Bug#748535: transition: gnutls28
On 07/09/14 14:11, Andreas Metzler wrote: On 2014-08-31 Andreas Metzler ametz...@bebt.de wrote: On 2014-08-03 Andreas Metzler ametz...@bebt.de wrote: [...] almost the final ones: Hello, This seems to be the final one for some time: nmu git-annex 5.20140831 . ALL . -m 'Rebuild against gnutls28' Scheduled. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-08-31 Andreas Metzler ametz...@bebt.de wrote: On 2014-08-03 Andreas Metzler ametz...@bebt.de wrote: [...] almost the final ones: Hello, This seems to be the final one for some time: nmu git-annex 5.20140831 . ALL . -m 'Rebuild against gnutls28' The transition seems to be almost done, there are just 7 relevant source packages left: * openldap: fixed in GIT, waiting for internal review of maintainership team. * gloox: Has a non-GnuTLS related FTBFS bug, which is forwarded upstream. Marked for autoremoval. * efl: Has a non-GnuTLS related FTBFS bug, which is forwarded upstream. * python-gnutls: NMU proposed, waiting for review.[1] * mod-gnutls: FTBFS since June * licq: waits for gloox. * mandos: waits for python-gnutls. cu Andreas [1] https://lists.debian.org/debian-mentors/2014/09/msg00125.html -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-08-03 Andreas Metzler ametz...@bebt.de wrote: [...] nmu libccrtp_2.0.9-2 . ALL . -m 'Rebuild against gnutls28' almost the final ones: nmu gtklp_1.2.7-2.3 . ALL . -m 'Rebuild against gnutls28' nmu samhain_3.1.0-6 . ALL . -m 'Rebuild against gnutls28' nmu suricata_2.0.3-1 . ALL . -m 'Rebuild against gnutls28' nmu x11vnc_0.9.13-1.1 . ALL . -m 'Rebuild against gnutls28' tia, cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-08-31 Andreas Metzler ametz...@bebt.de wrote: [...] nmu gtklp_1.2.7-2.3 . ALL . -m 'Rebuild against gnutls28' Nevermind about this one, it is only in sid and has an open FTBFS bug. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 31/08/14 16:26, Andreas Metzler wrote: On 2014-08-03 Andreas Metzler ametz...@bebt.de wrote: [...] nmu libccrtp_2.0.9-2 . ALL . -m 'Rebuild against gnutls28' almost the final ones: nmu gtklp_1.2.7-2.3 . ALL . -m 'Rebuild against gnutls28' Not done as you said. nmu samhain_3.1.0-6 . ALL . -m 'Rebuild against gnutls28' nmu suricata_2.0.3-1 . ALL . -m 'Rebuild against gnutls28' nmu x11vnc_0.9.13-1.1 . ALL . -m 'Rebuild against gnutls28' Done. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 03/08/14 19:39, Andreas Metzler wrote: On 2014-07-18 Emilio Pozuelo Monfort po...@debian.org wrote: On 17/07/14 18:33, Andreas Metzler wrote: nmu xbmc_2:13.1~rc1+dfsg1-1 . ALL . -m 'Rebuild against gnutls28' That one failed everywhere. I have opened a serious bug for it. Hello, I just had a successful local build, could you please retry? Done. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-07-17 Andreas Metzler ametz...@bebt.de wrote: On 2014-07-12 Andreas Metzler ametz...@bebt.de wrote: there is another binnmu candidate: [...] some new ones: [...] and more: nmu sipwitch_1.6.1-1 . ALL . -m 'Rebuild against gnutls28' nmu libccaudio2_2.1.3-1 . ALL . -m 'Rebuild against gnutls28' nmu libccrtp_2.0.9-2 . ALL . -m 'Rebuild against gnutls28' tia, cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-07-18 Emilio Pozuelo Monfort po...@debian.org wrote: On 17/07/14 18:33, Andreas Metzler wrote: nmu xbmc_2:13.1~rc1+dfsg1-1 . ALL . -m 'Rebuild against gnutls28' That one failed everywhere. I have opened a serious bug for it. Hello, I just had a successful local build, could you please retry? cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 03/08/14 19:14, Andreas Metzler wrote: On 2014-07-17 Andreas Metzler ametz...@bebt.de wrote: On 2014-07-12 Andreas Metzler ametz...@bebt.de wrote: there is another binnmu candidate: [...] some new ones: [...] and more: nmu sipwitch_1.6.1-1 . ALL . -m 'Rebuild against gnutls28' nmu libccaudio2_2.1.3-1 . ALL . -m 'Rebuild against gnutls28' nmu libccrtp_2.0.9-2 . ALL . -m 'Rebuild against gnutls28' Done. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-07-12 Andreas Metzler ametz...@bebt.de wrote: there is another binnmu candidate: [...] some new ones: nmu gobby-infinote_0.4.94-6 . ALL . -m 'Rebuild against gnutls28' nmu pycurl_7.19.3.1-1 . ALL . -m 'Rebuild against gnutls28' nmu xbmc_2:13.1~rc1+dfsg1-1 . ALL . -m 'Rebuild against gnutls28' TIA. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 17/07/14 18:33, Andreas Metzler wrote: nmu gobby-infinote_0.4.94-6 . ALL . -m 'Rebuild against gnutls28' nmu pycurl_7.19.3.1-1 . ALL . -m 'Rebuild against gnutls28' nmu xbmc_2:13.1~rc1+dfsg1-1 . ALL . -m 'Rebuild against gnutls28' Scheduled. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 17/07/14 18:33, Andreas Metzler wrote: nmu xbmc_2:13.1~rc1+dfsg1-1 . ALL . -m 'Rebuild against gnutls28' That one failed everywhere. I have opened a serious bug for it. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
