Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
On Wed, May 25, 2011 at 10:39 PM, Andreas Tille wrote: > I have not built the package - but packages with lintian *warnings* can > be uploaded (and IMHO the rpath issue is only a warning, but I might be > wrong). Could anybody post the output of "lintian -i" ? i cannot update the svn today. So I'll post the output of lintian I got the last time I tried: $ ls -rw-r--r-- 1 mathieu mathieu1609 May 11 13:04 ginkgocadx_2.4.1.1-1_amd64.changes $ lintian -i ginkgocadx_2.4.1.1-1_amd64.changes W: ginkgocadx source: newer-standards-version 3.9.2 (current is 3.9.1) N: N:The source package refers to a Standards-Version which is newer than the N:highest one lintian is programmed to check. If the source package is N:correct, then please upgrade lintian to the newest version. (If there is N:no newer lintian version, then please bug lintian-ma...@debian.org to N:make one.) N: N:Severity: normal, Certainty: certain N: W: ginkgocadx: spelling-error-in-description throught through N: N:Lintian found a spelling error in the package description. Lintian has a N:list of common misspellings that it looks for. It does not have a N:dictionary like a spelling checker does. It is particularly picky about N:spelling and capitalization in package descriptions since they're very N:visible to end users. N: N:Severity: minor, Certainty: certain N: W: ginkgocadx: new-package-should-close-itp-bug N: N:This package appears to be the first packaging of a new upstream N:software package (there is only one changelog entry and the Debian N:revision is 1), but it does not close any bugs. The initial upload of a N:new package should close the corresponding ITP bug for that package. N: N:This warning can be ignored if the package is not intended for Debian or N:if it is a split of an existing Debian package. N: N:Refer to Debian Developer's Reference section 5.1 (New packages) for N:details. N: N:Severity: normal, Certainty: certain N: W: ginkgocadx: binary-without-manpage usr/bin/ginkgocadx N: N:Each binary in /usr/bin, /usr/sbin, /bin, /sbin or /usr/games should N:have a manual page N: N:Note that though the man program has the capability to check for several N:program names in the NAMES section, each of these programs should have N:its own manual page (a symbolic link to the appropriate manual page is N:sufficient) because other manual page viewers such as xman or tkman N:don't support this. N: N:If the name of the man page differs from the binary by case, man may be N:able to find it anyway; however, it is still best practice to make the N:case of the man page match the case of the binary. N: N:If the man pages are provided by another package on which this package N:depends, lintian may not be able to determine that man pages are N:available. In this case, after confirming that all binaries do have man N:pages after this package and its dependencies are installed, please add N:a lintian override. N: N:Refer to Debian Policy Manual section 12.1 (Manual pages) for details. N: N:Severity: normal, Certainty: possible N: W: ginkgocadx: postinst-has-useless-call-to-ldconfig N: N:The postinst script calls ldconfig even though no shared libraries are N:installed in a directory controlled by the dynamic library loader. N: N:Refer to Debian Policy Manual section 8.1.1 (ldconfig) for details. N: N:Severity: minor, Certainty: certain N: W: ginkgocadx: postrm-has-useless-call-to-ldconfig N: N:The postrm script calls ldconfig even though no shared libraries are N:installed in a directory controlled by the dynamic library loader. N: N:Refer to Debian Policy Manual section 8.1.1 (ldconfig) for details. N: N:Severity: minor, Certainty: certain N: E: ginkgocadx: binary-or-shlib-defines-rpath usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 /usr/lib/openmpi/lib N: N:The binary or shared library sets RPATH. This overrides the normal N:library search path, possibly interfering with local policy and causing N:problems for multilib, among other issues. N: N:The only time a binary or shared library in a Debian package should set N:RPATH is if it is linked to private shared libraries in the same N:package. In that case, place those private shared libraries in N:/usr/lib/. Libraries used by binaries in other packages should N:be placed in /lib or /usr/lib as appropriate, with a proper SONAME, in N:which case RPATH is unnecessary. N: N:To fix this problem, look for link lines like: N:gcc test.o -o test -Wl,--rpath,/usr/local/lib N:or N:gcc test.o -o test -R/usr/local/lib N:and remove the -Wl,--rpath or -R argument. You can also use the chrpath N:utility to remove the RPATH. N: N:Refer to http://wiki.debian.org/RpathIssue for details. N: N:Severity: serious, Certainty: possible N: E: ginkgocadx: binary
