build of documentation completely broken?
Today, while trying to fix something in the thread documentation I noticed that the doc build does not work at all. The last version that worked cvs update -r RC_1_34_0 -D "06 Jan 2007 13:30" After this the cvs update -r RC_1_34_0 -D "06 Jan 2007 13:45" appears to be broken. Obviously this has been caused by a check in from Vladimir Prus. Either the changes expect from the user to change their setups, or something indeed has been broken. I can provide build logs for both cases if wanted. Please give further information. Thank you Roland
Roland Schwarz wrote:
Today, while trying to fix something in the thread documentation I noticed that the doc build does not work at all.
You did not say that command did you run, in which directory, and what error did you get.
The last version that worked
cvs update -r RC_1_34_0 -D "06 Jan 2007 13:30"
After this the
cvs update -r RC_1_34_0 -D "06 Jan 2007 13:45"
And which of the 24 (or 25?) existing timezones is this? As of few minutes ago, running "bjam -n" in the "doc" directory of an up-to-date RC tree works fine for me. I did not change anything is boostbook code. I have some guesses what breakage that could be, but perhaps you can try again? - Volodya
Vladimir Prus wrote:
Roland Schwarz wrote: You did not say that command did you run, in which directory, and what error did you get.
I thought this was obvious, since I am comparing two snapshots, one of them was working. -> cd boost_root/doc -> bjam
And which of the 24 (or 25?) existing timezones is this?
Good question! Of course I do know my timezone (UTC+1) but is this the time I specify in cvs update (didn't find out)? But I can tell you the time of the cvs check in notification e-mail that was generated: Sat, 06 Jan 2007 04:39:45 -0800 (There are actually three of them.)
As of few minutes ago, running "bjam -n" in the "doc" directory of an up-to-date RC tree works fine for me. I did not change anything is boostbook code. I have some guesses what breakage that could be, but perhaps you can try again?
I tried again - and still failed. And some observation which might be of help: I removed the doc directory and then cvs update(ed) afresh (thereby removing the xml subdir). Also I removed the bin.v2 dir before the build. I guess the error is somehow related to doxygen not rebuilding the xml (it is doxygen building this dir isn't it?). Hmm, why bjam -n? This doesn't update anything? I just run bjam. As already said, I can send you the build logs of both cases if you need them. Roland
Roland Schwarz wrote:
Vladimir Prus wrote:
Roland Schwarz wrote: You did not say that command did you run, in which directory, and what error did you get.
I thought this was obvious, since I am comparing two snapshots, one of them was working. -> cd boost_root/doc -> bjam
Given that I cannot reproduce this error, apparently, something is not obvious. And you still did not tell what error you've got.
And which of the 24 (or 25?) existing timezones is this?
Good question! Of course I do know my timezone (UTC+1) but is this the time I specify in cvs update (didn't find out)? But I can tell you the time of the cvs check in notification e-mail that was generated: Sat, 06 Jan 2007 04:39:45 -0800 (There are actually three of them.)
Ok, this is more precise I think.
As of few minutes ago, running "bjam -n" in the "doc" directory of an up-to-date RC tree works fine for me. I did not change anything is boostbook code. I have some guesses what breakage that could be, but perhaps you can try again?
I tried again - and still failed. And some observation which might be of help: I removed the doc directory and then cvs update(ed) afresh (thereby removing the xml subdir). Also I removed the bin.v2 dir before the build. I guess the error is somehow related to doxygen not rebuilding the xml (it is doxygen building this dir isn't it?).
No idea -- I still did not seen "the error".
Hmm, why bjam -n? This doesn't update anything? I just run bjam.
I use "bjam -n" to avoid waiting 5 times in case quickbook must be rebuilt.
As already said, I can send you the build logs of both cases if you need them.
Yes, please do. Just in case it's not clear, I would like all bug reports to say what the error is. - Volodya
Roland Schwarz wrote:
Vladimir Prus wrote:
Roland Schwarz wrote: You did not say that command did you run, in which directory, and what error did you get.
I thought this was obvious, since I am comparing two snapshots, one of them was working. -> cd boost_root/doc -> bjam
It was not obvious; you might have being building just Thread docs.
And which of the 24 (or 25?) existing timezones is this?
Good question! Of course I do know my timezone (UTC+1) but is this the time I specify in cvs update (didn't find out)? But I can tell you the time of the cvs check in notification e-mail that was generated: Sat, 06 Jan 2007 04:39:45 -0800 (There are actually three of them.)
As of few minutes ago, running "bjam -n" in the "doc" directory of an up-to-date RC tree works fine for me. I did not change anything is boostbook code. I have some guesses what breakage that could be, but perhaps you can try again?
I tried again - and still failed. And some observation which might be of help: I removed the doc directory and then cvs update(ed) afresh (thereby removing the xml subdir). Also I removed the bin.v2 dir before the build. I guess the error is somehow related to doxygen not rebuilding the xml (it is doxygen building this dir isn't it?).
Hmm, why bjam -n? This doesn't update anything? I just run bjam.
As already said, I can send you the build logs of both cases if you need them.
