Bug#947143: Bug#947212: Bug#947143: RFS: wordpress/5.3.2+dfsg1-0.1 [NMU] [RC] -- weblog manager
Hi Markus, Yes Nils was doing a nmu for me. Unless they are very keen I'll handle the backports. As you said the confusion is on the sponsorship. We were using a Mentors as a way of getting the package from him to me in the standard way. - Craig On Tue, 24 Dec. 2019, 4:27 am Markus Koschany, wrote: > Hello Niels, > > Am 23.12.19 um 15:04 schrieb DebBug: > > > Anyone to chime in? Craig? Markus? > > There is a bit of confusion here, so I try to explain the situation and > how we should proceed. Thank you for filing bug report #947212 to track > the security issues in Wordpress. This will help to answer those > questions raised by Adam. However there was already #946905 that you > could have been used as well. > > You have only recently added me to CC, presumably because I have done > some security uploads in the past for Wordpress. I don't know what you > have discussed with Craig and if he wants to review your work and > sponsor it later. Then you actually don't need to open a sponsorship > request on debian-mentors. > > Sponsorship requests are either of severity normal or important. Here it > would be ok to use important but the severity is merely an indicator and > it doesn't automatically guarantee that a bug is prioritized. Security > related bugs like #947212/#946905 are either of severity important or > grave. > > Version 5.3.2 seems to fix a couple of security vulnerabilities. No CVE > has been assigned yet. This version should be uploaded to unstable. > > If you want to fix Wordpress in Buster and Stretch as well, then you > have to go a different route. The security team is responsible for that. > As previously discussed I recommend to base security updates on upstream > releases for specific Wordpress branches. > > https://wordpress.org/download/releases/ > > Buster should be updated to version 5.0.8 and Stretch to 4.7.16. In both > cases you would base your work on the Wordpress packages in Buster and > Stretch. The changes to the debian files should be minimal, you would > merely rebase existing patches and repack the tarball to make it > compliant with the DFSG. > > In short: > > Version 5.3.2 -> unstable > Did Craig agree with the upload? > If there is simply no response because of the holiday season we could do > a NMU with a delay of 5 to 10 days. I assume you haven't made any major > changes to the package. > > After that: > Version 5.0.8 -> buster-security > Version 4.7.16 -> stretch-security > > You can already prepare the packages, then we contact the security team > and ask for approval. > > Regards, > > Markus > >
Bug#805172: RFS: vim-command-t/1.13-2 [ITP]
On Thu, Nov 26, 2015 at 3:45 AM Sam Morriswrote: > I think this happens because you have vim 2:7.4.488-7 (from jessie) > installed, which uses libruby2.1. vim in testing/unstable uses libruby2.2, > so upgrading to this version would get things working. > It would, but the packaging system should ensure that whatever version of vim is installed, the program will work or won't be installable. > In fact, I'll probably just remove the version requirements on those > build-dependencies entirely--they just get in the way when backporting > the package to run on jessie. > Are you sure? Isn't this a linking problem? To me the plugin is linked to ruby 2.2 and vim linked to ruby 2.1 It sort of needs a dpkg-shlibdeps type of thing going on, but with vim not ldd. - Craig
Bug#805172: RFS: vim-command-t/1.13-2 [ITP]
Hmm, something isn't too happy :( command-t.vim could not load the C extension Please see INSTALLATION and TROUBLE-SHOOTING in the help Vim Ruby version: 2.1.5-p273 Expected version: 2.2.3-p173 For more information type::help command-t Press ENTER or type command to continue So.. my pbuilder has ruby 2.2.3 but my vim is linked to libruby2.1.5 $ ldd /usr/bin/vim | grep ruby libruby-2.1.so.2.1 => /usr/lib/x86_64-linux-gnu/libruby-2.1.so.2.1 (0x7f891aef3000) But, I have libruby2.2 So, how do you tell this package to only depend on a vim that is linked to ruby2.2 or how do you tell command-t to link to the vim it knows about. On Tue, Nov 24, 2015 at 10:51 PM Sam Morris <s...@robots.org.uk> wrote: > On Tue, 2015-11-24 at 11:35 +, Craig Small wrote: > > > > On Mon, Nov 16, 2015 at 1:21 AM Sam Morris <s...@robots.org.uk> wrote: > > > I am looking for a sponsor for my package "vim-command-t" > > I tried building it but it looks like you are missing a dependency. > > > > > > make[1]: Entering directory '/build/vim-command-t-1.13' > > cd ruby/command-t && ruby extconf.rb > > mkmf.rb can't find header files for ruby at > > /usr/lib/ruby/include/ruby.h > > debian/rules:7: recipe for target 'override_dh_auto_configure' failed > > make[1]: *** [override_dh_auto_configure] Error 1 > > make[1]: Leaving directory '/build/vim-command-t-1.13' > > debian/rules:4: recipe for target 'build' failed > > make: *** [build] Error 2 > > dpkg-buildpackage: error: debian/rules build gave error exit status 2 > > > > - Craig > > Thanks for letting me know--seems I had ruby2.2-dev still installed on > my machine so didn't notice the missing dependency. I've uploaded 1.13- > 3 with the additional build-dependency added. > > -- > Sam Morris <https://robots.org.uk/> > CAAA AA1A CA69 A83A 892B 1855 D20B 4202 5CDA 27B9 > >
Bug#805172: RFS: vim-command-t/1.13-2 [ITP]
On Mon, Nov 16, 2015 at 1:21 AM Sam Morriswrote: > I am looking for a sponsor for my package "vim-command-t" > I tried building it but it looks like you are missing a dependency. make[1]: Entering directory '/build/vim-command-t-1.13' cd ruby/command-t && ruby extconf.rb mkmf.rb can't find header files for ruby at /usr/lib/ruby/include/ruby.h debian/rules:7: recipe for target 'override_dh_auto_configure' failed make[1]: *** [override_dh_auto_configure] Error 1 make[1]: Leaving directory '/build/vim-command-t-1.13' debian/rules:4: recipe for target 'build' failed make: *** [build] Error 2 dpkg-buildpackage: error: debian/rules build gave error exit status 2 - Craig
Re: Bug#792144: RFS: cunit/2.1-2.dfsg-3 -- Unit Testing Library for C [ITA]
On Mon, Sep 07, 2015 at 11:21:35AM +0300, Azat Khuzhin wrote: > On Sun, Jul 12, 2015 at 11:33:32AM +1000, Riley Baird wrote: > > d/copyright: > > -In the LGPL license text, you've accidentally referred to the wrong > > license: > > You should have received a copy of the GNU General Public License > > Hm, this is what dh_make have in: > /usr/share/debhelper/dh_make/licenses/lgpl2 So it is! I've fixed that now. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5
Re: error adding symbols: DSO missing from command line
On Fri, Sep 04, 2015 at 08:33:52PM +0200, Jakub Wilk wrote: > "help" is not a helpful subject. I updated it, so that is provides some > context. :) > If you want people to help you, you need to make it easy for them to > understand and reproduce the problem. In case of compiler errors, including > the compiler command that failed is absolute minimum. Especially for this one, the common problem here is that the order of options is wrong. e.g. ld thing.o -lqdb -lthingqbneeds So if libqdb needs symbols from libthingqdbneeds it will error like this. Sometimes the dpkg-buildflags (correctly) trips these things up. I've found them a useful check for getting upstream in order. But yes, as Jakub said, bit hard to debug this further without the actual command line. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5
Re: Some questions about copyrights
OK, Paul has talked about embedded copies which I agree with (leaving aside the whole jquery thing) in general. Use the already packaged libraries for that, your life will be better for it! Turning to copyright files, the preferred format is the old DEP-5 format, found at [1] The files stanzas, the last match wins. This means you want your general ones first, then your more specific ones. So almost always Files: * is first. That license, assuming you are not using embedded copies, looks like Unlicense[2] check that it is fully that and if not use a different short name. Also ensure you know the difference between the or later clauses. GPL-2+ means something different to GPL-2 and may mean it won't link to future GPL code. So you might have: Files: * Copyright: Copyright 1980 John Doe jd...@example.com License: GPL-3+ [gpl3 license stuff goes here] Files: somedir/* Copyright: Copyright 1999 John Doe2 f...@example.com License: Unlicense [unlicense stuff goes here] [1] https://www.debian.org/doc/packaging-manuals/copyright-format/1.0/ [2] http://spdx.org/licenses/Unlicense.html -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20150613032039.ga8...@enc.com.au
Re: dep5-copyright-license-name-not-unique
On Tue, Mar 03, 2015 at 07:54:23PM +0100, Helge Kreutzmann wrote: Thanks, this makes the file much shorter. I just wonder why lintian does not accept the longer form (which, by my reading, is allowed as well). My guess it is a coding error and when it was written the programmer didn't consider duplicate and identical license clauses. It is checking for duplicate license codes for different licenses. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20150304115841.ga20...@enc.com.au
Re: Bug#749152: RFS: top/2.2.3-1 [ITP]
On Sun, May 25, 2014 at 02:00:57AM +0200, Hugo Lefeuvre wrote: It should be done now. The new name of the binary is qtop. I've uploaded the package on mentos.debian.org.[0] As the procps maintainer, thanks for doing this. We really don't want to get the two confused! - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140525102706.ge18...@enc.com.au
Re: No upstream versioning
On Thu, Apr 24, 2014 at 09:14:08PM +1000, Benjamin Donald-Wilson wrote: Thanks for the advice. Just one more quick question, should I use date of the commit I'm using or the date that I package it? (last commit was roughly 5 months ago) Their commit date. The reason is there is at least some way of tracing back to where you got that particular tar file. So if I saw 20131224 I could look at the git log and see perhaps you went from commit on 24 Dec 2013. Your download date could be anything from their commit date to (if you lived in New Zealand and waited 30 minutes) tomorrow's date. That's really hard to work out. Oh, and you don't get oddness like this: * They commit and push 1st April * They commit and don't push 2nd April * You download on 5th April and use version 20140405 (you can't see their 2nd April commit) * They eventually push their commits * Why doesn't 20140405 have the commit for 2nd April? - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140424113626.ga5...@enc.com.au
Re: Question about a licensing problem
On Mon, Mar 17, 2014 at 01:17:04PM +0100, Gert Wollny wrote: one version, 2.21 is GPL, but already for reasonable sized brain image data the code crashes, and the way the code is written there is no easy way to fix this, and hence, Oscar went with version 3.0, which works. If the API is the same you might be able to have it depend on either, or do something where the default behaviour is to use the GPL one but be able to use the non-GPL one. The problem with that approach is that someone has to maintain the GPLed version for at least some bug fixes. I don't know how big a job that is, but it might be significant. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140317233402.ga5...@enc.com.au
Re: continue on a failing test (dh7)
On Wed, Feb 12, 2014 at 06:21:24AM +, Roelof Wobben wrote: But then Im going to need a javascript expert. I'm not that! One of the test gives undefined where it schould be given a exception. The whole file can be found here : https://github.com/linuxmint/cjs/blob/master/test/js/testLocale.js if needed I can give the error logs. I suppose the need to decide what they do with invalid input. Can upstream help here? -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140212091132.gb17...@enc.com.au
Re: continue on a failing test (dh7)
On Wed, Feb 12, 2014 at 09:20:06AM +0800, Paul Wise wrote: Please don't do that , tests are there for a reason. It is better to fix the problem or fix the test than ignore the test. To give an example of why this is important. procps had a problem where it would fail on certain buildds. At first I cursed the buildds, cursed the architectures and their flakyness. Nothing I could do would reproduce it, but some buildds would, at depressingly semi-regular times, complain. Eventually the problem was the test was working but made assumptions about the system. These are valid assumptions for a normal setup (you wouldn't run it like this) however the program shouldn't of crashed but complained or exited nicely. I was *that* close to having a rule in the test setup that basically said if the system is in this state, don't run test. This would of masked a real bug in the program; which admittedly not many people will ever see, but it shouldn't be there in the first place. The bug is fixed now, the test will need some adjusting to cater for the error message, but that's the correct way it should respond. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140212013727.ga29...@enc.com.au
Re: Rules for packaging forked software
On Sat, Feb 08, 2014 at 11:31:20PM +0800, Paul Wise wrote: Upstreams continue to disappoint me :( I tend ending up as my own upstream a lot of the time, so I provide my own disappointment. Forks are a pain, but they often can be in the long run a benefit. This is especially so if the parent project has gone dormant, which it sounds like it for at least one of the cases. procps is effectively a fork of a fork, for example - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140209113319.ga13...@enc.com.au
Re: Is this really the debian way ?