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Am Donnerstag, 26. Mai 2011, 09:31:19 schrieb Mathieu Malaterre: See below > On Wed, May 25, 2011 at 10:39 PM, Andreas Tille wrote: > > I have not built the package - but packages with lintian *warnings* can > > be uploaded (and IMHO the rpath issue is only a warning, but I might be > > wrong). Could anybody post the output of "lintian -i" ? > > i cannot update the svn today. So I'll post the output of lintian I > got the last time I tried: > > $ ls > -rw-r--r-- 1 mathieu mathieu1609 May 11 13:04 > ginkgocadx_2.4.1.1-1_amd64.changes > > $ lintian -i ginkgocadx_2.4.1.1-1_amd64.changes > W: ginkgocadx source: newer-standards-version 3.9.2 (current is 3.9.1) > N: > N:The source package refers to a Standards-Version which is newer than > the N:highest one lintian is programmed to check. If the source > package is N:correct, then please upgrade lintian to the newest > version. (If there is N:no newer lintian version, then please bug > lintian-ma...@debian.org to N:make one.) > N: > N:Severity: normal, Certainty: certain > N: > W: ginkgocadx: spelling-error-in-description throught through > N: > N:Lintian found a spelling error in the package description. Lintian > has a N:list of common misspellings that it looks for. It does not > have a N:dictionary like a spelling checker does. It is particularly > picky about N:spelling and capitalization in package descriptions > since they're very N:visible to end users. > N: > N:Severity: minor, Certainty: certain > N: > W: ginkgocadx: new-package-should-close-itp-bug > N: > N:This package appears to be the first packaging of a new upstream > N:software package (there is only one changelog entry and the Debian > N:revision is 1), but it does not close any bugs. The initial upload of > a N:new package should close the corresponding ITP bug for that > package. N: > N:This warning can be ignored if the package is not intended for Debian > or N:if it is a split of an existing Debian package. > N: > N:Refer to Debian Developer's Reference section 5.1 (New packages) for > N:details. > N: > N:Severity: normal, Certainty: certain > N: > W: ginkgocadx: binary-without-manpage usr/bin/ginkgocadx > N: > N:Each binary in /usr/bin, /usr/sbin, /bin, /sbin or /usr/games should > N:have a manual page > N: > N:Note that though the man program has the capability to check for > several N:program names in the NAMES section, each of these programs > should have N:its own manual page (a symbolic link to the appropriate > manual page is N:sufficient) because other manual page viewers such as > xman or tkman N:don't support this. > N: > N:If the name of the man page differs from the binary by case, man may > be N:able to find it anyway; however, it is still best practice to > make the N:case of the man page match the case of the binary. > N: > N:If the man pages are provided by another package on which this > package N:depends, lintian may not be able to determine that man pages > are N:available. In this case, after confirming that all binaries do > have man N:pages after this package and its dependencies are > installed, please add N:a lintian override. > N: > N:Refer to Debian Policy Manual section 12.1 (Manual pages) for > details. N: > N:Severity: normal, Certainty: possible > N: > W: ginkgocadx: postinst-has-useless-call-to-ldconfig > N: > N:The postinst script calls ldconfig even though no shared libraries > are N:installed in a directory controlled by the dynamic library > loader. N: > N:Refer to Debian Policy Manual section 8.1.1 (ldconfig) for details. > N: > N:Severity: minor, Certainty: certain > N: > W: ginkgocadx: postrm-has-useless-call-to-ldconfig > N: > N:The postrm script calls ldconfig even though no shared libraries are > N:installed in a directory controlled by the dynamic library loader. > N: > N:Refer to Debian Policy Manual section 8.1.1 (ldconfig) for details. > N: > N:Severity: minor, Certainty: certain > N: > E: ginkgocadx: binary-or-shlib-defines-rpath > usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 > /usr/lib/openmpi/lib > N: > N:The binary or shared library sets RPATH. This overrides the normal > N:library search path, possibly interfering with local policy and > causing N:problems for multilib, among other issues. > N: > N:The only time a binary or shared library in a Debian package should > set N:RPATH is if it is linked to private shared libraries in the same > N:package. In that case, place those private shared libraries in N: > /usr/lib/. Libraries used by binaries in other packages should N: >be placed in /lib or /usr/lib as appropriate, with a proper SONAME, in > N:which case RPATH is unnecessary. > N: > N:To fix this problem, look for link lines like: > N:gcc test.o -o test -Wl,--rpath,/usr/local/lib > N:or >