I though that in last mail I quite clearly complained about the missing error message. You still did not include it. Eventually, I figured out that the error comes from XSLT processor, not from bjam as I've assumed all the time. It appears the doxygen.jam version I've merged to branch is broken. I've reverted this merge. Now, this seem to mean that we have *two* bugs on HEAD that make building docs impossible. Not quite nice, we really should have better regression testing for docs. - Volodya
Vladimir Prus wrote:
I though that in last mail I quite clearly complained about the missing error message. You still did not include it.
Sorry for this, but I was not able to figure out which error you mean. Building the docs currently produces so much noise, I just didn't see it. With error I just meant the fact that the docs are not getting build, and that the doc/xml directory doesn't get created (which I guessed to be necessary from previous observations.) Also I offered to send you logs of both builds (I didn't want to post them on the list without need). I even tried to find the exact chekin of yours that caused this. In light of this I read your "complained" as "urged".
Eventually, I figured out that the error comes from XSLT processor, not from bjam as I've assumed all the time. It appears the doxygen.jam version I've merged to branch is broken. I've reverted this merge.
Oh I see, but it is caused by doxygen not producing correct results isn't it? This is what I reported.
Now, this seem to mean that we have *two* bugs on HEAD that make building docs impossible. Not quite nice, we really should have better regression testing for docs.
Yes, definitely. What would also be of help I guess is cleaner separation of source and targets as we have it for code. So the xml directory not being rebuilt would not have slipped through. Roland
Vladimir Prus wrote:
Now, this seem to mean that we have *two* bugs on HEAD that make building docs impossible. Not quite nice, we really should have better regression testing for docs.
At least 2 :-( With this mornings cvs and building from boost-root/doc I see: The mpi module has not been initialized; MPI libraries will not be built. hint: add "using mpi ;" to user-config.jam to enable MPI support. warning: Python location is not configured warning: the Boost.Python library won't be built c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:985: in virtual-target.register-actual-name from module virtual-target error: Duplicate name of actual target: <phtml>HTML.manifest error: previous virtual target { boostbook%boostbook.docbook-to-htmldir-HTML.manifest.HTMLDIR { boostbook%boostbook.boostbook-to-docbook-boost.docbook.DOCBOOK { src/boost.xml.XML } } } error: created from ./doc error: another virtual target { boostbook%boostbook.docbook-to-htmldir-HTML.manifest.HTMLDIR { boostbook%boostbook.boostbook-to-docbook-mpi.docbook.DOCBOOK { quickbook%quickbook.quickbook-to-boostbook-mpi.xml.XML { mpi.qbk.QUICKBOOK } } } } error: created from ../libs/parallel/doc/mpi error: added properties: <catalog>..\bin.v2\libs\parallel\doc\msvc-8.0\debug\threading-multi/catalog.xml error: removed properties: <catalog>..\bin.v2\doc\msvc-8.0\debug\threading-multi/catalog.xml <dependency>object(file-target)@1055 <dependency>object(file-target)@1114 <dependency>object(file-target)@1128 <dependency>object(file-target)@1144 <dependency>object(file-target)@1160 <dependency>object(file-target)@1253 <dependency>object(file-target)@1266 <dependency>object(file-target)@1295 <dependency>object(file-target)@1334 <dependency>object(file-target)@1355 <dependency>object(file-target)@390 <dependency>object(file-target)@610 <dependency>object(file-target)@676 <dependency>object(file-target)@724 <dependency>object(file-target)@782 <dependency>object(file-target)@798 <dependency>object(file-target)@810 <dependency>object(file-target)@863 <dependency>object(file-target)@876 <dependency> object(file-target)@988 <implicit-dependency>object(file-target)@1114 <implicit-dependency>object(file-target)@1128 <implicit-dependency>object(file-target)@1144 <implicit-dependency>object(file-target)@1160 <implicit-dependency>object(file-target)@1266 <implicit-dependency>object(file-target)@1295 <implicit-dependency>object(file-target)@1334 <implicit-dependency>object(file-target)@1355 <implicit-dependency>object(file-target)@798 <implicit-dependency>object(file-target)@810 <implicit-dependency>object(file-target)@876 <implicit-dependency>object(file-target)@988 <xsl:param>boost.libraries=../../libs/libraries.htm c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:459: in actualize-no-scanner from module object(file-target)@988c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:111: in object(file-target)@988.actualize from module object(file-target)@988c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:732: in actualize-source-type from module object(action)@1363c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:751: in actualize-sources from module object(action)@1363c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:698: in object(action)@1363.actualize from module object(action)@1363c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:274: in actualize-action from module object(file-target)@1364c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:467: in actualize-no-scanner from module object(file-target)@1364c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:111: in object(file-target)@1364.actualize from module object(file-target)@1364c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:463: in actualize-no-scanner from module object(file-target)@1372c:/data/boost/develop/boost/tools/build/v2/build\virtual-target.jam:111: in object(file-target)@1372.actualize from module object(file-target)@1372c:/data/boost/develop/boost/tools/build/v2\build-system.jam:475: in load from module build-system c:\data\boost\develop\boost\tools\build\v2/kernel\modules.jam:261: in import from module modules c:\data\boost\develop\boost\tools\build\v2/kernel/bootstrap.jam:132: in boost-build from module c:\data\boost\develop\boost\boost-build.jam:9: in module scope from module Anyone any ideas? John.
participants (3)
-
John Maddock -
Roland Schwarz -
Vladimir Prus