On Sat, Feb 08, 2014 at 07:13:59PM +, Roelof Wobben wrote: On the control file these packages are made: cjs - Mozilla-based javascript bindings for the GNOME platform libcjs0c - This is the shared library applications link to libcjs-dev - This package contains the development files applications need to build against. That is very common. You have the programs people use in cjs. It has a shared library or libraries in the lib package with the development library parts (so other people can use it) in a -dev package. That's pretty standard behaviour. DEB_DH_MAKESHLIBS_ARGS_libcjs0c = -Xusr/lib/cjs-1.0/ -V'libcjs0c (= $(DEB_UPSTREAM_VERSION)), libcjs0-$(LIBMOZJS)' -- -c4 So they have to use some sort of hack to find the files of the second package. That looks odd. Let's go through the args * Don't look for my libraries in /usr/lib/cjs-1.0 * The -V option adds libcjs0-$(LIBMOZJS) to the dependency, that looks wrong. Your library package creates the shlibs file so that other packages that link to it have a sensible dependency. As an example, libprocps0 has: libprocps 0 libprocps0 (= 1:3.3.2-1) while libprocps3 has: libprocps 3 libprocps3 So if something is built and linked with libprocps0 then it will have a line in its control file like: Depends: libprocps0 (= 1:3.3.2-1), [other things here] while if it was linked with libprocps3 it looks like: Depends: libprocps3, [other things here] That odd -V option is saying, if you link with libcjs then you need to also install either this or a newer version of libcjs (this is common) but you ALSO need libcjs0-(something here) package. That's very, very odd. shlibs is for finding library depdencies (I need libary XYZ, which package is it in?) not for... well just a dependency for dependency sake. I suspect, but it is only a guess, that libcjs0 should depend on this MOZJS thing. HOWEVER, if you MUST have this, the way in debhelper rules would be: override_dh_makeshlibs: dh_makeshlibs -Xusr/lib/cjs-1.0/ -V'libcjs0c (= $(DEB_UPSTREAM_VERSION)), libcjs0-$(LIBMOZJS)' -- -c4 What worries me is that if you need this MOZJS thing and you are not making debian packages, then shlibs doesn't apply. If it is important enough to do evil things to shlibs, then I suspect this theorectical non-debian program wouldn't work. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140209120049.gb13...@enc.com.au
Re: Is this really the debian way ?
On Sun, Feb 09, 2014 at 03:10:26PM +, Sune Vuorela wrote: On 2014-02-08, Roelof Wobben rwob...@hotmail.com wrote: DEB_DH_MAKESHLIBS_ARGS_libcjs0c = -Xusr/lib/cjs-1.0/ -V'libcjs0c (= $(DEB_UPSTREAM_VERSION)), libcjs0-$(LIBMOZJS)' -- -c4 In this case, it ensures that anything that links libcjs0c ensures that it is installed along with the libcjs0-$(LIBMOZJS) package. That's the problem, it doesn't. It ensures that any *Debian package* built this way will pull it in. shlibs are only looked at at Debian package build time and is used to assist in the generation of the Depends: line for that package being built. Assuming this LIBMOZJS thing is required for libcjs0 to work, then any program compiled against libcjs0 that isn't a Debian package has the potential to break. If libcjs0 needs that LIBMOZJS package to work, then it should be a dependency of libcjs0. It should not be some kludge using shlibs. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140209213711.gc13...@enc.com.au
Re: Is this really the debian way ?
On Sun, Feb 09, 2014 at 12:14:17PM +, Roelof Wobben wrote: I know that cjs depends on mozjs185 package. Is there a better way to make this work ? Without the hack ? cjs the program or libcjs0? Whichever one, add a Dependency. As you're running objdump over a library libcjs0 to find what version of mozjs it uses, I guess it is for libcjs0. That's another bad idea, objdump for the shlibs, fixed version for the -dev package. You can see over time these will diverge. OK, its even wierder. The rules file has: DEB_DH_MAKESHLIBS_ARGS_libcjs0c = -Xusr/lib/cjs-1.0/ -V'libcjs0c (= $(DEB_UPSTREAM_VERSION)), libcjs0-$(LIBMOZJS)' and: echo cjs:Provides=libcjs0-$(LIBMOZJS) debian/$(cdbs_curpkg).substvars So The rules file kludges shlibs so it depends in libcjs0 and libcjs0-mozjs185 and The package that provide libcjs0-mozjs185 is.. libcjs0 After all that mucking around you're back to depending on libcjs0 twice. I'd try to find whoever wrote this and ask is it really needed and if not nuke it. It would only make sense if the mozjs package is also provided by something else and in that case should be a dependency on libcjs0. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140209214952.gd13...@enc.com.au
Re: Moving a package (to non-free)
On Sat, Feb 01, 2014 at 05:19:12PM +0100, Niels Thykier wrote: File /srv/ftp-master.debian.org/dak/dak/process_policy.py, line 136, in binary_component_func .join(Component).one() File /usr/lib/python2.7/dist-packages/sqlalchemy/orm/query.py, line 2193, in one Multiple rows were found for one()) MultipleResultsFound: Multiple rows were found for one() The program was expecting only one row but there was more than one there in the database. I personally don't like one() for this reason. Checking the count and/or using first() is usually better, unless this is pointing to something else breaking (e.g. the thing that let the database have duplicate rows in the first place.) (assuming you haven't implemented it already) Indeed. Not really the uploaders fault though, to answer his question. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140202215456.ga4...@enc.com.au
Re: Restrictive Artwork License
On Sun, Dec 29, 2013 at 02:19:38PM -0500, Paul Tagliamonte wrote: Even considering code in the same jarfile as linked, I don't think you can link an image to code in the same way. I've seen it before. Takes an image and makes a C file, something like char my_img={ 0x11, 0x22, ... etc} Was evil, in so many ways. Sounds like the graphics are non-free and the binary is contrib to me, assuming they can be separated. If the program could run without the images then it could go in main. - Craig -- Craig Small (@smallsees) http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint:5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131230031712.gb23...@enc.com.au
Bug#728716: RFS: xchroot/2.3.2-9 [ITP] -- Hi Debian!