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Am Donnerstag, 26. Mai 2011, 09:31:19 schrieb Mathieu Malaterre: > On Wed, May 25, 2011 at 10:39 PM, Andreas Tille wrote: > > I have not built the package - but packages with lintian *warnings* can > > be uploaded (and IMHO the rpath issue is only a warning, but I might be > > wrong). Could anybody post the output of "lintian -i" ? > > i cannot update the svn today. So I'll post the output of lintian I > got the last time I tried: > > $ ls > -rw-r--r-- 1 mathieu mathieu1609 May 11 13:04 > ginkgocadx_2.4.1.1-1_amd64.changes > > $ lintian -i ginkgocadx_2.4.1.1-1_amd64.changes > W: ginkgocadx source: newer-standards-version 3.9.2 (current is 3.9.1) > N: > N:The source package refers to a Standards-Version which is newer than > the N:highest one lintian is programmed to check. If the source > package is N:correct, then please upgrade lintian to the newest > version. (If there is N:no newer lintian version, then please bug > lintian-ma...@debian.org to N:make one.) > N: > N:Severity: normal, Certainty: certain > N: > W: ginkgocadx: spelling-error-in-description throught through > N: > N:Lintian found a spelling error in the package description. Lintian > has a N:list of common misspellings that it looks for. It does not > have a N:dictionary like a spelling checker does. It is particularly > picky about N:spelling and capitalization in package descriptions > since they're very N:visible to end users. > N: > N:Severity: minor, Certainty: certain > N: > W: ginkgocadx: new-package-should-close-itp-bug > N: > N:This package appears to be the first packaging of a new upstream > N:software package (there is only one changelog entry and the Debian > N:revision is 1), but it does not close any bugs. The initial upload of > a N:new package should close the corresponding ITP bug for that > package. N: > N:This warning can be ignored if the package is not intended for Debian > or N:if it is a split of an existing Debian package. > N: > N:Refer to Debian Developer's Reference section 5.1 (New packages) for > N:details. > N: > N:Severity: normal, Certainty: certain > N: > W: ginkgocadx: binary-without-manpage usr/bin/ginkgocadx > N: > N:Each binary in /usr/bin, /usr/sbin, /bin, /sbin or /usr/games should > N:have a manual page > N: > N:Note that though the man program has the capability to check for > several N:program names in the NAMES section, each of these programs > should have N:its own manual page (a symbolic link to the appropriate > manual page is N:sufficient) because other manual page viewers such as > xman or tkman N:don't support this. > N: > N:If the name of the man page differs from the binary by case, man may > be N:able to find it anyway; however, it is still best practice to > make the N:case of the man page match the case of the binary. > N: > N:If the man pages are provided by another package on which this > package N:depends, lintian may not be able to determine that man pages > are N:available. In this case, after confirming that all binaries do > have man N:pages after this package and its dependencies are > installed, please add N:a lintian override. > N: > N:Refer to Debian Policy Manual section 12.1 (Manual pages) for > details. N: > N:Severity: normal, Certainty: possible > N: > W: ginkgocadx: postinst-has-useless-call-to-ldconfig > N: > N:The postinst script calls ldconfig even though no shared libraries > are N:installed in a directory controlled by the dynamic library > loader. N: > N:Refer to Debian Policy Manual section 8.1.1 (ldconfig) for details. > N: > N:Severity: minor, Certainty: certain > N: > W: ginkgocadx: postrm-has-useless-call-to-ldconfig > N: > N:The postrm script calls ldconfig even though no shared libraries are > N:installed in a directory controlled by the dynamic library loader. > N: > N:Refer to Debian Policy Manual section 8.1.1 (ldconfig) for details. > N: > N:Severity: minor, Certainty: certain > N: > E: ginkgocadx: binary-or-shlib-defines-rpath > usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 > /usr/lib/openmpi/lib > N: > N:The binary or shared library sets RPATH. This overrides the normal > N:library search path, possibly interfering with local policy and > causing N:problems for multilib, among other issues. > N: > N:The only time a binary or shared library in a Debian package should > set N:RPATH is if it is linked to private shared libraries in the same > N:package. In that case, place those private shared libraries in N: > /usr/lib/. Libraries used by binaries in other packages should N: >be placed in /lib or /usr/lib as appropriate, with a proper SONAME, in > N:which case RPATH is unnecessary. > N: > N:To fix this problem, look for link lines like: > N:gcc test.o -o test -Wl,--rpath,/usr/local/lib > N:or > N:g