Good morning, there is another binnmu candidate: nmu cadaver 0.23.3-2 . ALL . -m 'Rebuild against gnutls28' Cadaver does not build-depend directly on gnutls by inherits the build-dependency from neon27, which was upgraded to gnutls28 and has propagated to testing. I could not find any other candidate although gnome-vfs and and xmlsec1 have also switched.[1] cu Andreas [1] Except for gtklp, but it is not in testing and FTBFS. -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-07-12 Andreas Metzler ametz...@bebt.de wrote: Good morning, there is another binnmu candidate: nmu cadaver_0.23.3-2 . ALL . -m 'Rebuild against gnutls28' [...] actually there is a second one: nmu libaqbanking_5.4.3beta-2 . ALL . -m 'Rebuild against gnutls28' cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 12/07/14 09:34, Andreas Metzler wrote: On 2014-07-12 Andreas Metzler ametz...@bebt.de wrote: Good morning, there is another binnmu candidate: nmu cadaver_0.23.3-2 . ALL . -m 'Rebuild against gnutls28' [...] actually there is a second one: nmu libaqbanking_5.4.3beta-2 . ALL . -m 'Rebuild against gnutls28' Both scheduled. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-06-17 Emilio Pozuelo Monfort po...@debian.org wrote: [...] https://release.debian.org/transitions/html/gnutls28.html [...] These packages lose their gnutls dependency on a rebuild: nmu libnet-cups-perl_0.61-2 ALL . -m 'Rebuild to drop libgnutls26 dependency' nmu newsbeuter_2.7-1 ALL . -m 'Rebuild to drop libgnutls26 dependency' nmu xpp_1.5-cvs20081009-1 ALL . -m 'Rebuild to drop libgnutls26 dependency' cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 21/06/14 13:28, Andreas Metzler wrote: nmu libnet-cups-perl_0.61-2 ALL . -m 'Rebuild to drop libgnutls26 dependency' nmu newsbeuter_2.7-1 ALL . -m 'Rebuild to drop libgnutls26 dependency' nmu xpp_1.5-cvs20081009-1 ALL . -m 'Rebuild to drop libgnutls26 dependency' Scheduled. BTW there is a missing dot between the package list and the architecture list. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-06-17 Emilio Pozuelo Monfort po...@debian.org wrote: [...] BTW there are currently some issues with the gcrypt transition, see e.g.: https://buildd.debian.org/status/package.php?p=pycurl Are you aware of that? Not sure if that's part of the gnutls transition or if we should have another bug for that. I think gcrypt is separate. This specific problem seems to be caused by 745216.. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-06-16 Emilio Pozuelo Monfort po...@debian.org wrote: On 15/06/14 19:20, Andreas Metzler wrote: [...] Actually I was wrong, it simply propagated, since libetpan/testing is using libgnutls26. :-) Yep, it is done now. Do you want a tracker for the gnutls26 - gnutls28 transition as per your initial email? I would appreciate that, thank you. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
Control: forwarded -1 https://release.debian.org/transitions/html/gnutls28.html On 16/06/14 19:07, Andreas Metzler wrote: On 2014-06-16 Emilio Pozuelo Monfort po...@debian.org wrote: On 15/06/14 19:20, Andreas Metzler wrote: [...] Actually I was wrong, it simply propagated, since libetpan/testing is using libgnutls26. :-) Yep, it is done now. Do you want a tracker for the gnutls26 - gnutls28 transition as per your initial email? I would appreciate that, thank you. https://release.debian.org/transitions/html/gnutls28.html BTW there are currently some issues with the gcrypt transition, see e.g.: https://buildd.debian.org/status/package.php?p=pycurl Are you aware of that? Not sure if that's part of the gnutls transition or if we should have another bug for that. Regards, Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 15/06/14 19:20, Andreas Metzler wrote: On 2014-06-15 Andreas Metzler ametz...@bebt.de wrote: On 2014-06-10 Emilio Pozuelo Monfort po...@debian.org wrote: On 06/06/14 20:30, Emilio Pozuelo Monfort wrote: [...] Looks like libgadu and wireshark are the only remaining problems. Good morning, Which would both be solved now. However afaict the dependency circle has grown, since the libetpan transition #747534 was started. Oi. Actually I was wrong, it simply propagated, since libetpan/testing is using libgnutls26. :-) Yep, it is done now. Do you want a tracker for the gnutls26 - gnutls28 transition as per your initial email? Regards, Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-06-10 Emilio Pozuelo Monfort po...