On Tue, Nov 05, 2013 at 05:19:31PM +0800, Paul Wise wrote: The GNU GPL is the appropriate license here, please use it instead of contributing to license proliferation, as recommended by Debian's guide for upstreams: As someone who has written Free Software for nearly 20 years, I completely agree with what Paul is saying here. For your general requirements, the GPL seems to be the best fit. Coincidently or perhaps not, it is also the license most of the software I write uses. The GPL effectively says, here is what you can do with my software, but if you change or distribute it, you have to do the same too. I know its tempting to try to tweak licenses but so, so many people get it wrong. The problem is that we're usually better programmers than we are laywers which is something I'm personally very happy about. License choice is hard; even now (or two years ago) on new projects and not ones I inherit it takes time. What is even harder than license choice is license creation. Large companies have tried it and screwed it up. There is also the problem with code-sharing which if you use something that isn't one of the common ones, code sharing starts to become impossible. Alternatively I know right now any piece of GPL licensed code from any project I can reuse on my GPL projects. I don't even need to stop and think about it; it's just there. And the other way around, your project has a problem already solved in another GPL project? You got the code already. My strong suggestion to you is choose any one of the DFSG-free licenses. A weaker suggestion is that the GPL will serve most of your goals. As an author, you are absolutely free to choose any license for code you wrote, which is why they are only suggestions. Doing this will free you from writing less legal blah and more C code (or whatever your language preference is). To me that's a win for everyone. - Craig -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 signature.asc Description: Digital signature
Re: dh_shlibdeps and non-standard library paths
On Tue, Oct 08, 2013 at 03:44:11PM -0700, Daniel Farina wrote: I've been working on trying to hack up the packaging of PostGIS to try and cope better with multiple, simultaneously installed versions. To that end, I am setting --prefix=/opt/postgresql-9.3-postgis-2.1 and stuffing most of the supporting libraries it generates in there. I hope that's not for the actual packages but for some private package. dh_shlibdeps \ -l/build/buildd/postgresql-9.3-postgis-2.1-src-2.1.0/debian/tmp/opt/postgresql-9.3-postgis-2.1/lib Did you try -Lpostgresql-9.3-postgis-2.1 To me the -L option looks better for this sort of thing than the -l hack. Also on the command line dpkg-shlibdeps -v with the -S flag. -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131013030912.ga5...@enc.com.au
Re: Include in PATH
On Sat, Oct 12, 2013 at 09:51:52PM -0300, Beco wrote: I'm creating a makefile with two installation rules. One, the global installation, program with SGID and score in /var/games/ like some other games do. The second one would install the program and scores under ~/.gamedir/A Global installation allow me to call the game directly because /usr/games is in $PATH. You have to decide at what point you want to make the decision between global or user scores. At install time is a bad point. The two better options are configure time run time For configure time, use a ./configure option, say --enable-globalscores compile with this on and it looks for scores in /var/games and the only way to not use that is recompile. This is not as a big a limitation as you first think. The second way is at run time. You can make it an option or environment variable or some configuration item. If you have to configure stuff in the program already put it with that. You could just try the global directory and if it fails use the local directory. Now, for the second installation process, what would be the best way to include the game on PATH? The game *binary* should be in the PATH, don't mess with the path to find it. The game *score file* should be worked out by the binary. A config item, ./configure option or environment variable. Wether or not it is SGID, it is installed in the same spot. As a package distributor you can ask the user if they want SGID, set is a low priority and default to NO. Would it be polite to do something like: echo PATH=~./gamedir/:$PATH ~/.profile Uh no, you never do that. Have a sane default and let the user know how to set it. PATH is where the binaries are, not score/config/data files. -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131013032533.gb5...@enc.com.au
Re: dh_make and $DEBEMAIL
On Sat, Sep 28, 2013 at 12:42:02AM -0300, Beco wrote: Its said in the man page that dh_make would use my $DEBEMAIL, but the email I see is another one (I suppose composed of login @ host-name ). dh_make tries lots of things to get the right email address which is why you are seeing that email. As you know now, setting the wrong environment variable means it ignores it. PS. Also, hijacking my own thread, I think there is a bug with the short opt -c to set license. Is that so? Just the long option is working for It was bug #684258 and was fixed in dh-make 0.62 in February. Glad it seems, or almost seems, to be working for you now. - Craig (dh-make author) -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131002021520.ga2...@enc.com.au
Re: optional package in Build-Depends (how?)
On Mon, Feb 27, 2012 at 01:42:28PM +1100, Dmitry Smirnov wrote: Indeed it probably could be written as Build-Depends: libgpm-dev [linux-any] But the obvious drawback would be the requirement to know all architectures where this package is available. In this case Build-Depends configuration will be ineffective against changes. That's the problem I have with mudlet. libluajit-5.1-dev [amd64 armel i386 kfreebsd-i386], liblua5.1-0-dev [!amd64 !armel !i386 !kfreebsd-i386], It's not pretty and basically means if other arches come along and support libluajit I have to manually fix it. I don't think I could use or | because it didn't work on some build systems. A or nothing would be handy but I always get worried that you will miss linking because of a transistional bump then the program behaviour changes. Imagine if on the armel libluajit is not available temporarily. I think its better to fail to build than to issue out a package without that linking. Specifically to your testing, valgrind testing should probably be opportunistic, so test if valgrind is available and don't otherwise. I think dejagnu does it that way. - Craig -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120227030921.gb1...@enc.com.au
Re: Correct value for DEP5’s Format: field?
On Sun, Jan 08, 2012 at 06:42:27PM +0100, Daniel Stender wrote: expected URL: http://www.debian.org/doc/packaging-manuals/copyright-format/1.0 I've seen this url twice, including with the trailing slash. However, it gives me a 404 from two Australian locations. http://www.debian.org/doc/packaging-manuals/ shows no copyright-format directory. wget http://www.debian.org/doc/packaging-manuals/copyright-format/1.0/ --2012-01-09 14:01:38-- http://www.debian.org/doc/packaging-manuals/copyright-format/1.0/ Resolving www.debian.org (www.debian.org)... 2001:388:1034:2900::26, 150.203.164.38 Connecting to www.debian.org (www.debian.org)|2001:388:1034:2900::26|:80... connected. HTTP request sent, awaiting response... 404 Not Found 2012-01-09 14:01:38 ERROR 404: Not Found. Is this a mirror problem or a URL problem? The IP is for gluck. - Craig (I never knew gluck was so close, it must be only 100s of metres from me as I type this) -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120109030953.ga18...@enc.com.au
Re: dep5: Two licensing / One file
On Sat, Nov 19, 2011 at 04:08:45PM +0100, Mathieu Malaterre wrote: I am trying to use dep5 format to specify the license info of this file. How would one do ? Something like the following ? It looks like BSD-4-Clause to me. This has it's own problems in itself. Also, note that University of California have announced (in 1999) that clause 3 is no longer required[1]. SPDX also has some information about it but it's notes are unhelpful[2] I believe that means the second one could be BSD-3-Clause then. The first isn't BSD, it's BSD-ish. I could not find anything that matched it directly. This was the problem back then, people made up their own variations which meant some things were unclear. So to me, the simple answer is that you don't have a standard license so the short name is not a standard tag, make something that looks ok and put the entire two sets underneath. - Craig [1] ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change [2] http://spdx.org/licenses/BSD-4-Clause -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org old fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 NEW fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/2019205935.ga9...@enc.com.au
Re: dh --parallel (was: Re: RFS: lebiniou)
On Thu, Oct 20, 2011 at 11:06:54AM +0800, Paul Wise wrote: On Thu, Oct 20, 2011 at 2:47 AM, Joey Hess wrote: Only after reading debhelper's bug logs where the decision was researched and made to not --parallel by default, please. I was not able to find such a bug, could you point it out please? 532805 perhaps? dh_auto_build: prevents make jobs from being run simultaneously It's certainly a fascinating (for me) read anyhow. - Craig -- Craig Small VK2XLZ http://enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org old fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 NEW fingerprint: 5D2F B320 B825 D939 04D2 0519 3938 F96B DF50 FEA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111020044654.ga12...@enc.com.au
Re: RFS: kildclient (updated package)
On Sun, May 29, 2011 at 06:28:58PM -0300, Eduardo M KALINOWSKI wrote: I am looking for a sponsor for my package kildclient. I'll upload it for you. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110602210521.ga3...@enc.com.au
Re: New Backup Application
On Fri, May 20, 2011 at 10:47:34AM +1200, Paul Reddy wrote: I've built a reasonably comprehensive backup and disaster recovery app, and am looking for some help on the next steps, the first of which I believe is to find a mentor/sponsor from this group. Hello Paul, This does sound interesting but you probably should of put up a website or some more information about the program. Have you found someone to sponsor you yet? - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110522014333.ga23...@enc.com.au
Re: RFS: kildclient (updated packages, fixes FTBS)
On Fri, May 13, 2011 at 05:56:23PM -0300, Eduardo M KALINOWSKI wrote: Anyway, if some DD is thinking about uploading this package, please wait. I've submitted a quick fix for the FTBS bug 555000 to Dominic Hargreaves as per his offer in that bug report, so I'll prepare a new version soon. I'm a DD and work on a competitor to kildclient (mudlet) but we're all one big happy FOSS family here so when you get it sorted out let me know. Also, have you read the email regarding the perl transistion as kildclient uses perl? It might be worthwhile waiting for that to settle down first. pkg-config files for gtk+ will not list so many libraries as dependencies (that aren't really dependencies)? There doesn't seem to be a simple fix for gtk's dependency bloat. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110514122242.ga29...@enc.com.au
Re: RFS: ax-emergency-listen
On Mon, May 02, 2011 at 11:45:11AM +0200, Fabrizio Regalli wrote: - Missing copyright and license information: config.h, listen.h lack both of them and ax25dump.c, kissdump.c, utils.c don't have any license information. As such, they are not distributable. I need to talk with the author for try to solve this issue. They sound the like files out of the listen program in ax25-utils. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110502105859.ga18...@enc.com.au
