Re: spacing/breaking issue [time-sensitive]
On Mon, 2010-05-31 at 11:11 -0400, Kieren MacMillan wrote: [Lilypond 2.13.21] Hey all, I'm trying to get a score out this afternoon (for a competition deadline tomorrow!) and I'm running into a bug... Sorry, I'm on holiday and I haven't been checking email regularly, so this probably isn't much use any more, but anyway... My score looks quite nice on 17 systems spread over 5 pages -- I would like: Page 1: title + 3 Pages 2-4: 4 systems Page 5: 2 systems Unfortunately, minimal-breaking causes Lilypond to [try to] cram everything on one page, whereas optimal-breaking gives 3 systems for most pages and 4 on the last (even though ragged-last-bottom = ##t). No doubt one or more of my problems is related to the known bug regarding the missing properties blank-page-force = 0 blank-after-score-page-force = 0 but I've explicitly set that as well. Which known bug is this? I searched the bug tracker for blank and blank-page-force with no results. Cheers, Joe ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Joe, No doubt one or more of my problems is related to the known bug regarding the missing properties blank-page-force = 0 blank-after-score-page-force = 0 but I've explicitly set that as well. Which known bug is this? I searched the bug tracker for blank and blank-page-force with no results. http://lists.gnu.org/archive/html/lilypond-user/2010-01/msg00086.html Cheers, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On Mon, 2010-06-07 at 13:00 -0400, Kieren MacMillan wrote: Hi Joe, No doubt one or more of my problems is related to the known bug regarding the missing properties blank-page-force = 0 blank-after-score-page-force = 0 but I've explicitly set that as well. Which known bug is this? I searched the bug tracker for blank and blank-page-force with no results. http://lists.gnu.org/archive/html/lilypond-user/2010-01/msg00086.html Oops, I already fixed this one but it seems that I sent the email saying so only to Reinhold and not to the list. Cheers, Joe ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On Mon, May 31, 2010 at 4:11 PM, Kieren MacMillan kieren_macmil...@sympatico.ca wrote: My score looks quite nice on 17 systems spread over 5 pages -- I would like: Page 1: title + 3 Pages 2-4: 4 systems Page 5: 2 systems Other than manually inserting breaks, what can I do to get Lilypond to Do The Right Thing™? I think that manual breaks are the easiest solution. You could separate it into distinct \bookparts, but that would be much more work than manually adding \break and \pageBreak. Cheers, - Graham ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Graham, I think that manual breaks are the easiest solution. You could separate it into distinct \bookparts, but that would be much more work than manually adding \break and \pageBreak. OK... =\ No disrespect meant to anyone, in particular Joe who has done a fabulous job increasing the flexibility and scope of the spacing code… but: MAO MAO-ING MAO, can we please resolve these spacing issues ASAP? Are we really comfortable telling people -- especially newbies -- that, in order to have Lilypond output good-looking scores, they have to: 1. manually insert breaks in the music (ruining the separation of content from presentation, unless they use \tag which is stupid MAO-ing overkill); or, 2. manually insert breaks in some score-specific variable with skips that they have to count and keep synchronized with the music? Oh, and p.s. Noob: manually inserting \pageBreaks does not, by itself, guarantee that Lilypond will actually break pages *only* there: you also have to mess around with the mao-ing #'page-break-permission, and maybe something else we've forgotten to warn you about. Sorry for the rant -- I blame the deadline pressure -- but it really is a bit ludicrous, and does not give Lilypond the best possible chance of impressing anyone out of the box. If this issue isn't on your critical 2.14-killer list, it should be. Cheers, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Graham, For the record, I just added \pageBreaks where I want them... and Lilypond outputted 25 pages: it added 20 blank pages and scattered them amongst the 5 with actual music. MAO!!! =( Kieren. On 2010-May-31, at 11:36, Kieren MacMillan wrote: Hi Graham, I think that manual breaks are the easiest solution. You could separate it into distinct \bookparts, but that would be much more work than manually adding \break and \pageBreak. OK... =\ No disrespect meant to anyone, in particular Joe who has done a fabulous job increasing the flexibility and scope of the spacing code… but: MAO MAO-ING MAO, can we please resolve these spacing issues ASAP? Are we really comfortable telling people -- especially newbies -- that, in order to have Lilypond output good-looking scores, they have to: 1. manually insert breaks in the music (ruining the separation of content from presentation, unless they use \tag which is stupid MAO-ing overkill); or, 2. manually insert breaks in some score-specific variable with skips that they have to count and keep synchronized with the music? Oh, and p.s. Noob: manually inserting \pageBreaks does not, by itself, guarantee that Lilypond will actually break pages *only* there: you also have to mess around with the mao-ing #'page-break-permission, and maybe something else we've forgotten to warn you about. Sorry for the rant -- I blame the deadline pressure -- but it really is a bit ludicrous, and does not give Lilypond the best possible chance of impressing anyone out of the box. If this issue isn't on your critical 2.14-killer list, it should be. Cheers, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On Mon, May 31, 2010 at 4:41 PM, Kieren MacMillan kieren_macmil...@sympatico.ca wrote: For the record, I just added \pageBreaks where I want them... and Lilypond outputted 25 pages: it added 20 blank pages and scattered them amongst the 5 with actual music. MAO!!! =( Stick the \pageBreak in the 1st violin part, not the global{} part? I'm guessing here, but maybe two simultaneous \pageBreaks aren't treated/recognized as simultaneous? Cheers, - Graham ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On 31 May 2010 16:11, Kieren MacMillan kieren_macmil...@sympatico.ca wrote: Unfortunately, minimal-breaking causes Lilypond to [try to] cram everything on one page You have to use manual breaks with minimal-breaking (though I think a bug's crept in too since 2.13.19: both regression tests for minimal-breaking have started complaining). No doubt one or more of my problems is related to the known bug regarding the missing properties blank-page-force = 0 blank-after-score-page-force = 0 but I've explicitly set that as well. Which bug's that? Cheers, Neil ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On Mon, May 31, 2010 at 4:36 PM, Kieren MacMillan kieren_macmil...@sympatico.ca wrote: MAO MAO-ING MAO, can we please resolve these spacing issues ASAP? At the risk of making a crude masturbation joke, if I could wave a magic wand to fix stuff, I'd be waving my wand all day and night. Sorry for the rant -- I blame the deadline pressure -- but it really is a bit ludicrous, and does not give Lilypond the best possible chance of impressing anyone out of the box. If this issue isn't on your critical 2.14-killer list, it should be. For the record, there are NO KNOWN BUGS with the new spacing algorithm. Every single spacing regression bug that has been sent to bug-lilypond has been fixed. ** nb: actually, this might be incorrect. There's about 5-10 bugs that were sent to bug-lilypond, but for which nobody has processed them. One of those might contain a nicely-written bug report with a Tiny example which demonstrates this problem. So rather: every single spacing regression bug that was entered to the issue tracker has been fixed. Because of our recent discussion about contributors getting started, I've spent an hour rewriting the CG chapter about Issues and the website help us page. Hopefully in a few weeks I can start recruiting new people to join the Bug Squad -- perhaps even you? this requires absolutely zero interaction with the source code and compilation, after all -- so that we won't lose good bug reports. (or even non-good bug reports) Hopefully after the deadline is over, you could create a tiny example of your problem and report it as a bug. Or, since Joe doesn't seem to be very picky about spacing bugs, and is probably willing to assure you that he will not show copywritten material to anybody else, you might be able to send your current score directly to him. But I want to be absolutely clear that I can do is organize the bug reports and attempt to herd the cattish developers. If I have no reports (again, leaving aside the lost ones), I can't try to manipulate any developers into working on those problems. (with regards to the lost bug reports: again, all I can do is to organize people willing to work on bugs. If we