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
On Thu, May 26, 2011 at 9:49 AM, Sebastian Hilbert wrote: >> So in summary rpath is an Error. >> > > Is this something upstream should take care of (which I have CCed) or is this > a packaging issue ? I asked upstream why they used rpath in the first place but never got an answer ... I have not digged any further on why an RPATH for openmpi was there, could be a debian only issue. From quick browsing the cmake script it looks like RPATH is being forced in upstream build system. 2cts -- Mathieu -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/BANLkTi=nwnjtds5mxwdpdvvuctpmwt7...@mail.gmail.com
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Hi Mathieu, On Thu, May 26, 2011 at 09:31:19AM +0200, Mathieu Malaterre wrote: > > i cannot update the svn today. So I'll post the output of lintian I > got the last time I tried: at least I can confirm that svn commits work in principle if you adapt your .ssh/known_hosts to the new server key. If it is no technical reason on your side which prevents you from commiting we might just wait. > $ ls > -rw-r--r-- 1 mathieu mathieu1609 May 11 13:04 > ginkgocadx_2.4.1.1-1_amd64.changes > > $ lintian -i ginkgocadx_2.4.1.1-1_amd64.changes > W: ginkgocadx source: newer-standards-version 3.9.2 (current is 3.9.1) > W: ginkgocadx: spelling-error-in-description throught through Both are easy to fix. > W: ginkgocadx: new-package-should-close-itp-bug I could issue an ITP on behalf of the Debian Med team. Just tell me that I should do this (or do it yourself and CC to the list) to make sure we will not have duplicate ITPs. > W: ginkgocadx: binary-without-manpage usr/bin/ginkgocadx Could be left for later versions - we should cooperate with upstream here. > W: ginkgocadx: postinst-has-useless-call-to-ldconfig > W: ginkgocadx: postrm-has-useless-call-to-ldconfig That's a usual warning in connection with the rpath issue. > E: ginkgocadx: binary-or-shlib-defines-rpath > usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 > /usr/lib/openmpi/lib So this is really an issue. > N:To fix this problem, look for link lines like: > N:gcc test.o -o test -Wl,--rpath,/usr/local/lib > N:or > N:gcc test.o -o test -R/usr/local/lib > N:and remove the -Wl,--rpath or -R argument. You can also use the chrpath > N:utility to remove the RPATH. > N: > N:Refer to http://wiki.debian.org/RpathIssue for details. We should verify whether " -Wl,--rpath," strings occure in the build log and if possible patch the build system to avoid these. If this is not successfill, I'd consider overriding the error and upload to experimental (leaving a note about this in README.source to document it properly). > E: ginkgocadx: embedded-library > usr/lib/ginkgocadx/libCADxCore.so.2.4.1.1: sqlite As I said I would just link against the Debian packaged sqlite. > So in summary rpath is an Error. Yes. If you need any help I'd volunteer for this, but we should find some way to base our work on the same code in debian/. So we need to either get svn working again or find some other way to colaborate. Thanks for your work on this Andreas. -- http://fam-tille.de -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110526081835.ga20...@an3as.eu
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Le Thu, May 26, 2011 at 10:18:35AM +0200, Andreas Tille a écrit : > Hi Mathieu, > > at least I can confirm that svn commits work in principle if you adapt > your .ssh/known_hosts to the new server key. If it is no technical > reason on your side which prevents you from commiting we might just > wait. Hi all, on my side, I had to change (I actually just removed) ~/.subversion on svn.debian.org… I do not know why. Cheers, -- Charles -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110526082706.gf15...@merveille.plessy.net