Re: What to do about the situation with id3v2
Hello Stefan, As someone who has been in your situation many times (I look after more orphans than an animal shelter) you do need to think about the longer term. I took over psmisc in 2000, for example. I saw some suggested you make the library a basic shim to the currently developed one and that's probably a good compromise if it doesn't look like too much work. The ideal would be for programs to use a supported library. I would say that the library change would be about the same effort as the command-line one but there's more ongoing support. It's the ongoing work that you need to consider. Also I wonder how many of the other programs know the id3 library state. I was actually looking for an id3 tagger for a program of mine and wasn't aware. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110429224409.gb10...@enc.com.au
Re: override_dh_auto_configure not called
On Thu, Apr 28, 2011 at 10:04:37AM +0200, Mathieu Malaterre wrote: %: dh --parallel --with quilt --buildsystem=cmake $@ and dpkg-source: info: using source format `3.0 (quilt)' look wrong as 3.0 (quilt) does the patching work for you. Also the $@ is right after dh not after your options. override_dh_auto_configure: dh_auto_configure -- -DCMAKE_BUILD_TYPE:STRING=Release -DCMAKE_SKIP_RPATH:BOOL=ON This is a little circular, you probably should replace it with override_dh_auto_configure: ./configure -DCMAKE_BUILD_TYPE:STRING=Release -DCMAKE_SKIP_RPATH:BOOL=ON I'm not sure if you use configure or something else here. The build targets look fine. -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110429225518.gc10...@enc.com.au
Re: Re: installing an end user editable file
On Sun, Feb 13, 2011 at 06:24:06PM +, james frize wrote: I'm looking at installing the links.txt file to ~/.packagename/links.txt, but I don't seem to be able to use the ~wildcard. Is this in rules file? If so you cannot do that at all. It doesn't make any sense for a package to install a file in the users home directory. Generally there is a /etc/package/whatever file for defaults and any user specific files would be in /usr/share/doc/package/examples The file in examples should really tell people what to do, e.g.: # This is an example file for whatever, you can copy this to # ~/.whatever/myfile There are many problems with getting a package installer to install files in a user's home directory. Perhaps the first one I can think of is, which user's directory? - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110213213840.ga32...@enc.com.au
Re: warning: patches have not been applied, applying them now (use --no-preparation to override)
On Tue, Feb 01, 2011 at 12:29:22PM +0100, Mathieu Malaterre wrote: However the build failure is still there. What I do not understand is that the debian-changes-3.6.0-1 patch is generated on the fly by dpkg, therefore I cannot remove it (quilt pop): A rough rule of thumb is if you get a debian-changes file and you weren't expecting it, you're doing something wrong. The hard bit is working out what the something wrong is. /usr/share/doc/autotools-dev/README.Debian.gz is a good file to read for starters. It sounds like you have old config.* files and are updating them. That's a good start! The clean target can delete or rebuild the config.* files and perhaps even run autogen.sh For the patching problem, you can ignore the changes because your script is going to nuke the files anyhow. It's explained pretty well at Raphael's blog entry at: http://raphaelhertzog.com/2011/01/28/3-ways-to-not-clutter-your-debian-source-package-with-autogenerated-files/ Just under A new way to avoid the problem I use this feature in pidgin-musictracker because upstream always forgets to regenerate his po files so my diff would be filled with useless updates. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110201223717.ga18...@enc.com.au
Re: DEP5 and multiple copyrights for same file
On Thu, Jan 20, 2011 at 10:27:41PM +0100, Siegfried-Angel Gevatter Pujals (RainCT) wrote: For extra points, make the first one: License: GPL-2 See /usr/share/common-licenses/GPL-2. I believe it needs the full GPL-2 excerpt just like a normal copyright file has got. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/2011012059.gb21...@enc.com.au
Re: Time of a package to be processed by FTP-masters
On Thu, Dec 16, 2010 at 07:31:56PM +0100, Manuel A. Fernandez Montecelo wrote: Why is this important? In this lapse of time of 2 months, my package (1.7.1 version) was already obsoleted (by 1.7.2, with a lot of bug fixes). So I was wondering if it's worth to invest time in creating a package for 1.7.2, or if the effort might be futile because the package won't be reviewed and approved under any circumstance until freeze is over. By that time 1.7.3 might be out, and my effort wasted again, so I would choose not to create the 1.7.2 package at this moment. Freeze is a trying time for lots of bits of Debian and definitely for the ftp masters. The problem is your question about when they can look at your package revolves around when with the next version of Debian be released, and we all know the answer to that is when it's ready. If there is no real changes between the versions, I'd wait until the one sitting in the queue has made it through. If it is something important I'd update the pending package. It can be frustrating because you've done all this work and then.. nothing, but look on the bright side; you're not one of those poor ftp masters! - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101216215118.ga17...@enc.com.au
Re: RFS: php-net-smtp (updated package)
On Fri, Nov 26, 2010 at 10:28:21PM +0100, Guillaume Delacour wrote: So the only version to consider is 1.4.2-3 ot fix the grave bug, thanks in advance if you could take time to review and sponsor it (I've cc'ed Thomas because he wanted sponsor it as soon as possible). It has an older standards version. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101127121605.ge15...@enc.com.au
Re: RFS: php-net-smtp (updated package)
On Sat, Nov 27, 2010 at 07:07:40PM +0100, Guillaume Delacour wrote: Yes, RT and Thomas said me to not upgrade Standards-Version to 3.9.1 to not introduce any other changes in the package. If needed, i could upgrade the package and had some minor improvments. Ah ok, that makes sense now. It's been uploaded. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101127215534.ga10...@enc.com.au
Re: RFS: php-net-smtp (updated package)
On Mon, Nov 22, 2010 at 07:15:03PM +0100, Guillaume Delacour wrote: I am looking for a sponsor for the new version 1.4.4-1 of my package php-net-smtp. mentors has 1.4.2-3 on it, not 1.4.4-1. I'll sponsor it but sort out the versions first. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 signature.asc Description: Digital signature
Re: RFS: php-log (updated package)
On Mon, Nov 22, 2010 at 07:13:00PM +0100, Guillaume Delacour wrote: I am looking for a sponsor for the new version 1.12.3-1 of my package php-log. I've had a look at it and it all looks fine to upload so I'll sponsor this one. - Craig -- Craig Small VK2XLZhttp://www.enc.com.au/ csmall at : enc.com.au Debian GNU/Linux http://www.debian.org/ csmall at : debian.org GPG fingerprint: 1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 signature.asc Description: Digital signature
Re: RFS: notorious-women
On Fri, Jun 18, 2010 at 09:17:52AM +0200, Didier 'OdyX' Raboud wrote: Bilal Akhtar wrote: Just take the case of my package gnome-media-player. I submitted the RFS two months ago, and got a large number of replies, everyone saying How can you use this package name? It is *not* the official GNOME's Media Player (Totem Is) and so you should change the package name. No one ever Seems like an extremely sensible decision to me. I find funny how the it is in Ubuntu, it must be good argument is used so often. The fact that Ubuntu accepted that package with said name doesn't necessarily make it good or suitable as-is for Debian. Different distros, different policies and different people. I get that too. Sometimes it is valid, sometimes it is the equivalent of small children saying well all my friends are doing it why can't I? i.e. the only argument for it is someone somewhere didn't object to it. I've seen some really bad patches go into Ubuntu, for example. This doesn't mean the entire distribution is terrible or all Unbuntu-specific patches are bad, but I would advise you check them yourself first. Getting back on topic, the name isn't too bad, though I do agree a name like fortunes-fr-*whatever* or fortunes-*whatever*-fr would be better, assuming that it is a fortune file and the quotes are in french, as opposed to french women quoted in english. I also read that someone said notorious is the same as famous people. I while strictly speaking they may be definied the same, for most english speakers it means in a negative way. Someone who is a well-known criminal would be notorious. A famous person who saves lives of others probably is not. A grey area is where the person is famous, but notorious for saying socially iffy quotes. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100618235551.gc12...@enc.com.au
Re: A lot of pending packages
On Wed, Jun 09, 2010 at 10:44:00PM +, Sune Vuorela wrote: On 2010-06-09, Lorenzo De Liso blackz...@gmail.com wrote: I'm a simple debian contributor: I'm trying to get my work in debian through a sponsor [1] [2]. The problem is that I'm waiting for a sponsor since 7 days+ (and not only me, in mentors.debian.net there are 20+ pending packages) [3]. Why are they in pending status and nobody wants to upload them? I know, we all are busy with the real life things, but a When I'm sponsoring packages, which happens from time to time, it is normally packages that I somehow have a interest in. I think that many other sponsors feel it the same way. That's exactly how I work when sponsoring packages. I look after 7 of them and all 7 have a reason for being there. There is only 9 packages that are asking for sponsors. For example, my interests is mostly around KDE, and I really try to avoid python stuff. That kind of rules your two packages out for me. Whereas for me that would be my worst nightmare. A gui toolkit I don't use and haven't got install and a language I don't understand. However, the variety of interests and skills is a good thing. What Sune said is pretty good advice, you may also be able to ask people who look after similiar packages. I sponsored purple-plugin-pack because I maintaint pidgin-musictracker. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100609233105.gb27...@enc.com.au
Re: About package with potential legal risk
On Wed, Jun 09, 2010 at 10:04:36AM +1000, Ben Finney wrote: Rather, I think Noel is saying that replacing them quickly, with any freely-licensed functional images (even ugly ones), is best. Once freely-licensed replacement images are working, other people can suggest improvements and do the work of making better images if needed. There used to be a website where Free Software projects could ask for icons. The tree icon for pstree in the psmisc package came from there, for example. Unfortunately it appears to have gone. It's a shame really, as I'm sure there would be some graphically talented people out there that wouldn't mind drawing icons etc for a project every now and then, but there doesn't seem a way to find them. procps (See bug #192635) has wanted a menu icon since 2003. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100609002741.ga22...@enc.com.au
Re: Using dbconfig-common in maintainer scripts.