@debian.org wrote: On 06/06/14 20:30, Emilio Pozuelo Monfort wrote: Ah, good point! libgadu seems to be the only problem so far (it was failing on kbsd and didn't migrate because of that. I have retried it and it failed again). Looks like libgadu and wireshark are the only remaining problems. Good morning, Which would both be solved now. However afaict the dependency circle has grown, since the libetpan transition #747534 was started. Oi. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-06-15 Andreas Metzler ametz...@bebt.de wrote: On 2014-06-10 Emilio Pozuelo Monfort po...@debian.org wrote: On 06/06/14 20:30, Emilio Pozuelo Monfort wrote: [...] Looks like libgadu and wireshark are the only remaining problems. Good morning, Which would both be solved now. However afaict the dependency circle has grown, since the libetpan transition #747534 was started. Oi. Actually I was wrong, it simply propagated, since libetpan/testing is using libgnutls26. :-) Thanks for the handholding. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 06/06/14 20:30, Emilio Pozuelo Monfort wrote: Ah, good point! libgadu seems to be the only problem so far (it was failing on kbsd and didn't migrate because of that. I have retried it and it failed again). Looks like libgadu and wireshark are the only remaining problems. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 05/06/14 20:08, Andreas Metzler wrote: On 2014-06-05 Emilio Pozuelo Monfort po...@debian.org wrote: On 29/05/14 08:09, Andreas Metzler wrote: [...] Okay. So I have got gnutls28 (3.2.14-2) ready for upload (except for dch -r) If you want to take a look at it the package is available on people.debian.org in ~ametzler/GNUTLS/. (debdiff is also attached). Could you please give me a heads-up when I may upload to unstable? I was going to ack this, but I see gnutls/exp failed on kbsd-i386. But you've said you've got 3.2.14-2 ready, so I assume you'll upload 3.2 and not 3.3. So, go ahead! Hello Emilio, I have just uploaded 3.2.15-2 (3.2.14 + bugfixes, including CVE-2014-3466) to unstable. - For paranoia's sake I have verified that it builds on kbsd-i386. ;-) Great. It built everywhere, so I scheduled the level 1 rdeps and all were successfully rebuilt. I have just scheduled all the remaining binnmus. I foresee trouble with mod-gnutls. Regards, Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-06-06 Emilio Pozuelo Monfort po...@debian.org wrote: On 05/06/14 20:08, Andreas Metzler wrote: On 2014-06-05 Emilio Pozuelo Monfort po...@debian.org wrote: [...] I was going to ack this, but I see gnutls/exp failed on kbsd-i386. Just for the archives: I cannot reproduce the error with 3.3 on fisher. [...] I have just uploaded 3.2.15-2 (3.2.14 + bugfixes, including [...] Great. It built everywhere, so I scheduled the level 1 rdeps and all were successfully rebuilt. I have just scheduled all the remaining binnmus. Thank you. I foresee trouble with mod-gnutls. sid FTBFS but the version in testing is linked against libgnutls26. Afaict it should not hurt this transition. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 06/06/14 19:39, Andreas Metzler wrote: On 2014-06-06 Emilio Pozuelo Monfort po...@debian.org wrote: On 05/06/14 20:08, Andreas Metzler wrote: On 2014-06-05 Emilio Pozuelo Monfort po...@debian.org wrote: [...] I was going to ack this, but I see gnutls/exp failed on kbsd-i386. Just for the archives: I cannot reproduce the error with 3.3 on fisher. Ok, I've given it back. [...] I have just uploaded 3.2.15-2 (3.2.14 + bugfixes, including [...] Great. It built everywhere, so I scheduled the level 1 rdeps and all were successfully rebuilt. I have just scheduled all the remaining binnmus. Thank you. I foresee trouble with mod-gnutls. sid FTBFS but the version in testing is linked against libgnutls26. Afaict it should not hurt this transition. Ah, good point! libgadu seems to be the only problem so far (it was failing on kbsd and didn't migrate because of that. I have retried it and it failed again). Regards, Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