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Charles, On Thu, May 26, 2011 at 10:27 AM, Charles Plessy wrote: > Le Thu, May 26, 2011 at 10:18:35AM +0200, Andreas Tille a écrit : >> Hi Mathieu, >> >> at least I can confirm that svn commits work in principle if you adapt >> your .ssh/known_hosts to the new server key. If it is no technical >> reason on your side which prevents you from commiting we might just >> wait. > > Hi all, > > on my side, I had to change (I actually just removed) ~/.subversion on > svn.debian.org… I do not know why. > Here is what I have (before & after removing the $HOME/.subversion): $ svn co svn+ssh://malat-gu...@svn.debian.org/svn/debian-med/trunk/packages/dicom3tools svn: To better debug SSH connection problems, remove the -q option from 'ssh' in the [tunnels] section of your Subversion configuration file. svn: Network connection closed unexpectedly I simply gave up for now on this. -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/banlktikgzvgzv-vb6mxk01rstk0aatt...@mail.gmail.com
SVN issues (Was: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6)
Hi, On Thu, May 26, 2011 at 11:25:18AM +0200, Mathieu Malaterre wrote: > Here is what I have (before & after removing the $HOME/.subversion): As I said - you need to fix .ssh/known_hosts. > $ svn co > svn+ssh://malat-gu...@svn.debian.org/svn/debian-med/trunk/packages/dicom3tools > svn: To better debug SSH connection problems, remove the -q option > from 'ssh' in the [tunnels] section of your Subversion configuration > file. > svn: Network connection closed unexpectedly > > I simply gave up for now on this. I just copied the solution to the packaging list. For further reference read here: http://lists.debian.org/debian-edu/2011/05/msg00100.html I can confirm that I yesterday followed the strategy described in this mail and it worked. So try ssh svn.debian.org Good luck Andreas. -- http://fam-tille.de -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110526094046.gd20...@an3as.eu
Re: SVN issues (Was: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6)
On Thu, May 26, 2011 at 11:40 AM, Andreas Tille wrote: > > ssh svn.debian.org thanks that works now :) I have not touched the package since r6811, so the rpath issue is still there. Hopefully Carlos will be able to guide us. Thanks -- Mathieu -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/BANLkTi=fwotpcsvggnx5go82pfxcuek...@mail.gmail.com
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Hello, thank you all for your work on this. On 05/26/2011 10:18 AM, Andreas Tille wrote: >> W: ginkgocadx: postinst-has-useless-call-to-ldconfig >> W: ginkgocadx: postrm-has-useless-call-to-ldconfig > That's a usual warning in connection with the rpath issue. >> E: ginkgocadx: binary-or-shlib-defines-rpath >> usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 >> /usr/lib/openmpi/lib > So this is really an issue. But I think lintian is wrong here. The error should just be overridden. If libvisualizator.so is not used in any other context thatn ginkgo-cadx, then the location in Plugins is just right and the binary should get that rpath setod ld.conf.d be extended or ... but for the moment I am rather happy about that rpath. >> N:To fix this problem, look for link lines like: >> N:gcc test.o -o test -Wl,--rpath,/usr/local/lib >> N:or >> N:gcc test.o -o test -R/usr/local/lib >> N:and remove the -Wl,--rpath or -R argument. You can also use the chrpath >> N:utility to remove the RPATH. >> N: >> N:Refer to http://wiki.debian.org/RpathIssue for details. > We should verify whether " -Wl,--rpath," strings occure in the build log > and if possible patch the build system to avoid these. If this is not > successfill, I'd consider overriding the error and upload to > experimental (leaving a note about this in README.source to document it > properly). Seconded. >> E: ginkgocadx: embedded-library >> usr/lib/ginkgocadx/libCADxCore.so.2.4.1.1: sqlite > As I said I would just link against the Debian packaged sqlite. This is of course annoying. I'd ignore that, too, though. Just make it appear in experimental as soon as possible. The SQLite is shipping as a source with ginkgo-cad, right? Otherwise we might be risking a violation with the GPL (if sqlite is GPLed). >> So in summary rpath is an Error. > Yes. If you need any help I'd volunteer for this, but we should find > some way to base our work on the same code in debian/. So we need to > either get svn working again or find some other way to colaborate. I'd consider it a limitation of lintian, nothing blocking an upload to unstable. Ok, maybe that is part why I don't volunteer as a helper to the ftp masters. The sqlite thingy I'd prefer fixed for unstable, though. Best regards, Steffen -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4dde4249.9030...@gmx.de
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian6