On Mon, Apr 12, 2010 at 08:31:01AM +1000, Nikolai Lusan wrote: What happens is that debconf asks the questions (not all the questions I would like), Not asking all the questions sounds like a dpkg-configure priority level. the entries in /etc/dbconfig-common/postfix-cluebringer are all empty strings. Even the first one dbc_install ? If dbc_install is anything but true then most if not all of the subsequent lines will be false. dbc_generate_include=template:/etc/cluebringer/cluebringer.conf dbc_generate_include_perms=660 dbc_generate_include_owner=root:www-data dbc_generate_include_args=-o template_infile=/usr/share/doc/postfix-cluebringer/templates/cluebringer.conf -U OK, so you're using the templates. I've only used the db_get strings but that generally looks like a better way. If I have problems with my postinst, I generally make a verbose postinst script that blurts out everything it is doing etc. It's often found what I'm missing. these packages don't rely on postfix-cluebringer I would like to be able to access the debconf/dbconfig-common database entries for postfix-cluebringer (if it is installed) to optionally configure them automagically for the user. Is this possible? Did you try making a dummy package and using db_get postfix-cluebringer/myvariable While debugging the postinst script I also noticed a whole lot of variables not mentioned in any documentation or example being used by dbconfig-common and I was wondering if any of these are usable in the config/postinst script (for forcing debconf to ask specific questions If its not documented then you might, but then the maintainer might change the name/function of that variable too. I'd not depend on these. They're shell scripts, so not exactly great at information hiding. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100411233630.ga10...@enc.com.au
Re: what it's like to be old debian maintainer
On Tue, Mar 23, 2010 at 03:04:13AM +0100, J??r??my Lal wrote: i'm wondering what happens when you spent hours and years dealing with debian... is it something you don't regreat ? I was wondering if you meant someone who is old who is a Debian maintainer or someone who has been involved in Debian for a long time. I've been working in the Debian project for over 13 years. While there have been highs and lows I don't regret the time. I'm in the learning curve and it's still very interesting, Debian is evolving. It gets better and generally learns from its mistakes. A good example of that is the glibc/libc transistions but there is so much going on now. You'll get better too usually after making some mistakes (In 1997, I once installed /bin/ps as /bin; hopefully noone remembers that) I've been playing around with computer and programming since the very early 80s and writing free software since 1994 and I'm still learning new things. So don't worry about running out of new things to see and do in Debian. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100323042927.ga22...@enc.com.au
Re: FS: tacacs+ (updated package)
On Mon, Feb 08, 2010 at 04:38:23PM +0100, Tourneur Henry-Nicolas wrote: I am looking for a sponsor for my package tacacs+. I've sponsored and uploaded this package. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: RFS: tacacs+
On Thu, Feb 04, 2010 at 06:30:26PM +0100, Tourneur Henry-Nicolas wrote: I just did so and normally, I should have fixed all of the lintian Errors and Warnings. Could somebody review my package again ? Thanks everybody for taking time to help me. I was wondering why you have put the binaries into /usr/bin and not /usr/sbin. Both are not programs standard users run. tac_plus is the TACACS+ daemon and has a man page in section 8 tac_pwd is used by system administrators to make DES passwords to put into the configuration file. Again it has a section 8 man pages. To me this sort of stuff belongs in /usr/sbin. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: RFS: tacacs+
On Wed, Feb 03, 2010 at 09:54:20PM +0100, Tourneur Henry-Nicolas wrote: I did upload a new version of the package, which should fix most of those issue. I'll have a look at it. 1° I didn't manage to get #1 fixed, I guess I should use dh_clean or dh_prep but I don't know how to. I'll look at your package to see what you haven't cleaned up, but if you are using dh_clean you can add a file debian/clean and put the files you want delted in there. Alternatively in the clean target delete the file. 2° Which version of lintian do you use to get those warnings and errors ? I got an updated sid version of lintian but no messages :( I guess I fixed #2, #3 and 4 but it's just based on your description of the problem (thanks for that). This is how I got those messages, this is for your older package: csm...@elmo:~/debian/tacacs$ lintian tacacs+_4.0.4.19-1_amd64.changes W: tacacs+ source: patch-system-but-direct-changes-in-diff users_guide W: tacacs+ source: out-of-date-standards-version 3.7.3 (current is 3.8.4) E: tacacs+: duplicate-updaterc.d-calls-in-postinst tacacs_plus E: tacacs+: duplicate-updaterc.d-calls-in-postrm tacacs_plus N: 2 tags overridden (2 warnings) csm...@elmo:~/debian/tacacs$ lintian -V Lintian v2.3.3 The trick is to run lintian on the changes file, as it does both source and binary package checking, otherwise it only does source (.dsc) or binary (.deb) - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: RFS: tacacs+
On Tue, Feb 02, 2010 at 11:08:37PM +0100, Tourneur Henry-Nicolas wrote: * License : No license, only a copyright file If there was no license, then you couldn't distribute it. The license is what is in the COPYING file which is very MIT-like. I'm pretty sure it's ok. The package is almost lintian clean, only one small warning about some patches issue but I don't fully understand the tag description, could you help me ? I get 4 lintian messages. W: tacacs+ source: patch-system-but-direct-changes-in-diff users_guide W: tacacs+ source: out-of-date-standards-version 3.7.3 (current is 3.8.4) E: tacacs+: duplicate-updaterc.d-calls-in-postinst tacacs_plus E: tacacs+: duplicate-updaterc.d-calls-in-postrm tacacs_plus #1 - You are using dpatch patch system, but you have difference in your source files that are not in dpatch. It looks like you are not deleting generated files, such as users_guide and the debhelper log files. #2 - change the standards line to 3.8.4 in debian/control. #3 and #4 are the same thing - You have duplicated the automatically added sections. Delete debian/postinst, debian/prerm as they are fully autocally added files. For debian/postrm just get the purge entry to delete your log directory. About the license, I guess we should check that everything is fine but what the copyright file mainly says is that we are free to copy/use/modify/distribute the source as we like so that looks fine it think :) The copyright (ie who wrote it) has changed a bit, but the license (ie how you use it) has not. It would be nice to have that confirmed. Also your single dbpatch has an author of root@, you probably want to fix that. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: dh_installinit: what to do if upstream provides a initscript?
On Fri, Jan 08, 2010 at 02:56:56PM +0100, Darshaka Pathirana wrote: , |-o, --onlyscripts |Only modify postinst/postrm/prerm scripts, do not actually install |any init script or default files. May be useful if the init script |is shipped and/or installed by upstream in a way that doesn’t make |it easy to let dh_installinit find it. ` Must have had not enough sleep and/or not enough coffee. ;) I'll put that into the newmaint guide, it's a small enough thing but could catch people. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: no explicit license on upstream
On Fri, Dec 04, 2009 at 03:39:51PM +0100, Gabriele Giacone wrote: My question is about licenses: dependencies have been licensed well otherwise they would not be in Debian but ispy directory and ispy website resources don't explicitly refer to a specific license. Should upstream create a LICENSE file? Or a manifest/disclaimer on download page would be better? Or 1 and 2? Unfortunately without any sort of license file or some sort of note or anything at all then the program is unlicensed. A license generally means something permitting you to do something you normally cannot. So no license means no permission. Now that's probably not the intention of the writers of the software but that is unfortunately the outcome of the program as it stands. The world is filled with programs that are not-quite-rightly licensed. One of the nice things with Debian is we try to fix it, when possible. This isn't changing the minds of the upstream developers, but just getting them to be a bit more explicit about what they were thinking. My suggestion is if they want to choose a license, any license, then use one that most other software like it uses, and failing that use GPL. I personally started using that for my software in 1994 and don't regret it. However, if they have a preference for it, BSD is pretty good too. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: How to write emacs dependency ?
On Wed, Nov 11, 2009 at 06:51:01PM +0100, Sven Joachim wrote: On 2009-11-11 16:43 +0100, Christoph Egger wrote: Depends: emacs21 | emacs22 | emacs23 | xemacs21, gnus | emacs22 | emacs23 | xemacs21 This does not give you any guarantee that somebody trying to run darcsum on emacs21 has the gnus package installed. It does, let's say you have emacs21 installed but not gnus. The first clause matches emacs21, so its satisfied. The second, nothing matches, so you have an unmet dependency. For instances off emacs22, emacs23 and xemacs21 then whatever is installed meets both clauses. How about simply dropping support for emacs21? It's an obsolete Emacs flavor version that has been removed from sid already. That is possibly the better idea in that case. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: problem sign .changes
On Tue, Apr 07, 2009 at 12:38:07AM +0200, Jaromír Mike? wrote: how-to sign .dsc and .changes files? debsign file.changes I tried $ debsing -sgpg file.changes and have error: secret key not available gpg is the default. I already generated my key by $gpg --gen-key command. do you have ~/.gnupg/secring.gpg ? - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Building localy
On Mon, Mar 23, 2009 at 09:23:24PM +0100, Jaromír Mike? wrote: dh_clean -k Undefined subroutine Getopt::Long::GetOptionsFromArray called at /usr/share/perl5/Debian/Debhelper/Dh_Getopt.pm line 76. make: *** [install] Error 255 It's not your rules, it looks like your perl is missing a module or you are using an old perl. It should also fail on the command line if you type dh_clean - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Need to debianize again?
On Mon, Mar 16, 2009 at 06:13:40PM +0100, tombs wrote: I'm trying to work with a package that was already debianized before. But nothing more than that. The first question to ask is, is the packaging done by the previous maintainer ok? If it is more or less what you want that's a different situation to a bunch of debian files that are wrong. If I was happy about how the package was, then just edit the control and changelogs and off you go. Also a rough rule of thumb is that upstream debian files are wrong. That's a gross generalisation but one I'd start with (ie treat the upstream debian files with caution) - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: debian/rules
On Thu, Mar 12, 2009 at 05:10:38AM -0300, Rogério Brito wrote: On Mar 11 2009, Jaromír Mike?? wrote: # Add here commands to clean up after the build process. -$(MAKE) -C source/ PREFIX=/usr clean Hummm, here you should also make sure that you don't ignore errors from make (see the hyphen that was put there). Are current versions of dh_make producing such templates? Reported in bug 432667 and fixed in dh-make 0.45 which was released May 2008. Offtopic perhaps, but I'm working on the next release of dh-make which looks like extra features rather than bug-fixing. Some upstream have unusual ways of doing things indeed. :-) You can say that again! - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: FLOSS universities??