have so few volunteers offering to help organize bugs that we can't process all bug reports, then we'll obviously lose bug reports. *shrug* ) Cheers, - Graham ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Graham, Stick the \pageBreak in the 1st violin part, not the global{} part? I'm guessing here, but maybe two simultaneous \pageBreaks aren't treated/recognized as simultaneous? Thanks for the guess... but I *do* have it in only one part (by creating a separate, score-specific variable, and then -ing it in only with the violin staff). I've got it working now, though I don't really know how: I've changed a bunch of properties/settings in a bunch of different files (which is likely to MAO up my other scores). I'll do a post-mortem on this in the next few days -- right now, I've got to get this printed up and couriered off. Thanks, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Neil, Unfortunately, minimal-breaking causes Lilypond to [try to] cram everything on one page You have to use manual breaks with minimal-breaking (though I think a bug's crept in too since 2.13.19: both regression tests for minimal-breaking have started complaining). Wait... one HAS to use \pageBreak if you set minimal-breaking? Does that really make sense? No doubt one or more of my problems is related to the known bug regarding the missing properties blank-page-force = 0 blank-after-score-page-force = 0 but I've explicitly set that as well. Which bug's that? See the thread including http://lists.gnu.org/archive/html/lilypond-user/2010-01/msg00086.html Cheers, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Graham, At the risk of making a crude masturbation joke, if I could wave a magic wand to fix stuff, I'd be waving my wand all day and night. =) For the record, there are NO KNOWN BUGS with the new spacing algorithm. So http://lists.gnu.org/archive/html/lilypond-user/2010-01/msg00086.html is fixed? Hopefully in a few weeks I can start recruiting new people to join the Bug Squad -- perhaps even you? Definitely. Hopefully after the deadline is over, you could create a tiny example of your problem and report it as a bug. Or, since Joe doesn't seem to be very picky about spacing bugs, and is probably willing to assure you that he will not show copywritten material to anybody else, you might be able to send your current score directly to him. I have no problem re: copyright. My problem is that the source is (a) currently scattered over a number of \include files, and I need to optimize. Thanks for your contributions, as always! Kieren Defcon Down To Two Now That The Score Is In Printable Condition MacMillan. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On 31 May 2010 17:07, Kieren MacMillan kieren_macmil...@sympatico.ca wrote: Wait... one HAS to use \pageBreak if you set minimal-breaking? According to the texidocs for the minimal-breaking regression tests, yes. Does that really make sense? It doesn't seem to offer anything useful over inserting manual breaks while forbidding automatic breaking, but I'm sure there's more to it than this. Cheers, Neil ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Neil, Wait... one HAS to use \pageBreak if you set minimal-breaking? According to the texidocs for the minimal-breaking regression tests, yes. According to Notation, The ly:minimal-breaking function performs minimal computations to calculate the page breaking: it fills a page with as many systems as possible before moving to the next one. 1. No mention of \pageBreak (which is introduced in the next section) being required. 2. This description is clearly false, at least in the case of my score: it was saving 4 systems for the last page, where it looked just fine with 2 (and the other two on previous pages). It doesn't seem to offer anything useful over inserting manual breaks while forbidding automatic breaking, but I'm sure there's more to it than this. Probably -- I'll look into it during my post-mortem. At the very least, it should do what it says in Notation. [Or Notation should be rewritten.] Cheers, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On 2010-05-31 17:41, Kieren MacMillan wrote: Hi Graham, For the record, I just added \pageBreaks where I want them... and Lilypond outputted 25 pages: it added 20 blank pages and scattered them amongst the 5 with actual music. Hi, Kieren, I assume you already defined page-count = #5? There's a known issue (at least it's been on the mailing lists, no clue whether