Control: tags -1 confirmed On 29/05/14 08:09, Andreas Metzler wrote: On 2014-05-26 Julien Cristau jcris...@debian.org wrote: On Mon, May 26, 2014 at 20:29:18 +0200, Andreas Metzler wrote: So now we need a libgnutls28 - libgnutls-deb0-28 transition as a first step. - How should I go about this, should a open a separate new transition bug report? I think we can use this one. Good morning, Okay. So I have got gnutls28 (3.2.14-2) ready for upload (except for dch -r) If you want to take a look at it the package is available on people.debian.org in ~ametzler/GNUTLS/. (debdiff is also attached). Could you please give me a heads-up when I may upload to unstable? I was going to ack this, but I see gnutls/exp failed on kbsd-i386. But you've said you've got 3.2.14-2 ready, so I assume you'll upload 3.2 and not 3.3. So, go ahead! Regards, Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-06-05 Emilio Pozuelo Monfort po...@debian.org wrote: On 29/05/14 08:09, Andreas Metzler wrote: [...] Okay. So I have got gnutls28 (3.2.14-2) ready for upload (except for dch -r) If you want to take a look at it the package is available on people.debian.org in ~ametzler/GNUTLS/. (debdiff is also attached). Could you please give me a heads-up when I may upload to unstable? I was going to ack this, but I see gnutls/exp failed on kbsd-i386. But you've said you've got 3.2.14-2 ready, so I assume you'll upload 3.2 and not 3.3. So, go ahead! Hello Emilio, I have just uploaded 3.2.15-2 (3.2.14 + bugfixes, including CVE-2014-3466) to unstable. - For paranoia's sake I have verified that it builds on kbsd-i386. ;-) cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' signature.asc Description: Digital signature
Bug#748535: transition: gnutls28
On 2014-05-26 Julien Cristau jcris...@debian.org wrote: On Mon, May 26, 2014 at 20:29:18 +0200, Andreas Metzler wrote: So now we need a libgnutls28 - libgnutls-deb0-28 transition as a first step. - How should I go about this, should a open a separate new transition bug report? I think we can use this one. Good morning, Okay. So I have got gnutls28 (3.2.14-2) ready for upload (except for dch -r) If you want to take a look at it the package is available on people.debian.org in ~ametzler/GNUTLS/. (debdiff is also attached). Could you please give me a heads-up when I may upload to unstable? cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' Warning: these package names were in the second list but not in the first: -- libgnutls-deb0-28 Warning: these package names were in the first list but not in the second: -- libgnutls28 [The following lists of changes regard files as different if they have different names, permissions or owners.] Files only in first set of .debs, found in package libgnutls-xssl0 -- lrwxrwxrwx root/root /usr/share/doc/libgnutls-xssl0 - libgnutls28 Files only in first set of .debs, found in package libgnutls28 -- -rw-r--r-- root/root /usr/lib/i386-linux-gnu/libgnutls.so.28.30.5 -rw-r--r-- root/root /usr/share/doc/libgnutls28/AUTHORS -rw-r--r-- root/root /usr/share/doc/libgnutls28/NEWS.Debian.gz -rw-r--r-- root/root /usr/share/doc/libgnutls28/NEWS.gz -rw-r--r-- root/root /usr/share/doc/libgnutls28/README.gz -rw-r--r-- root/root /usr/share/doc/libgnutls28/THANKS.gz -rw-r--r-- root/root /usr/share/doc/libgnutls28/changelog.Debian.gz -rw-r--r-- root/root /usr/share/doc/libgnutls28/changelog.gz -rw-r--r-- root/root /usr/share/doc/libgnutls28/copyright -rw-r--r-- root/root /usr/share/locale/cs/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/de/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/en@boldquot/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/en@quot/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/eo/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/fi/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/fr/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/it/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/ms/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/nl/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/pl/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/sv/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/uk/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/vi/LC_MESSAGES/gnutls28.mo -rw-r--r-- root/root /usr/share/locale/zh_CN/LC_MESSAGES/gnutls28.mo lrwxrwxrwx root/root /usr/lib/i386-linux-gnu/libgnutls.so.28 - libgnutls.so.28.30.5 Files only in first set of .debs, found in package libgnutls28-dbg -- -rw-r--r-- root/root /usr/lib/debug/.build-id/07/de1cc0c9079a51a3743463d9fccedc59bf98ba.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/0c/6011ad2a436bd7546a45525e6a4fc98a51f1d0.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/1c/13e071b79cb81d3a4f404602383448753f90d4.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/45/c50555b20b48657b0ae864537ecd953bfd3580.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/59/d443d1c09b70b0d4c7251c9c7141c0b1be2cfa.