On Fri, Feb 13, 2009 at 05:19:03PM -0700, Tom D. Davidson wrote: Please forgive me if this is inappropriate for this list. It is one of those which email list should this go to problems. I have a one shot opportunity to return to school and am after a Software Engineering degree or a project based Computer Science program the focuses on FLOSS tools and methodologies. It would be good if you could find it, but if not do not despair. Surely the theoretical and abstract part of software engineering could apply for FLOSS or non-FLOSS. Even some low-level stuff crosses over; C is still C even if you have to use that infernal visual thing. A good solid start to me would be more important. When I was at uni FLOSS wasn't even really an idea, hell even Linux was radical, but my undergraduate thesis was the first two pieces of Free Software I wrote. If you want it too, it has a way of sneaking in. Good luck on your search and studies! - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Patching Upstream Source
On Sun, Jan 04, 2009 at 02:49:31AM -0600, Boyd Stephen Smith Jr. wrote: Can changes need to be made to files provided by upstream, e.g. a Debian-specific patch or a patch not yet integrated upstream, be done in the .diff.gz and be compliant with Standards-Version 3.8.0? Or, do they need to be done as part of the patch target in debian/rules, e.g. via dpatch? I believe either are fine, as far as policy is concerned. However unless you have trivial patches you will find using dpatch makes life so much easier. I used to just use the old diff.gz way (hey, it was the only way then) and switched to using dpatch. Now I only ever use dpatch. It is a lot easier to handle the patches that way. The dpatch files are still in the diff.gz anyhow. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to debian-mentors-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: hi from argentina, info about complex package
On Tue, Oct 21, 2008 at 11:54:23AM -0200, David Damian Faerman wrote: I am new in this list, I have read a lot about make a simple package but I can't find any info about the other package types... (multiple binary, library, kernel module). Those sound like the dh_make types that it asks you about if you don't specify on the command line, except simple is single. I always recommend you try to make a single package first. There is enough to learn about that for starters. Debian packages are generally more difficult to build than others, but you get a much better result in the end. The order you list them is probably the order from less to more complicated. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Library Packaging
On Fri, Oct 03, 2008 at 09:44:44PM +0100, Neil Williams wrote: Generally, I don't recommend that anyone on this list seriously considers packaging shared libraries until they have a couple of ordinary binary packages in the archive. I would agree with that. Debian packaging is quite daunting for someone who's not seen it before, let alone packaging a library. Some upstream Makefiles are packager-friendly and others it looks like they go out of their way to make life tough. dh-make helps, but it cannot cater for everything. Another problem with your email is you don't actually say what is going wrong. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 'cdarch' has a dash version
On Fri, Oct 03, 2008 at 10:28:02PM +0200, Jann Horn wrote: I uploaded a package, but it causes the following message: Your package does not seem to be lintian clean. 'lintian' is a tool to verify if source package contain obvious packaging errors. These warnings/errors were found: W: cdarch source: native-package-with-dash-version Is cdarch truly a native package? If so, why did they use a dash in the version string. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: dh_make
On Wed, Sep 24, 2008 at 02:27:06PM +0200, Alessio Giovanni Baroni wrote: always I am. In drqueue (0.64.3, last) there is also a python interface. The actual package (drqueue-0.60.0-1.deb) have not this interface. I must to include it? And if yes, in dh_make can I to select multiple binaries? dh_make won't help you there, as this situation is a little unsual and dh_make can only cater for usual cases. A few more lines in the control file and some work with dh_install should do it. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: New debhelper tool proposal
On Mon, Jun 30, 2008 at 05:42:32PM +0800, Paul Wise wrote: On Tue, Jul 1, 2008 at 12:08 AM, Luca Bruno [EMAIL PROTECTED] wrote: What I suggest is a tool that does the following stuff after dh_make has been called: Why not add these smarts to dh_make itself? Some of them don't look like they should be done at dh_make run time. If they (the ones that should be done at the dh_make run time) can be split out and put into a bug I can have a look at them. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: A simply project to read
On Wed, May 21, 2008 at 12:36:41AM +0200, Alexandre González wrote: I'm searching for a simple project to read the code. I download the code of a lot of programs, but I don't know how start, a lot of archives and a lot of unreadable (for me) code. Any suggestion? The package hello might be a good start. Also, look for some package that is close to the sort of thing you are interested in and look at a few there. All packages in main have to have their source code available so you have plenty to choose from. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Easiest way to Debianise a package?
On Fri, Apr 11, 2008 at 06:00:41PM +0200, Ivan Vucica wrote: 1. What are your recommendations with regards to what to use to perform initial debianisation of the project? Meaning - what should I use? Just debhelper? CDBS? I'd use dh-make, but I'm obviously biased in that regard. I didn't like CDBS myself. 2. Should data files be produced by same debianisation directory (by which I mean folder containing files copyright, rules, etc)? Depending on the relative sizes, data files should probably go in their own package. copyright goes into ALL Debian binary packages. rules files are required by all Debian source packages. 3. Should source file generated by autotools contain the data files? Will that make debianisation any easier? Generally make dist will create the Makefile.in and configure files but not the Makefile itself. Whatever make dist does should be good enough. 4. In case my questions are wrong: What would be your steps for producing packages project and project-data, considering what I mentioned above and directory structure below? I'd have a debian/project-data.files file and use dh_movefiles What would be your desires for upstream -- what should upstream do to make your life easier? Use autotools correctly (difficult and I'm usually the upstream!) Don't put undistributable files in the source package. Don't muck up the copyrights of various files, for example a mixture of files with conflicting copyrights. Don't mandatorily link to something not in Debian and that could not ever be in Debian. The autotools usually take care of it, but make sure its easy to shift the destination of the install target. For example you don't want make install to install something in /bin with no way of changing it with either a PREFIX or DESTDIR option. What would you do after fetching SVN (not tar.gz) that contains data as below, and whose make dist would generate .tar.gz containing just sources? There are some tools to make things easier, I'd just download the svn and make dist. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
FTP excuses for lprng dont make sense to me
Hello, I'm trying to work out what is going on with the lprng package. Specifically why it is not getting into testing. The excuses file says: # lprng (- to 3.8.A~rc4-1) * Maintainer: Craig Small * 11 days old (needed 10 days) * Ignoring high urgency setting for NEW package * out of date on hppa: lprng (from 3.8.A~rc2-1) * out of date on armel: lprng (from 3.8.A~rc2-1) (but armel isn't * keeping up, so nevermind) * lprng (source, i386, alpha, amd64, arm, hppa, ia64, mips, mipsel, * powerpc, s390, sparc, armel) has new bugs! * Updating lprng introduces new bugs: #462605 * Not considered It looks like HPPA and Armel just take their sweet time so its not a build problem, but a taking too long to get up-to-date problem. Now, where is my previous version (3.8.A~rc2-1) gone? It's not like lprng is new, its years old! since 1996. Bug 462605 was fixed, not introduced, in 3.8.A~rc4-1. So, what's going on here? - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Config files which are writable by www-data
On Sun, Feb 10, 2008 at 01:20:14PM +0100, Roland Gruber wrote: but when I copy the files at install time (postinst) then /usr/share/doc should be no problem? If the administrator deletes the files in /usr/share/doc afterwards then my application will have no problems. I got a similiar situation. If the program is going to read-in the files, even as templates, put them into /usr/share/packaganame If they are just examples and the admin can copy and edit them then /usr/share/doc/packagename/examples is where they should go. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to set BTS tags nicely
On Tue, Feb 12, 2008 at 02:47:10PM +0100, David Paleino wrote: To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] I do the same except BCC so if anyone replies they don't spammed by the BTS complaining about their reply being commands it doesn't understand. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Re: how to dh_install exclude filename ?
On Tue, Jan 29, 2008 at 12:08:02AM -0200, Andre Felipe Machado wrote: many thanks for your suggestions. Unfortunately, it did not work as expected. Reading the debian/rules file [0] you could see that it is the binary target in a multibinary rules file. It seems that the --exclude argument is simply ignored. That does appear to be strange, I have used the -X or --exclude option for dh_install and it did what it was told. Perhaps you are using include then excludes and it just getting confused? Could you create a manifest file and just use that? - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: updating the patches
On Sun, Nov 25, 2007 at 08:42:55PM -0500, Kamaraju S Kusumanchi wrote: I have a very naive question. I am involved in maintaining texmacs package. I thought it was a good question myself, not naive at all! dptach. The files involved in these patches have been altered upstream in the latest version 1.0.6.12 (Debian Sid currently has 1.0.6.11). So the old patches no longer apply. What is the most efficient way in this scenario? I think it would depend on how much the files have altered, or rather how big and nasty the .rej file is. If it doesn't look too bad, then I would use dpatch-edit-patch to apply the patch and work on that version of the files to fix the problems up. Once you are happy, exit out of the subshell and you have a new patch. If it is patching very messily, then you may want to move the patch out of the way, use dpatch-edit-patch again (but it won't apply the moved patch) and then manually make the changes using the moved patch as your guide. You don't mention what the patches are for, but it is a good idea to consider if you require them anymore. For example you may of had a work-around for some bug and the upstream has put a proper fix in, so its not required. As an aside, dpatch is pretty neat little program that has defintely solved the majority of headaches for patching, though not all. Certainly better than the bad old days :) - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How much free must a package be to be included in non-free (was: Re: CC by-SA 3.0 is DFSG-free?)
On Mon, Jul 02, 2007 at 10:43:05PM -0300, Rogério Brito wrote: Can anybody explain how packages go into non-free? I mean: how much free the package has to be to be considered to non-free and which issues are blocker that would forbid the package into entering non-free? You/we have to be able to legally distribute it. A binary-only distribution agreement may mean it can go in non-free, but a binary-only agreement which says you must download it from a specific website could not (but you could make a contrib installer in that case). I only recall acrobat, or rather the installer for the reader, being in contrib. The problem is that non-free can be there for many reasons. Most of the licenses used in those packages are just plain awful and probably more restrictive than intended. That's why working out if a package can go into non-free, or cannot be distributed at all, is often very hard. I'd say it is harder to do than the main/non-free decision. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: dependancy question?
