Builds of lilypond against texinfo 5.1 (which recently landed in unstable) have been failing, as detailed below. Could you please look into these errors? Thanks! ---------------------------------------------------------------------- Please check the logfile changes.makeinfo.log for errors: out/common-macros.texi:189: warning: @end ifset should only appear at a line beginning out/changes.texi:55: raising the section level of @section which is too low make[2]: *** [out/lilypond-changes.info] Error 1 make[2]: *** Waiting for unfinished jobs.... Please check the logfile usage.makeinfo.log for errors: out/common-macros.texi:189: warning: @end ifset should only appear at a line beginning out/usage/running.texi:31: warning: @vindex should only appear at a line beginning (possibly involving @rlearning) out/usage/running.texi:48: warning: @vindex should only appear at a line beginning (possibly involving @rweb) out/usage/running.texi:948: warning: @vindex should only appear at a line beginning (possibly involving @ruser) out/usage/running.texi:952: warning: @vindex should only appear at a line beginning (possibly involving @ruser) out/usage/running.texi:1115: warning: @vindex should only appear at a line beginning (possibly involving @rlearning) out/usage/running.texi:1124: warning: @vindex should only appear at a line beginning (possibly involving @ruser) out/usage/running.texi:1149: warning: @vindex should only appear at a line beginning (possibly involving @ruser) out/usage/lilypond-book.texi:32: warning: @vindex should only appear at a line beginning (possibly involving @rweb) out/usage/lilypond-book.texi:32: warning: @vindex should only appear at a line beginning (possibly involving @rweb) out/usage/lilypond-book.texi:1177: @itemx must follow @item out/usage/lilypond-book.texi:1181: @itemx must follow @item out/usage/lilypond-book.texi:1185: @itemx must follow @item out/usage/lilypond-book.texi:1190: @itemx must follow @item out/usage/lilypond-book.texi:1199: @itemx must follow @item out/usage/lilypond-book.texi:1203: @itemx must follow @item out/usage/lilypond-book.texi:1208: @itemx must follow @item out/usage/lilypond-book.texi:1231: @itemx must follow @item out/usage/external.texi:274: warning: @vindex should only appear at a line beginning (possibly involving @rweb) out/usage/external.texi:323: warning: @vindex should only appear at a line beginning (possibly involving @rinternals) out/usage/external.texi:324: warning: @vindex should only appear at a line beginning (possibly involving @rinternals) out/usage/external.texi:587: warning: @vindex should only appear at a line beginning (possibly involving @rweb) out/usage/external.texi:607: warning: @vindex should only appear at a line beginning (possibly involving @ruser) out/usage/suggestions.texi:68: warning: @vindex should only appear at a line beginning (possibly involving @ruser) out/usage/suggestions.texi:119: warning: @vindex should only appear at a line beginning (possibly involving @ruser) make[2]: *** [out/lilypond-usage.info] Error 1 Please check the logfile contributor.makeinfo.log for errors: ./common-macros.itexi:189: warning: @end ifset should only appear at a line beginning out/contributor.texi:50: warning: @vindex should only appear at a line beginning (possibly involving @rweb) out/contributor.texi:50: warning: @vindex should only appear at a line beginning (possibly involving @rweb) contributor/introduction.itexi:24: warning: @vindex should only appear at a line beginning (possibly involving @rweb) contributor/introduction.itexi:24: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) contributor/introduction.itexi:24: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) contributor/introduction.itexi:26: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) contributor/introduction.itexi:26: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) contributor/introduction.itexi:26: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) contributor/introduction.itexi:26: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) contributor/introduction.itexi:26: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) contributor/introduction.itexi:120: warning: @vindex should only appear at a line beginning (possibly involving @rweb) contributor/quick-start.itexi:347: superfluous argument to @end quotation: . (possibly involving @warning) contributor/quick-start.itexi:458: warning: @vindex should only appear at a line beginning (possibly involving @rcontrib) included/compile.itexi:31: warning: @vindex should only appear at a line beginning (possibly involving @rweb) contributor/doc-work.itexi:91: warning: @vindex should only appear at a line beginning (possibly involving @rweb) contributor/issues.itexi:740: warning: @vindex should only appear at a line beginning (possibly involving @rweb) contributor/regressions.itexi:90: warning: @vindex should only appear at a line beginning (possibly involving @rweb) contributor/regressions.itexi:518: warning: @vindex should only appear at a line beginning (possibly involving @rglosnamed) contributor/regressions.itexi:519: warning: @vindex should only appear at a line beginning (possibly involving @rinternalsnamed) included/compile.itexi:925: raising the section level of @unnumberedsubsubsec which is too low make[2]: *** [out/lilypond-contributor.info] Error 1
Control: clone -1 -2 Control: reassign -2 texinfo Control: retitle -2 @itemx which does not follow @item should be a warning instead of an error Control: found -2 5.1.dfsg.1-2 Control: affects -2 lilypond Control: severity -2 important This is because texinfo recently changed @itemx failing to follow @item from a warning to an error, as near as I can tell. This should probably be demoted to a warning and result in @item being interpreted as an @item and the offending packages fixed. [If you search google for @itemx you'll see lots of other cases of things failing to build.] Don Armstrong
control: affects -1 krb5 Hi. I'm trying to get a security update of krb5 into sid, and this issue is breaking that build. I'm not thrilled at the concept of making significant documentation changes to the upstream docs in order to get this to build.