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/8b/25569aec6f31f9d4bd75431c2bc43ceb8cc891.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/90/672981f4ee0a7ebfcaa484641fa453ccb8ffd4.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/97/34be2ed4374e9945566a5366b807c28fd637b1.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/9d/8bd6a9ecfe0792a140dc07918010150c9c.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/ba/7a91c27bc2c539eeb677d9b55e6ee561507328.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/cb/b7282a1d5a87f39dd521dae7e92b10a89918c9.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/db/9f3cee135d6fe3bc6811f79fcf4449912f0fdb.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/de/518716f12c6bc90bca89f9cb1f60e400ab0af3.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/e0/5c25777abf863448452bfc745342f62b8ec2cf.debug -rw-r--r-- root/root /usr/lib/debug/.build-id/f7/baa2eae51b3d49d64792bfd65f8d51b418dccc.debug Files only in first set of .debs, found in package libgnutls28-dev -- lrwxrwxrwx root/root
Bug#748535: transition: gnutls28
On 2014-05-24 Julien Cristau jcris...@debian.org wrote: On Sat, May 24, 2014 at 15:21:12 +0200, Andreas Metzler wrote: [...] i.e. the library packge name would change to libgnutls-deb0-28. - Any thoughts about it? Looks reasonable to me. Thanks! Hello, I had uploaded changed packages to experimental on Saturday and they just passed through NEW. So now we need a libgnutls28 - libgnutls-deb0-28 transition as a first step. - How should I go about this, should a open a separate new transition bug report? thanks, cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On Mon, May 26, 2014 at 20:29:18 +0200, Andreas Metzler wrote: So now we need a libgnutls28 - libgnutls-deb0-28 transition as a first step. - How should I go about this, should a open a separate new transition bug report? I think we can use this one. Cheers, Julien signature.asc Description: Digital signature
Bug#748535: transition: gnutls28
On 2014-05-20 Julien Cristau jcris...@debian.org wrote: On Tue, May 20, 2014 at 19:47:48 +0200, Andreas Metzler wrote: [...] Afaik it is not possible to change the versioning scheme and move from soname.1 foo@0 bar@1 baz@1 to soname.1 foo@1 bar@1 baz@1 without breaking the ABI, is it? Yep. So we would need to change the binary package name and do a soname bump, correct? I wonder if building gnutls with -Bsymbolic wouldn't already be an improvement, by avoiding internal calls getting resolved to the wrong lib... Thinking about gnutls_init or the fact that some functions now support more algoritms I don't think that would be enough. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On 2014-05-24 Andreas Metzler ametz...@bebt.de wrote:
[...]
So we would need to change the binary package name and do a soname bump,
correct?
[...]
Hello,
Which I would do with attached patch, which causes this:
(SID)ametzler@argenau:/tmp/GNUTLS/gnutls-3.2.14/debian/tmp/usr/lib/i386-linux-gnu$
ls -l libgnutls{-deb0,}.so*
lrwxrwxrwx 1 ametzler ametzler 25 May 24 12:58 libgnutls-deb0.so.28 -
libgnutls-deb0.so.28.30.5
-rwxr-xr-x 1 ametzler ametzler 4356504 May 24 12:58 libgnutls-deb0.so.28.30.5
lrwxrwxrwx 1 ametzler ametzler 25 May 24 12:58 libgnutls.so -
libgnutls-deb0.so.28.30.5
i.e. the library packge name would change to libgnutls-deb0-28. - Any
thoughts about it?
BTW: Upstream discussion starts here:
http://lists.gnutls.org/pipermail/gnutls-devel/2014-May/006936.html.
thanks, cu Andreas
--
`What a good friend you are to him, Dr. Maturin. His other friends are
so grateful to you.'
`I sew his ears on from time to time, sure'
Description: Bump soname and symbol-versioning.
Fix the symbol versioning to use a different version than in gnutls26,
since this breaks the ABI also bump the soname.
Use a Debian specific soname and symbol-versioning tag, to make sure
that there is no conflict with future upstream versions.
Author: Andreas Metzler ametz...@debian.org
Bug-Debian: http://bugs.debian.org/7748742
Origin: vendor
Forwarded: no|not-needed|url proving that it has been forwarded
Last-Update: -MM-DD
--- gnutls28-3.2.14.orig/lib/Makefile.am
+++ gnutls28-3.2.14/lib/Makefile.am
@@ -115,6 +115,7 @@ libgnutls_la_SOURCES = $(HFILES) $(COBJE
gnutls.asn pkix.asn libgnutls.map
libgnutls_la_LDFLAGS = -no-undefined \
+ -release deb0 \
-version-info $(LT_CURRENT):$(LT_REVISION):$(LT_AGE)
libgnutls_la_LIBADD = ../gl/libgnu.la x509/libgnutls_x509.la \
--- gnutls28-3.2.14.orig/lib/libgnutls.map
+++ gnutls28-3.2.14/lib/libgnutls.map
@@ -18,7 +18,7 @@