On Tue, May 15, 2007 at 10:18:11PM -0500, Paul Elliott wrote: Build-Depends: debhelper (= 4.0.0), autotools-dev, libboost-dev, libboost-dev, libboost-regex-dev, libgconfmm-2.6-dev, libgtkmm-2.4-dev, libsigc++-2.0-dev, libboost-filesystem-dev (= 1.33.1 ) libboost-regex-dev (= 1.33.1 ) Why have two libboost-dev? Also there is a libbost-regex-dev , one with a version and one without Depends: ${shlibs:Depends}, ${misc:Depends}, libgtkmm-2.4-1c2a, libgconfmm-2.6-1c2, libsigc++-2.0-0c2a, libboost-filesystem1.33.1, libboost-regex1.33.1 Unless you are doing strange things, this is wrong. Let the shlibdeps work this out for you. The problem can be that while a package, with a recompile, will work absolutely fine with libfoo version 0.1 and 0.2, as they have the same API, there may be some dynamic linking problems. Hard coding the library dependencies is almost always a bad idea and will bite you in the end. I would like to say that peless depends on boost, libgtkmm 2.4, libgconfmm 2.6, and libsigc++ 2.0 without being so specific about the versions. But there appears to be no way to do this, the versions seem to be built into the package names! dpkg-shlibdeps will work out the right things for you, in the case of abinary requiring a specific library and that's the way it should be done. Generally try it without specifying the libraries and see if it looks ok. Have a look in debian/peless/DEBIAN/config too and see if that finds your dependencies. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: manpage tools
On Wed, Mar 28, 2007 at 05:01:56PM +0200, Magnus Holmgren wrote: What tools do you prefer for writing manpages (e.g. for commands that lack one from upstream)? cp and vim OK, not very sophisticated I know but I can bang out a man page pretty quickly this way. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFS: smarty
On Tue, Mar 13, 2007 at 10:41:25PM +0300, Thierry Randrianiriana wrote: I am looking for a sponsor for the new version 2.6.18-1 of my package smarty. Found one yet? If not I'll do this and the -doc - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ csmall at : enc.com.au http://www.debian.org/ Debian GNU/Linux, software should be Free -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bad practice to make a package depend on a specific kernel image
On Tue, Dec 12, 2006 at 01:40:47PM -0500, Jerry DuVal wrote: Is it bad practice to make a package depend on a specific kernel image? This might be a loaded question, but I was just trying to get an opinion. All of the boxes using this package are of the same configuration. Generally speaking, yes it is bad practice. kernel modules packages do, but they are tightly coupled to the kernel (could be considered part of it), so it is ok. Probably for anything else it is a case of bad programming. At the very least they should try to run, notice the missing feature because the kernel is less than version X and gracefully exit. We live in a strange world though, there is probably some other rare reasons why you could depend on a specific version. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: question about reporting bugs on my own package
On Mon, Sep 18, 2006 at 07:48:08AM -0700, tony mancill wrote: This is up to you; I've done it both ways in the past. If the bug is affects the usability of the package, and it could take a while to fix it, go ahead and report the bug to the BTS. That will let other users know that the maintainer is aware of the issue. You may even be able to describe a workaround in the bug report. If the bug is trivial or purely aesthetic then the bug report would be less useful. However, it doesn't hurt to file a bug report, even if the issue is very minor, so if you're in doubt, I would suggest filing one. I would have to second what Tony says above. I maintain packages as a Debian developer and as upstream and yes while there is no rule about it Tony's description is a pretty good way of determining what you can do. Another way of looking at it is; would the bug report be useful to users of your package. This could be either because they know it is not just their system, because you have put in a work-around or so that they know Debian/upstream is aware of it. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: header sanity check?
On Thu, Jul 06, 2006 at 09:05:28PM -0700, Tyler MacDonald wrote: So it's really easy to package a -dev package with a header file, that #include's a header file in a package that it doesn't depend on. Yes it is, it's a compile time problem of someones program. to catch this or if there are any requirements. pbuilder won't even catch it if the Build-Depends for the source package contains all the -dev packages needed. If that particular header file is not used in the compiling then it cannot be caught, as it doesn't result in an error. It is not as easy as it sounds. We have enough trouble with libraries (it's a similiar sort of issue). Consider if you include stdio.h You need to depend on libc6-dev right? Until libc7 comes along, then you're in trouble. Also a lot of headers have things like #ifdef HAVE_FOO_H and only include foo.h if you have it. Lot's of portable autoconf'ed stuff does it like that. Or perhaps parts of the headers only activate under certain circumstances. A program that could use mysql or postgresql databases, for example, may need some of their headers if you use that feature. It would be nice to trap all sorts of problems, but I believe this would create more than it would solve. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: COPYING says GPL, but all headers say LGPL
On Wed, Jun 28, 2006 at 09:07:18AM +0200, Bas Wijnen wrote: On Wed, Jun 28, 2006 at 01:12:10PM +1000, Craig Small wrote: On Wed, Jun 28, 2006 at 11:02:01AM +0900, Charles Plessy wrote: It appears that, although the COPYING file and the website claim that TreeView X is GPL, all the source files have Library GPL headers (except the Nexus ones of course). Technically, what is the licence of TreeView X? LGPL, or GPL? Both or neither, anyhow its a problem. No, it's not. The LGPL specifies that the user may redistribute the software under the LGPL, or the GPL. So you (as packager) can simply choose to do your redistribution under the terms of the GPL. It's a good idea to mention this in the copyright file, of course. Ah yes, I wasn't aware of clause 3 of the LGPL: 3. You may opt to apply the terms of the ordinary GNU General Public License instead of this License to a given copy of the Library. Generally speaking, it could be a problem with the headers saying one thing and the COPYING file saying another, but not in this case. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: COPYING says GPL, but all headers say LGPL
I have one last question: the package uses autoconf. Do I have to mention every autoconf-related file and remind their copyright holder and licence in the copyright file of the Debian package? I am a bit reluctant to do this. Too much information kills information. I don't for any of my autoconfed packages. The output files have do whatever type of license, as does most output files of this type. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: COPYING says GPL, but all headers say LGPL
On Wed, Jun 28, 2006 at 06:13:44PM -0700, Tyler MacDonald wrote: This makes me wonder about ppport.h in perl-XS packages; by default, the generated ppport.h comes with a note saying it's licensed under the same terms as perl itself. However, if you run perl ppport.h --strip, all documentation *and* licensing information was removed. My feeling is that if the authors of Devel::PPPort allow the license to be removed (while still leaving other comments intact in the header file), that means that they really dont care how the released file is licensed. They really should state that if it is their intention. A lot of those builder type programs do it, such as auto* bison and dh-make (I changed it after someone pointed out the problem with not having the specific exception). - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: About the libraries shipped with the sources of a package.
On Tue, Jun 27, 2006 at 03:13:41PM +0900, Charles Plessy wrote: I am packaging Treeview X (GPL) and the upstream tarball contains also the TreeLib (LGPL) and the Nexus library (GPL). For the moment, Treeview X is statically linked to them, just as what is done when building by ./configure make. OK, checking on what this was, I think I've read too much about biology and done my head in. The binary package does not contain the libraries per se, but bits and pieces of them in the binary program. If I understand correctly, I will have to mention their licences in the copyright file. It's statically linked, wether or not this is a good idea depends on if the library could be used elsewhere, it appears to use wxwidgets which are already packaged and you should use those ones. Is adding something like this appropriate? Treeview X is statically linked to libfoo, and libbar. Libfoo (c) MrFoo 1999 is distributed under the terms of the GPL, and libbar (c) MrBar 1998 is distributed under the terms of the LGPL (adding extract and link similar to what is already done for the GPL in this file) If you ship with them, then yes. Do I need separate statements for the statically linked binary file and the full sources of the libraries in the source package? All your statements must cover all the software you ship. How you arrange them is not that important as long as it makes sense. I also wanted to know wether these libraries were already distributed in other source packages, but packages.debian.org did not find anything. But is it searching the sources packages as well ? treelib itself sounds just like a bunch of things that treeviewX needs, so I suspect not. Nexus appears to be some sort of generic format, eg biococoa can read NEXUS files, whatever the hell they are. Lastly, I am wondering wether packaging the libraries separately would be useful if Treeview X would be the only program to use them (in particular, I know that packaging libraries are not recommended to beginners in the art of packaging). Nexus class libraries are found at http://hydrodictyon.eeb.uconn.edu/ncl/ I would strongly reccomend that these were packaged separately and that you use the ones directly from the author's site. Now we have a clash of names, as treelib could imply it has something to do with treeview (eg it is a library for treeview). However, wxwidgets has a treeview (it is the directory tree thing for things like viewing a filesystem). Which one is it? While it would be quicker and simpler to just use what it comes with and go nuts and statically compile the lot, it is not the best solution. Keep specific libraries within the package, anything sourced elsewhere should use the original source and be separate. We get a better Debian distribution in the end. A small example, what happens if someone wants to package the python bindings for Nexus? If the library is locked into treeviewx then we have duplicates. If the library is separate, we can just have the bindings depend on the library in the usual Debian fashion. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: COPYING says GPL, but all headers say LGPL
On Wed, Jun 28, 2006 at 11:02:01AM +0900, Charles Plessy wrote: It appears that, although the COPYING file and the website claim that TreeView X is GPL, all the source files have Library GPL headers (except the Nexus ones of course). Technically, what is the licence of TreeView X? LGPL, or GPL? Both or neither, anyhow its a problem. I would email the upstream author and ask them that you can see both licenses and which one is he/she/they using? procps had something like that and I got my clarification. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: debian/rules::dh_* comments as rejection criteria (Was: Re: A list of common gotchas in Debian packaging)
On Fri, May 05, 2006 at 01:20:33PM -0400, Joey Hess wrote: It seems silly to me that the ftpmasters would take especial umbrage to A while not caring about B, and while probably not checking for C even though it is nearly identical to A in effect. I think the ftpmasters should have much better things to do than waste time about this whole issue. Does it really matter if these commented lines are there? Some people don't like them; I don't like emacs. It doesn't mean I'd go around rejecting packages that used that editor. And yes, I believe this whole issue is as trivial and pointless as an editor war. There are packages that need more work, plenty of packages! Why are we even bothering about this at all? - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: PDF files and dh_compress