it's in the bug tracker and/or some known issues section in the docs) as follows: If page-count is defined in the 2.13 series (at least the earlier releases), \pageBreak produces page-count many breaks for the whole score; thus, you have to either set page-count = #1 (sic!) and use manual page breaks, or use \override NonMusicalPaperColumn #'page-break-permission = ##f and omit the page-count definition. By the way, I have no clue whether the same holds for system-count and 'line-break-permission. IMHO, this could also be show stopper for 2.14... HTH, and good luck with the deadline, Alexander ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On 2010-05-31 19:14, Alexander Kobel wrote: On 2010-05-31 17:41, Kieren MacMillan wrote: Hi Graham, For the record, I just added \pageBreaks where I want them... and Lilypond outputted 25 pages: it added 20 blank pages and scattered them amongst the 5 with actual music. Hi, Kieren, I assume you already defined page-count = #5? There's a known issue [...] Darn. Note to myself: Look at the tracker before posting. http://code.google.com/p/lilypond/issues/detail?id=884 This is it, and once more my explanation is not really true. But at least the workaround should work. :-/ Cheers, Alexander ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Le 31 mai 2010 à 17:57, Neil Puttock a écrit : On 31 May 2010 16:11, Kieren MacMillan kieren_macmil...@sympatico.ca wrote: Unfortunately, minimal-breaking causes Lilypond to [try to] cram everything on one page You have to use manual breaks with minimal-breaking (though I think a bug's crept in too since 2.13.19: both regression tests for minimal-breaking have started complaining). I jump into this conversation. Facing the same bug myself on a personal project, I am investigating it. Nicolas ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On Mon, May 31, 2010 at 07:14:30PM +0200, Alexander Kobel wrote: I assume you already defined page-count = #5? There's a known issue (at least it's been on the mailing lists, no clue whether it's in the bug tracker and/or some known issues section in the docs) as follows: If page-count is defined in the 2.13 series (at least the earlier releases), \pageBreak produces page-count many breaks for the whole score; thus, you have to either set page-count = #1 (sic!) and use manual page breaks, or use \override NonMusicalPaperColumn #'page-break-permission = ##f and omit the page-count definition. By the way, I have no clue whether the same holds for system-count and 'line-break-permission. IMHO, this could also be show stopper for 2.14... Since none of those are in the tracker as a Critical issue, none of them are show-stoppers for 2.14. If we get proper bug reports about them, showing a file that worked in 2.10 or 2.12 but fails in 2.14, then it would become a show-stopper. Cheers, - Graham ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Le 31 mai 2010 à 17:11, Kieren MacMillan a écrit : [Lilypond 2.13.21] Hey all, I'm trying to get a score out this afternoon (for a competition deadline tomorrow!) and I'm running into a bug... My score looks quite nice on 17 systems spread over 5 pages -- I would like: Page 1: title + 3 Pages 2-4: 4 systems Page 5: 2 systems Unfortunately, minimal-breaking causes Lilypond to [try to] cram everything on one page, whereas optimal-breaking gives 3 systems for most pages and 4 on the last (even though ragged-last-bottom = ##t). No doubt one or more of my problems is related to the known bug regarding the missing properties blank-page-force = 0 blank-after-score-page-force = 0 but I've explicitly set that as well. Other than manually inserting breaks, what can I do to get Lilypond to Do The Right Thing™? If you compile LilyPond yourself, you might try the following patch: diff --git a/lily/page-breaking.cc b/lily/page-breaking.cc index 0ef1a30..2251e50 100644 --- a/lily/page-breaking.cc +++ b/lily/page-breaking.cc @@ -1137,6 +1137,7 @@ Page_breaking::pack_systems_on_least_pages (vsize configuration, vsize first_pag Page_spacing space (page_height (first_page_num, false), this); cache_line_details (configuration); + compute_line_heights (); for (vsize line = 0; line cached_line_details_.size (); line++) { Real prev_force = space.force_; ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On Mon, May 31, 2010 at 08:23:40PM +0200, Nicolas Sceaux wrote: If you compile LilyPond yourself, you might try the following patch: ~~~ ... + compute_line_heights (); ROTFLMAO! That