# You should have received a copy of the GNU Lesser General Public License
# along with this program. If not, see http://www.gnu.org/licenses/
-GNUTLS_1_4
+GNUTLS_DEBIAN_0_1_4
{
global:
gnutls_alert_get;
@@ -502,7 +502,7 @@ GNUTLS_1_4
*;
};
-GNUTLS_2_8
+GNUTLS_DEBIAN_0_2_8
{
global:
gnutls_certificate_set_x509_simple_pkcs12_mem;
@@ -536,9 +536,9 @@ GNUTLS_2_8
gnutls_session_ticket_key_generate;
gnutls_session_ticket_enable_client;
gnutls_session_ticket_enable_server;
-} GNUTLS_1_4;
+} GNUTLS_DEBIAN_0_1_4;
-GNUTLS_2_10
+GNUTLS_DEBIAN_0_2_10
{
global:
gnutls_x509_crt_get_issuer_alt_name2;
@@ -564,9 +564,9 @@ GNUTLS_2_10
gnutls_hmac_output;
gnutls_certificate_set_verify_function;
gnutls_safe_renegotiation_status;
-} GNUTLS_2_8;
+} GNUTLS_DEBIAN_0_2_8;
-GNUTLS_2_12
+GNUTLS_DEBIAN_0_2_12
{
global:
gnutls_certificate_set_retrieve_function;
@@ -654,9 +654,9 @@ GNUTLS_2_12
gnutls_certificate_get_issuer;
gnutls_x509_crq_verify;
gnutls_global_set_time_function;
-} GNUTLS_2_10;
+} GNUTLS_DEBIAN_0_2_10;
-GNUTLS_3_0_0 {
+GNUTLS_DEBIAN_0_3_0_0 {
global:
gnutls_x509_trust_list_verify_crt;
gnutls_x509_trust_list_add_crls;
@@ -777,9 +777,9 @@ GNUTLS_3_0_0 {
gnutls_certificate_set_x509_system_trust;
gnutls_session_set_premaster;
gnutls_ocsp_resp_check_crt;
-} GNUTLS_2_12;
+} GNUTLS_DEBIAN_0_2_12;
-GNUTLS_3_1_0 {
+GNUTLS_DEBIAN_0_3_1_0 {
global:
gnutls_pkcs11_get_pin_function;
gnutls_pkcs11_obj_list_import_url2;
@@ -923,7 +923,7 @@ GNUTLS_3_1_0 {
gnutls_certificate_get_crt_raw;
gnutls_record_check_corked;
gnutls_db_get_default_cache_expiration;
-} GNUTLS_3_0_0;
+} GNUTLS_DEBIAN_0_3_0_0;
GNUTLS_PRIVATE {
global:
Bug#748535: transition: gnutls28
On Sat, May 24, 2014 at 15:21:12 +0200, Andreas Metzler wrote:
On 2014-05-24 Andreas Metzler ametz...@bebt.de wrote:
[...]
So we would need to change the binary package name and do a soname bump,
correct?
[...]
Hello,
Which I would do with attached patch, which causes this:
(SID)ametzler@argenau:/tmp/GNUTLS/gnutls-3.2.14/debian/tmp/usr/lib/i386-linux-gnu$
ls -l libgnutls{-deb0,}.so*
lrwxrwxrwx 1 ametzler ametzler 25 May 24 12:58 libgnutls-deb0.so.28 -
libgnutls-deb0.so.28.30.5
-rwxr-xr-x 1 ametzler ametzler 4356504 May 24 12:58 libgnutls-deb0.so.28.30.5
lrwxrwxrwx 1 ametzler ametzler 25 May 24 12:58 libgnutls.so -
libgnutls-deb0.so.28.30.5
i.e. the library packge name would change to libgnutls-deb0-28. - Any
thoughts about it?
Looks reasonable to me. Thanks!
Cheers,
Julien
signature.asc
Description: Digital signature
Bug#748535: transition: gnutls28
On 20/05/14 19:47, Andreas Metzler wrote: On 2014-05-20 Julien Cristau jcris...@debian.org wrote: On Mon, May 19, 2014 at 19:04:43 +0200, Julien Cristau wrote: On Sun, May 18, 2014 at 08:52:53 +0200, Andreas Metzler wrote: [...] Apparently loading both versions of gnutls in the same process leads to crashes. That screams at bugs in gnutls symbol versioning. Are you aware of such issues? BTW: Do you have a reference to (one of) these issues? See #748742. Emilio -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On Mon, May 19, 2014 at 19:04:43 +0200, Julien Cristau wrote: On Sun, May 18, 2014 at 08:52:53 +0200, Andreas Metzler wrote: Package: release.debian.org Severity: normal User: release.debian@packages.debian.org Usertags: transition Hello, I would like to ship jessie without GnuTLS 2.x (gnutls26) as it is dead upstream and _old_. The last upstream release was in February 2013, marking the end of a multi-year series of bugfix-only releases. Transitioning to GnuTLS 3.x is possible nowadays because GMP has switched to LGPLv3+/GPLv2+. Apparently loading both versions of gnutls in the same process leads to crashes. That screams at bugs in gnutls symbol versioning. Are you aware of such issues? So one issue seems to be that while gnutls26 and gnutls28 use versioned symbols, they use the same versions for their symbols. Meaning you get the same issues as without versions, i.e. calling some functions from each. Is there a good reason gnutls28 can't use its own separate set of symbol versions? Cheers, Julien signature.asc Description: Digital signature
Bug#748535: transition: gnutls28
