are on Windows, we encourage you to test this release on any scores you might have, especially larger scores. If you encounter any trouble, please get back to us on the bug-lilypond list (bug-lilyp...@gnu.org). If this bug still occurs under real-world conditions, we would like to know it. Thank you.
Re: Fixing regressions and serious issues
Gosh, to both. All are welcome to the OLL forum should they so feel like it. One reason not to open a forum is that there are people who may not want their postings disseminated and indexed, for whatever reason, even though this instance is purely a technical discussion platform. If you want to contribute to the OLL development then I'm afraid you need to sign up for a github account. Andrew On 19/09/2022 6:41 am, anthony wrote: On 18/09/2022 20:38, Jean Abou Samra wrote: Requiring an account to post is understandable, but I'd make any forum public for reading by default unless there are reasons not to. Subscribing to read posts is not a big barrier, but people are driven by curiosity, and if you need to take administrative steps before reading a post it can turn you away. From my point of view, the FEWER accounts I have the happier I am. Having to create an account is a massive turn-off that, if I don't have a burning desire for what's on offer, I just won't bother. (And what happens if I create an account, forget all about it, and then try and create a new account six months later? The friction is greatly increased because I can't create a new account, but can't get into the old account either ...) The other thing is, how much personal information does it want? The best internet security is just not to be on the net - if a site does not have your information, it can't lose it! Even when I *want* something, I regularly abandon any attempt half-way through because I don't want to provide the information they're demanding. Cheers, Wol