email is epic p0wnage. Given Kieren's problems trying to compile lilypond... given his efforts to produce good scores so close to a deadline... given the simplicity of the patch... and given that I'm certain that Nicolas isn't trying to be at all mean and is genuinely trying to help... that was absolutely beautiful. The best unintentional snark I've seen in years. ;) Cheers, - Graham ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Graham, 1) I see no indication that it was ever added to the issue tracker; hence my no KNOWN bugs still qualifies. And I don't think that we've ever claimed that we'd acknowledge bugs sent to the user list. No, the fact that the bug discussion was between two main developers does not mean that the bug is known. OK... I guess I'm just surprised that Joe didn't submit it, since it's his code, he personally confirmed/acknowledged the bug, and he appears to be a pretty conscientious dev. Ok, I'll release 2.13.23 fairly soon, so that the updated bug guildelines are online. You'll need a day or two to recover from this deadline, anyway. :) At least. :) Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hey Graham, ROTFLMAO! So *that's* where mao comes from... ;) That email is epic p0wnage. ;-p Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On 31 May 2010 22:29, Kieren MacMillan kieren_macmil...@sympatico.ca wrote: OK... I guess I'm just surprised that Joe didn't submit it, since it's his code, he personally confirmed/acknowledged the bug, and he appears to be a pretty conscientious dev. Hmm, are you sure this bug still exists? I've just tested Reinhold's snippet from that thread, and it works fine in 2.13.23. Looking at the code, the method which fills in the penalty details, Page_breaking::blank_page_penalty (), is only used by the page-turn-breaking code (so it can't affect any other page-breaking algorithms). Cheers, Neil ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On 31 May 2010 22:52, Neil Puttock n.putt...@gmail.com wrote: Hmm, are you sure this bug still exists? I've just tested Reinhold's snippet from that thread, and it works fine in 2.13.23. Aha, found the fix: http://git.savannah.gnu.org/gitweb/?p=lilypond.git;a=commit;h=337f9c142a82cc9e844ea3ce4beacfc4c24d5993 (precedes 2.13.11) Looks like Joe was pretty conscientious here. ;) Cheers, Neil ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Hi Neil, Aha, found the fix: http://git.savannah.gnu.org/gitweb/?p=lilypond.git;a=commit;h=337f9c142a82cc9e844ea3ce4beacfc4c24d5993 (precedes 2.13.11) Looks like Joe was pretty conscientious here. ;) I knew my faith wasn't misplaced. =) Thanks, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Graham, Since none of those are in the tracker as a Critical issue, none of them are show-stoppers for 2.14. May I say that http://code.google.com/p/lilypond/issues/detail?id=884 being labelled as medium causes me to wonder about the whole bug prioritizing system. Is there some logic to what makes something critical? Cheers, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
On Mon, May 31, 2010 at 07:59:20PM -0400, Kieren MacMillan wrote: Graham, Since none of those are in the tracker as a Critical issue, none of them are show-stoppers for 2.14. May I say that http://code.google.com/p/lilypond/issues/detail?id=884 being labelled as medium causes me to wonder about the whole bug prioritizing system. Is there some logic to what makes something critical? A regression. If 884 created good output in 2.12, then by all means point this out and it'll be re-categorized. Note that even truly embarrassing mistakes are not Critical -- those are High. I wouldn't call 884 truly embarrassing, though. At least, not in comparison to various other longstanding bugs. (are you starting to understand why I'm so short-tempered when people talk about grand new features and the like?) Cheers, - Graham ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: spacing/breaking issue [time-sensitive]
Graham, Note that even truly embarrassing mistakes are not Critical -- those are High. I wouldn't call 884 truly embarrassing, though. But undocumented output suffixes is? At least, not in comparison to various other longstanding bugs. (are you starting to understand why I'm so short-tempered when people talk about grand new features and the like?) Because you're not short-tempered when people talk about anything else? ;-p Cheers, Kieren. ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user