On 2014-05-20 Julien Cristau jcris...@debian.org wrote: On Mon, May 19, 2014 at 19:04:43 +0200, Julien Cristau wrote: On Sun, May 18, 2014 at 08:52:53 +0200, Andreas Metzler wrote: [...] Apparently loading both versions of gnutls in the same process leads to crashes. That screams at bugs in gnutls symbol versioning. Are you aware of such issues? BTW: Do you have a reference to (one of) these issues? So one issue seems to be that while gnutls26 and gnutls28 use versioned symbols, they use the same versions for their symbols. Meaning you get the same issues as without versions, i.e. calling some functions from each. Is there a good reason gnutls28 can't use its own separate set of symbol versions? Hello, Afaict it originated here http://lists.gnu.org/archive/html/gnutls-devel/2007-02/msg6.html and somehow actually appeared at some later point starting here https://www.gitorious.org/gnutls/gnutls/commit/5fe8d8687eb193df0c105e2e6fe926e3676b1bb8 and there was some discussion following http://lists.gnu.org/archive/html/gnutls-devel/2009-05/msg00067.html The main rationale seemed to be that one gets better automatic dependency info on rpm based distributions. (I have not used rpm for a long time, so I do not know.) As a side-effect it helps as documentation. Anyway, my understanding of symbol versioning is that soname.0: foo@0 bar@0 soname.1 foo@0 bar@1 baz@1 would work perfectly well if foo() did exactly the same thing in soname.0 and soname.1 and it would not matter whether the version from soname.0 or soname.1 was chosen at runtime. And that was what upstream was aiming for. Obviously it is a lot easier to get wrong. Afaik it is not possible to change the versioning scheme and move from soname.1 foo@0 bar@1 baz@1 to soname.1 foo@1 bar@1 baz@1 without breaking the ABI, is it? cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On Tue, May 20, 2014 at 19:47:48 +0200, Andreas Metzler wrote: On 2014-05-20 Julien Cristau jcris...@debian.org wrote: On Mon, May 19, 2014 at 19:04:43 +0200, Julien Cristau wrote: On Sun, May 18, 2014 at 08:52:53 +0200, Andreas Metzler wrote: [...] Apparently loading both versions of gnutls in the same process leads to crashes. That screams at bugs in gnutls symbol versioning. Are you aware of such issues? BTW: Do you have a reference to (one of) these issues? Andreas Henriksson was having issues with totem-pl-parser. IIRC he had a backtrace mixing the two gnutlses. They're no longer on paste.debian.net though so I'm hoping he'll be able to fill the gaps :) So one issue seems to be that while gnutls26 and gnutls28 use versioned symbols, they use the same versions for their symbols. Meaning you get the same issues as without versions, i.e. calling some functions from each. Is there a good reason gnutls28 can't use its own separate set of symbol versions? Afaict it originated here http://lists.gnu.org/archive/html/gnutls-devel/2007-02/msg6.html and somehow actually appeared at some later point starting here https://www.gitorious.org/gnutls/gnutls/commit/5fe8d8687eb193df0c105e2e6fe926e3676b1bb8 and there was some discussion following http://lists.gnu.org/archive/html/gnutls-devel/2009-05/msg00067.html The main rationale seemed to be that one gets better automatic dependency info on rpm based distributions. (I have not used rpm for a long time, so I do not know.) As a side-effect it helps as documentation. Anyway, my understanding of symbol versioning is that soname.0: foo@0 bar@0 soname.1 foo@0 bar@1 baz@1 would work perfectly well if foo() did exactly the same thing in soname.0 and soname.1 and it would not matter whether the version from soname.0 or soname.1 was chosen at runtime. And that was what upstream was aiming for. Obviously it is a lot easier to get wrong. Indeed. Afaik it is not possible to change the versioning scheme and move from soname.1 foo@0 bar@1 baz@1 to soname.1 foo@1 bar@1 baz@1 without breaking the ABI, is it? Yep. I wonder if building gnutls with -Bsymbolic wouldn't already be an improvement, by avoiding internal calls getting resolved to the wrong lib... Cheers, Julien signature.asc Description: Digital signature
Bug#748535: transition: gnutls28
Hello! On Tue, May 20, 2014 at 08:02:22PM +0200, Julien Cristau wrote: On Tue, May 20, 2014 at 19:47:48 +0200, Andreas Metzler wrote: On 2014-05-20 Julien Cristau jcris...@debian.org wrote: On Mon, May 19, 2014 at 19:04:43 +0200, Julien Cristau wrote: On Sun, May 18, 2014 at 08:52:53 +0200, Andreas Metzler wrote: [...] Apparently loading both versions of gnutls in the same process leads to crashes. That screams at bugs in gnutls symbol versioning. Are you aware of such issues? BTW: Do you have a reference to (one of) these issues? Andreas Henriksson was having issues with totem-pl-parser. IIRC he had a backtrace mixing the two gnutlses. They're no longer on paste.debian.net though so I'm hoping he'll be able to fill the gaps :) [...] I don't have the backtrace anymore, but crossing over from gnutls26 to gnutls28 happened via the _gnutls_rnd_ops.rnd structure/abstraction. If you want to reproduce this, take the (broken) patch from https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=746641#37 and rebuild totem-pl-parser with it. After installing the patched totem-pl-parser, start totem and goto channels, click the magnifying glass icon in the top right to search, click local in the search entry and pick YouTube instead, enter random search word and press enter Regards, Andreas Henriksson -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