Am Donnerstag, 26. Mai 2011, 14:06:33 schrieb Steffen Möller: > > As I said I would just link against the Debian packaged sqlite. > > This is of course annoying. I'd ignore that, too, though. > Just make it appear in experimental as soon as possible. > The SQLite is shipping as a source with ginkgo-cad, right? > Otherwise we might be risking a violation with the GPL > (if sqlite is GPLed). http://www.sqlite.org/copyright.html Public domain. Whatever that means. Sebastian -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201105261422.52691.sebastian.hilb...@gmx.net
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian6
On Thu, May 26, 2011 at 02:22:52PM +0200, Hilbert, Sebastian wrote: > > > As I said I would just link against the Debian packaged sqlite. > > > > This is of course annoying. I'd ignore that, too, though. > > Just make it appear in experimental as soon as possible. > > The SQLite is shipping as a source with ginkgo-cad, right? > > Otherwise we might be risking a violation with the GPL > > (if sqlite is GPLed). > > http://www.sqlite.org/copyright.html > > Public domain. Whatever that means. It means, do as you like. Karsten -- GPG key ID E4071346 @ gpg-keyserver.de E167 67FD A291 2BEA 73BD 4537 78B9 A9F9 E407 1346 -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110526122746.ge2...@hermes.hilbert.loc
Re: libbam-dev ships sam.h in /usr/include/samtools
On 05/26/2011 08:19 AM, Charles Plessy wrote: Le Thu, May 26, 2011 at 08:08:20AM +0200, Alex Mestiashvili a écrit : Would it be possible just to roll back to latest commit done by you , and start over from that point with my changes and the new upstream 1.0.2 ? Yes, this is something that I was considering: recover your changes with git-format-patches, reset (--hard) to the latest clean commit, and apply the relevant patches with git am. Cheers, Hi Charles , I've tried reset --hard , but when I was applying the patches with git-am I was getting errors like error: AUTHORS: already exists in index ... I also tried reset --soft but I can't push changes to the remote . for example I do : git clone git reset --soft 0b7256275dc8b8f7495a46b6b33739573fd0d1c0 git add . git commit -a and git log shows my commit right after yours . but it get's rejected : ! [rejected]master -> master (non-fast-forward) ! [rejected]pristine-tar -> pristine-tar (non-fast-forward) Is it possible to roll back everything as it was before my commits in the repository ? Alex -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4dde7a22.5040...@biotec.tu-dresden.de
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
On Wed, 25 May 2011, Andreas Tille wrote: Unfortunately I am not yet able to commit anything after the alioth rework. There is still an error: svn: Can't create directory '/svn/debian-med/db/transactions/6851-1.txn': Read-only file system Uhmm. :-( Any location for http download might do for the moment - I need to check commit tomorrow. Did you find anything? As everybody else but me could commit something, maybe I lost the write access to the repository? So maybe it would help to add me to the project once again? Thorsten -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/pine.lnx.4.64.1105262002090.29...@tor.gallien.in-chemnitz.de
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
On Thu, 26 May 2011, Steffen Möller wrote: E: ginkgocadx: binary-or-shlib-defines-rpath usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 /usr/lib/openmpi/lib So this is really an issue. But I think lintian is wrong here. The error should just be overridden. No, the rpath is really added to the library. Unfortunately there are two different paths added any only the openmpi needs to be removed. Anyway, I put my version of the debian directory to http://debian.alteholz.de/ginkgocadx-debian.tar.bz2 This should solve the rpath issue and the sqlite linking. As my next svn-buildpackage is still running, I am not sure whether everything is really working. Thorsten
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Hi, all. The openmpi dependency may be caused by ITK because a huge portion of it is template based. I'll review the "rpath problem explain" reference that Sebastian sent to us, but we are too busy right now to deep in the problem. [ For some notes about the issue, please read below the mail paste ] Mathiew, Our response may have lost. It was: ==8<== El 11/05/2011, a las 14:16, Carlos Barrales escribió: > Hi Mathieu. > > I'm not sure. It was the only way I was able to pass shlibs scanning. > > -- > Carlos Barrales Ruiz. > carlos.barra...