forwarded 707195 http://savannah.gnu.org/bugs/?38997 stop On 15.05.13 Sam Hartman (hartmans@debian.org) wrote: Hi. Forwarded to upstream. H.
I'd strongly recommend making this bug sevirity serious. I don't think we want to be unable to build a bunch of stuff in testing.
Sorry guys, I have the feeling you are missing something... That is not my problem. This has to be fixed either upstream or in the eronous .texi file. What do you mean wiht "security update" for sid - there is no such concept. Security updates are for stable. And in stable it will build properly. Then communicate this to upstream. They will have the same problems in many other distributions sooner or later. Again, that is out of discussion. With this reasoning any lib upgrade that bumps .so or changes API would be forbidden in SID. Of course we waited till after release, but now it is time, and docs should be fixed *TILL* the *NEXT* release, which is in 2(?) years. We reported it upstream (thanks Hilmar), but I consider this a bug in the texii files that need to be fixed. I had and have to fix often bugs/mis-code in other packages because a library changed in sid, that is called normal development process. Norbert
The concern I have is that I'd like to upload a very small patch that fixes a security issue to the krb5 package in sid and get it into testing quickly. I am nervous when I think about including that patch along with a bunch of doc changes. In general, when we're going to break a bunch of builds, we tend to file bugs first, fix the broken packages, and then break the dependency. So, for example we don't do library updates to break a bunch of stuff without first trying to make sure that the broken stuff actually will rebuild against the new library. In this instance, I think the right approach is to introduce a debian-specific patch that turns it into a warning not an error, file bugs against the broken packages. If upstream accepts the issue, then close those bugs. After some reasonable time remove the debian-specific patch if upstream doesn't agree with the issue.
On 15.05.13 Norbert Preining (preining@logic.at) wrote: Hi, Agreed on this. I didn't explicitely submit this as bug, but rather as "consider to revert that change". Having that bug as serious in Debian is definitly out of discussion. H.
tags 707195 - fixed-upstream tags 707195 + wontfix stop On 15.05.13 Sam Hartman (hartmans@debian.org) wrote: Hi, The texinfo people (K. Berry) refused to revert that change and turn the error back into a warning. The only thing you can do is to bother your upstream: they should fix their texinfo files. H.
Hi!
Have the maintainers of texinfo at least considered analyzing the
impact of this decision and started a Mass-Bug-Filling? I'm really
bothered by having tens of unreported RC bugs sitting somewhere
Christoph
Christoph Egger <christoph@debian.org> writes:
Additionally some information on how to fix problems due to the texinfo
change e.g. at [0] would be highly appreciated. I'm looking specifically
at #708460 and related issues [1].
Christoph
[0] http://wiki.debian.org/qa.debian.org/FTBFS
[1]
;; &foo would be amusingly bold in the pdf thanks to TeX/Texinfo
;; interactions,so we escape the ampersand -- amusingly for TeX.
;; sbcl.texinfo defines macros that expand @&key and friends to &key.
Hi Sam, the occurrences of @itemx without a preceeding @item into @item I have added an entry to the FTBFS wiki page: http://wiki.debian.org/qa.debian.org/FTBFS Norbert
Hi Christopher, Links please? I tried to search the bug database using http://bugs-search.debian.org/cgi-bin/search.cgi which ended in an Internal Server Error ... @itemx problem. Sample patch: @@ -1001,20 +1001,20 @@ ... @end example -@itemx --skip-lily-check +@item --skip-lily-check Do not fail if no lilypond output is found. It is used for LilyPond Info documentation without images. -@itemx --skip-png-check +@item --skip-png-check Do not fail if no PNG images are found for EPS files. It is used for LilyPond Info documentation without images. .... The texinfo manual, ch 11.4.3, states: Use '@item' for the first entry, and '@itemx' for all subsequent entries; '@itemx' must always follow an '@item' command, with no blank line intervening. Concerning the other item, please send me the problematic texinfo code (minimal example if possible). Thanks Norbert
affects 707195 - src:krb5 affects 707195 - krb5 thanks Upstream krb5 uses Sphinx instead of texinfo for documentation now.
notforwarded 707195 stop Marking as not forwarded. H.