On Sun, May 18, 2014 at 08:52:53 +0200, Andreas Metzler wrote: Package: release.debian.org Severity: normal User: release.debian@packages.debian.org Usertags: transition Hello, I would like to ship jessie without GnuTLS 2.x (gnutls26) as it is dead upstream and _old_. The last upstream release was in February 2013, marking the end of a multi-year series of bugfix-only releases. Transitioning to GnuTLS 3.x is possible nowadays because GMP has switched to LGPLv3+/GPLv2+. Apparently loading both versions of gnutls in the same process leads to crashes. That screams at bugs in gnutls symbol versioning. Are you aware of such issues? Also, why does libgnutls28-dbg conflict with libgnutls26-dbg? Cheers, Julien signature.asc Description: Digital signature
Bug#748535: transition: gnutls28
On 2014-05-19 Julien Cristau jcris...@debian.org wrote: On Sun, May 18, 2014 at 08:52:53 +0200, Andreas Metzler wrote: [...] I would like to ship jessie without GnuTLS 2.x (gnutls26) as it is dead upstream and _old_. The last upstream release was in February 2013, marking the end of a multi-year series of bugfix-only releases. Transitioning to GnuTLS 3.x is possible nowadays because GMP has switched to LGPLv3+/GPLv2+. Apparently loading both versions of gnutls in the same process leads to crashes. That screams at bugs in gnutls symbol versioning. Are you aware of such issues? Hello, Darn. No, first time I heard of this. Also, why does libgnutls28-dbg conflict with libgnutls26-dbg? Oversight, I will fix it. cu Andreas -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#748535: transition: gnutls28
Package: release.debian.org Severity: normal User: release.debian@packages.debian.org Usertags: transition Hello, I would like to ship jessie without GnuTLS 2.x (gnutls26) as it is dead upstream and _old_. The last upstream release was in February 2013, marking the end of a multi-year series of bugfix-only releases. Transitioning to GnuTLS 3.x is possible nowadays because GMP has switched to LGPLv3+/GPLv2+. Relevant bugs (as reported so far) are usertagged https://bugs.debian.org/cgi-bin/pkgreport.cgi?tag=gnutls3;users=ametz...@debian.org. The newer gnutls version is mostly API compatible, only a handful (less than 5) packages showed build-breakage due to removed functions in my test. (Sadly openldap is one of the candidates (ITS#7430 aka ITS#6359). Almost all breakage is due to gnutls switching from gcrypt to nettle, breaking the assumption the -lgcrypt works if -lgnutls does. While there is a obvious solution to this problem (Package: libgnutls-dev / Depends: libgnutls28-dev, libgcrypt20-dev | libgcrypt11-dev) it is probably not the right one to actually ship. Most of this gcrypt usage superfluous, only adapting gcrypt behavior on the assumption that it is used by gnutls. Of the rest, a big part is only using and handful of gcrypt functions (typicall md5 or sha1) and would do well with doing this with the GnuTLS crypto API instead of adding another dependency. I am not sure how to go about this I am looking at packages one at a time. Perhaps it would be better to do a big move like this: #1 Let libgnutls-dev depend on libgnutls28-dev, libgcrypt20-dev | libgcrypt11-dev. #2 Rebuild everything, transition to testing. #3 Get rid of unnecessary gcrypt usage one at a time, add libgcrypt20-dev dependency where really necessary #4 Drop libgnutls-dev's dependency on gcrypt. While this looks good, I expect there will be some #2a Some packages break, they build successfully but break at runtime cu Andreas Ben file: title = gnutls28; is_affected = .depends ~ libgnutls26 | .depends ~ libgnutls-dev | .depends ~ libgnutlsxx27 | .depends ~ libgnutls28 | .depends ~ libgnutls28-dev | .depends ~ libgnutlsxx28; is_good = .depends ~ libgnutls28 | .depends ~ libgnutls28-dev | .depends ~ libgnutlsxx28; is_bad = .depends ~ libgnutls26 | .depends ~ libgnutls-dev | .depends ~ libgnutlsxx27; -- `What a good friend you are to him, Dr. Maturin. His other friends are so grateful to you.' `I sew his ears on from time to time, sure' signature.asc Description: Digital signature