@metaemotion.com > http://healthcare.metaemotion.com > > > On Wed, May 11, 2011 at 1:13 PM, Mathieu Malaterre > wrote: > Carlos, > > Even after applying your patch I get an error reported by lintian: > > E: ginkgocadx: binary-or-shlib-defines-rpath > ./usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 > /usr/lib/openmpi/lib > E: ginkgocadx: binary-or-shlib-defines-rpath > ./usr/lib/ginkgocadx/libCADxCore.so.2.4.1.1 /usr/lib/openmpi/lib > > > Do you really need to use rpath ? ==>8== What I mean is: If I disable rpath, cmake build is done, but dh_shlibdeps fails so it can't found libCADxCore.so for ginkgocadx binary as well as libvisualization.so Maybe a easy fix should be to split ginkgo package in three ones with: * ginkgocadx executable binary and manual. * libCADxCore.so with its translations. * libvisualization.so with its translations. Doing this, dh_shlibdeps should be able to find deps so they must be previously installed. I don't know if dh_shlibdeps stands for LD_LIBRARY_PATH export or any other kind of path location to find those libraries or wathever. I don't know how this rule works exactly. El 26/05/2011, a las 10:16, Mathieu Malaterre escribió: > On Thu, May 26, 2011 at 9:49 AM, Sebastian Hilbert > wrote: >>> So in summary rpath is an Error. >>> >> >> Is this something upstream should take care of (which I have CCed) or is this >> a packaging issue ? > > I asked upstream why they used rpath in the first place but never got > an answer ... > I have not digged any further on why an RPATH for openmpi was there, > could be a debian only issue. From quick browsing the cmake script it > looks like RPATH is being forced in upstream build system. > > 2cts > -- > Mathieu -- Carlos Barrales Ruiz. carlos.barra...@metaemotion.com http://www.metaemotion.com/ -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/696bdb2b-6f8e-4a97-91cf-9d08518e8...@metaemotion.com
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
Am Donnerstag, 26. Mai 2011, 09:31:19 schrieb Mathieu Malaterre: response from upsteam (unquoted) see below. start of response Hi, all. The openmpi dependency may be caused by ITK because a huge portion of it is template based. I'll review the "rpath problem explain" reference that Sebastian sent to us, but we are too busy right now to deep in the problem. [ For some notes about the issue, please read below the mail paste ] Mathiew, Our response may have lost. It was: ==8<== El 11/05/2011, a las 14:16, Carlos Barrales escribió: > Hi Mathieu. > > I'm not sure. It was the only way I was able to pass shlibs scanning. > > -- > Carlos Barrales Ruiz. > carlos.barra...@metaemotion.com > http://healthcare.metaemotion.com > > > On Wed, May 11, 2011 at 1:13 PM, Mathieu Malaterre wrote: > Carlos, > > Even after applying your patch I get an error reported by lintian: > > E: ginkgocadx: binary-or-shlib-defines-rpath > ./usr/lib/ginkgocadx/Plugins/libvisualizator.so.2.4.1.1 > /usr/lib/openmpi/lib > E: ginkgocadx: binary-or-shlib-defines-rpath > ./usr/lib/ginkgocadx/libCADxCore.so.2.4.1.1 /usr/lib/openmpi/lib > > > Do you really need to use rpath ? ==>8== What I mean is: If I disable rpath, cmake build is done, but dh_shlibdeps fails so it can't found libCADxCore.so for ginkgocadx binary as well as libvisualization.so Maybe a easy fix should be to split ginkgo package in three ones with: * ginkgocadx executable binary and manual. * libCADxCore.so with its translations. * libvisualization.so with its translations. Doing this, dh_shlibdeps should be able to find deps so they must be previously installed. I don't know if dh_shlibdeps stands for LD_LIBRARY_PATH export or any other kind of path location to find those libraries or wathever. I don't know how this rule works exactly. El 26/05/2011, a las 10:16, Mathieu Malaterre escribió: > On Thu, May 26, 2011 at 9:49 AM, Sebastian Hilbert > wrote: >>> So in summary rpath is an Error. >>> >> >> Is this something upstream should take care of (which I have CCed) or is this >> a packaging issue ? > > I asked upstream why they used rpath in the first place but never got > an answer ... > I have not digged any further on why an RPATH for openmpi was there, > could be a debian only issue. From quick browsing the cmake script it > looks like RPATH is being forced in upstream build system. > > 2cts > -- > Mathieu -- Carlos Barrales Ruiz. carlos.barra...@metaemotion.com http://www.metaemotion.com/ -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201105262101.55072.sebastian.hilb...@gmx.net
Re: Ginkgo-CADx package - was: Install on ubuntu 10.04.2 or debian 6
On Thu, May 26, 2011 at 08:03:49PM +0200, Thorsten Alteholz wrote: > > Did you find anything? What exactly is the problem / error message if you try to commit? > As everybody else but me could commit something, > maybe I lost the write access to the repository? So maybe it would help > to add me to the project once again? I just verified that your (alteholz-guest) status is "Junior Developer" which should grant your commit permissions. Can you please try again and report the exact error? Kind regards and thanks for helping on Ginkgo CADx anyway Andreas. -- http://fam-tille.de -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110526204915.ga15...@an3as.eu
Re: libbam-dev ships sam.h in /usr/include/samtools
Hi Alex, I found that format-patch can be given a path, so that it only produces patch for it. Here is how I have prepared a new cufflinks repository: debcheckout -a git://git.debian.org/debian-med/cufflinks.git --git-track='*' cd cufflinks/ git format-patch 0b7256275dc8b8f7495a46b6b33739573fd0d1c0 debian/ # 0b7256275dc8b8f7495a46b6b33739573fd0d1c0 is my last commit in this repository mv 00* .. git reset --hard 0b7256275dc8b8f7495a46b6b33739573fd0d1c0 git am --committer-date-is-author-date ../0001-new-debian-dir-for-cufflinks.patch git am --committer-date-is-author-date ../0002-added-headers-for-patches-according-http-dep.debian..patch uscan --force-download git import-orig --pristine-tar ../cufflinks_1.0.2.orig.tar.gz git am --committer-date-is-author-date ../0003-Imported-upstream-version-1.0.2.patch git am --committer-date-is-author-date ../0004-added-ITP-number-for-the-initial-release.patch # patches 5,6,7 were redundant… git am --committer-date-is-author-date ../0008-added-docs.patch git am --committer-date-is-author-date ../0009-added-missing-dependency.patch git am --committer-date-is-author-date ../0010-libbam-dev-provides-necessary-headers-only-from-0.1..patch git diff origin/master # Contents of the new master branch were identical to origin/master # Then I went in Alioth, and created a new empty cufflinks repository with setup-repository # I temporarly disabled its commit hook, to not spam with the following: git push --all ; git push --tags If you like the new repository, I will delete the old one. The dates are not completely consistent, but overall things look better and simpler than before. I hope that I have not horrified others more used to Subversion. The repository as it was before the correction was completely functional, and the issue that I have fixed is borderline cosmetic :) Have a nice day, -- Charles Plessy Debian Med packaging team, http://www.debian.org/devel/debian-med Tsurumi, Kanagawa, Japan -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110527055622.gc1...@merveille.plessy.net
Re: libbam-dev ships sam.h in /usr/include/samtools
On 05/27/2011 07:56 AM, Charles Plessy wrote: Hi Alex, I found that format-patch can be given a path, so that it only produces patch for it. Here is how I have prepared a new cufflinks repository: debcheckout -a git://git.debian.org/debian-med/cufflinks.git --git-track='*' cd cufflinks/ git format-patch 0b7256275dc8b8f7495a46b6b33739573fd0d1c0 debian/ # 0b7256275dc8b8f7495a46b6b33739573fd0d1c0 is my last commit in this repository mv 00* .. git reset --hard 0b7256275dc8b8f7495a46b6b33739573fd0d1c0 git am --committer-date-is-author-date ../0001-new-debian-dir-for-cufflinks.patch git am --committer-date-is-author-date ../0002-added-headers-for-patches-according-http-dep.debian..patch uscan --force-download git import-orig --pristine-tar ../cufflinks_1.0.2.orig.tar.gz git am --committer-date-is-author-date ../0003-Imported-upstream-version-1.0.2.patch git am --committer-date-is-author-date ../0004-added-ITP-number-for-the-initial-release.patch # patches 5,6,7 were redundant… git am --committer-date-is-author-date ../0008-added-docs.patch git am --committer-date-is-author-date ../0009-added-missing-dependency.patch git am --committer-date-is-author-date ../0010-libbam-dev-provides-necessary-headers-only-from-0.1..patch git diff origin/master # Contents of the new master branch were identical to origin/master # Then I went in Alioth, and created a new empty cufflinks repository with setup-repository # I temporarly disabled its commit hook, to not spam with the following: git push --all ; git push --tags If you like the new repository, I will delete the old one. The dates are not completely consistent, but overall things look better and simpler than before. I hope that I have not horrified others more used to Subversion. The repository as it was before the correction was completely functional, and the issue that I have fixed is borderline cosmetic :) Have a nice day, Dear Charles , Thank you for the help and the explanation . The repository looks much better , so I think we should leave it . Such a nice features shouldn't be horrifying at all :) Best regards , Alex -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4ddf46de.70...@biotec.tu-dresden.de