On Sat, May 06, 2006 at 11:41:17PM -0400, Yaroslav Halchenko wrote: So is there a recommendation anywhere in dev ref or deb policy regarding the PDF files? Shouldn't it be recommended (withing dev ref or deb policy) to keep PDFs not compressed with gzip on top (at least in -doc packages)? Obviousely dh_compress doesn't bother checking if there is a good reason to compress the file (like some threshold gain, after which file has to be compressed). I doubt that it is worth implementing, but I think it should at least take care about not compressing pdf's in -doc packages. What do you think? Fix the policy not the tool. If there is a good reason not to compress pdf files, then it should be put into the policy as an exception. Currently if dh_compress did not compress pdf files, it breaks 12.3 of policy. There may be very good technical reasons for not compress pdf's, that's fine, but we should not be fixing this just for one set of packaging tools. is worth to provide linda/lintian warnings about twice compressed files or at least compressed pdfs in -doc packages. Once policy is changed, yes. But not before. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: PDF files and dh_compress
On Sun, May 07, 2006 at 11:58:42PM -0400, Yaroslav Halchenko wrote: Of cause we can call PDFs as text documentation, but often with the same success as png's with text in them. I bet policy intended to address regular textual (ASCII) files with documentation. The text documentation is significant. We're only assuming we know what they were trying to describe here. Ironically .zip is among them (I believe PDF internally uses zip compression, doesn't it?) Whatever the compression is, it is not too good. I'm getting about 10-30% compression on pdfs. Random text is just under 50% for me. Once policy is changed, yes. But not before. So as to me, there is no real need in fixing the policy, but rather this question has to be addressed in dev ref, or in packaging practices. ? It needs to be clarified somewhere. Is a 10-30% compression ratio enough to make it worthwhile? It is not debhelper's place to decide this but in our documentation. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Non-Debian packaging practice
On Sun, Mar 12, 2006 at 08:47:28PM -0500, Joe Smith wrote: Russ Allbery [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED] Joe Smith [EMAIL PROTECTED] writes: So I must ask why do people dislike the autotools? Are there really problems that outweigh the benefits of being able to compile the program on strange architectures with little difficulty? Remember that the autotools let one compile the program in the same way on a GNU/Linux system, Cygwin, BSD, and a large number of commercial unixes. Problem #1 is Bad Documentation. There is also a large amount of indirection involved to get things to happen just the way you want them to. A simple program is pretty simple. However I had a lot of difficultly getting the thing to, say, generate a file nicely. It is now doing it but it sure took a long time to work out. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Non-Debian packaging practice
On Fri, Mar 10, 2006 at 10:35:22PM -, StealthMonger wrote: Debian discourages creating Debian-native packages: This type of packaging is only appropriate for the debian-specific packages, which will never be useful in another distribution. [1] But creating it for other distributions requires some knowledge of what those other distributions expect of a package. I am/was upstream for a few packages that are used by multiple distributions. They're not too picky usually about their needs as generally the distribution has to fit in with the upstream, but there would be some points: - Have a clear copyright and license statement, especially if there are multiple authors and/or you don't use one of the common ones - Make sure your dependencies are clearly defined so that the author knows what to look for, if you use auto* make sure you check for those libraries. If you need a specific version, make sure you check that too - Make sure the thing builds twice or can do build, clean, build Seriously, there is stuff out there that won't! - When it comes for installing, allow the installer to put the files where they want. A DESTDIR or prefix variable helps, but so does those not-so-obviously-named things that auto* provide. The current interest here is primarily in packages consisting of shell scripts, as opposed to compilable code, but presumably the question arises in either case. For shell scripts, if you use just /bin/sh don't use bashisms, use checkbashisms program to err, check for bashisms. In general, writing something that works *for me* is easy, writing something that works *for everyone* is a lot harder. You'll probably not hear from package maintainers that much, over 5 years maintaining one thing and I've heard from some a handful of times. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFS: plans -- web calendar
On Mon, Sep 12, 2005 at 10:36:39AM +1000, Ben Finney wrote: On 12-Sep-2005, Carlos Parra Camargo wrote: IMO, the description should be about the application, it shouldn't be a comparative. Agreed. I'd say rather that enough information about the application should be included so that the reader can make their own comparison. You can do both. If your package has feature flaming unicycles while package foo does not, just mention that feature, but no need to mention that foo doesn't have it. Putting it another way. You are a user and you want a web calendar thingy? Which package satisfies your requirements better? Obviously there is only so much a package description can do. - Craig (who longs for the day one of these things is buildable AND integrates into OE) -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: A question regarding unofficial packages (for Xen)
On Wed, Jul 13, 2005 at 03:16:45AM -0500, Yvette Chanco wrote: Apologies if the answer to this is obvious, or if this is the wrong list. My question involves Xen which has been maintained through 2.0.5-3 (in experimental) by Adam Heath. I am not in any way trying to highjack this package - I was very happy with the original. However, since the release of Xen 2.0.6 (May 28th) there have been a few posts to the xen-users list suggesting that Adam is too busy to deal with this (including somebody who spoke with him on IRC), as well as calls to get a team of maintainers together. There have been some volunteers (as well as people, myself included, providing temporary, unofficial debs based on Adam's work) but no Debian Developers. It's really going to come down to what Adam thinks he can do. If he believes he needs help then you could suggest that he sponsors you (or a bunch of you). I do know there is a lot of interest in Xen. I contacted some people on the Xen list to see if we could work together. I tried to reach Adam (although his lack of response could Adam is [EMAIL PROTECTED], did you try that? The only things I could think to do were keep working on my packages to make them clean and conform with policy, and post to this list to ask for advice. Does anybody have thoughts on what the best course of action is? You really need to try to get hold of Adam. There may be a very good reason why the latest version is not been uploaded; sometimes the reason is not obvious if you don't know how Debian packages interact. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [RFH] Dealing with HTML templates in Web Software
On Thu, Apr 28, 2005 at 02:48:59PM +0200, Alexis Sukrieh wrote: So on the first hand, it would be decent to put templates files under /etc/bugzilla, in order not to break user's customizations ; and on the other hand, major upgrades could lead to dangerous user-unfriendly merges. What would you advice me to do? What is the best location for Web applications template files? Is that issue already discussed somewhere so I can read previous thoughts about that? You could try to do it like how request tracker does and have two directory trees. The first one is the one as the package maintainer fiddles with. The second is where the admin puts their modified files. The magic comes with how you handle the two trees. Of course a well-designed system will make sure the templates are just templates (eg formatting with no code at all) so there is not that many changes. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFS: phpicalendar -- clean, logical iCal/vCalendar web interface
On Mon, Apr 11, 2005 at 10:56:16PM -0700, [EMAIL PROTECTED] wrote: Both Chad I really look forward to making this package part of Debian - please consider sponsoring it : ) What's happened to the website for the project? It's still down. That's a shame because I'm looking for something that does ical, preferbly two-way too. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: upgrading database tables and data
On Sun, Apr 10, 2005 at 12:29:29PM -0600, David Everly wrote: Does anyone know of a package that I can look at that has a mechanism to run a series of certain SQL scripts based on upgrading from a certain version to a certain version? For instance, when upgrading from 1.1-1 to 1.1-2, I want to delete a row from a table, when upgrading from 1.1-2 to 1.2-1, I want to alter a table, and when upgrading from 1.1-1 to 1.2-1, I want to do both of these actions (first the delete, then the alter). I do this in the JFFNMS package. If you want to see how I did it make sure you get one that uses the compare-versions features and not the old way with a switch statement. I do ask the admin if they want this done automatically. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFS: tex4ht -- LaTeX and TeX for Hypertext (HTML)
On Wed, Apr 06, 2005 at 04:18:06PM +0530, Kapil Hari Paranjape wrote: I am looking for a sponsor who will help me to upload tex4ht. Hello Kapil, I use LaTeX and even general some webpages with it (using hyperlatex though). If you haven't got a sponsor then I'll do it. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Help about license
On Mon, Feb 28, 2005 at 07:18:38PM +0100, Frank Küster wrote: After this there should be either a copy of the GPL, or a reference where it can be found. On Debian, it often looks as what I paste below, but for upstream this is obviously not possible. Also note that it is *not* sufficient to put a copy of the GPL in the tarball. Instead, there *must* be a statement that the software is intended to be covered by that license! Some people may consider that to be something theoretical and really if the GPL file is there, all is well? Perhaps a real-life example could illustrate why that assumption is a bad idea. There was a program that allowed group chat, consider it a cut-down version of a irc daemon, for hamradio operators. A quick check and you would think it was licensed under the GPL, no problem to go into main. However a more throurough check showed that while some parts of it were licensed under the GPL, parts were not. In fact there were about 4 types of licenses, including some that conflicted. To make matters worse, some of the files were written by others but had no license at all. In effect, the software could not be legally distributed. Probably any one of the 20 developers could object. As far as Debian or any other distributor was concerned, it was a complete write-off. The real shame is I don't think that any of the developers wanted it that way. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Some questions
On Tue, Mar 01, 2005 at 04:34:44PM +0100, Lars Roland wrote: 2) When you update the database software from, say, version 1.0 to 1.1 then it would be preferable to NOT generate all the tables again (that is, mysoftware-db 1.1 should just alter or extend the tables that mysoftware-db 1.0 created). Can I somehow create a package that depends on earlier versions of itself ? - or is there some other scheme that is used when you are in a situation where software depends on earlier versions of itself ?. I have that problem with JFFNMS. There are two seaprate issues. On a new install, create the database On an upgrade, upgrade database schema. First thing is to ask the admin, they might not like you playing with the database! I do all this magic in the postinst script. A new install will be called as postinst configure An upgrade will be called as postinst configure (last version) So you do a test like [ -z $2 ] to see if it is a new install. If so I have a mysql or postgresql dump and I just pipe that into the database client. Now the upgrades are trickier. There are a set of files for each version, called jffnms-(oldver)-(newver).mysql. It doesn't have to be multiple files for each version, but there definitely should be a file per version change, unless there is no DB changes. I then start at the oldest version (0.7.2) and use dpkg --compare-versions $2 lt 0.7.3-1 EG, was the last configured version ($2) less than 0.7.3-1, which means it was 0.7.2-something I then call a function upgrading the database from 0.7.2-0.7.3 I keep doing that for each step there are database changes, so there are 8 of these in postinst. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFS: gaim-themes : Smiley themes collection for gaim
On Thu, Mar 03, 2005 at 07:40:19PM +0100, Martin Braure de Calignon wrote: I am looking for a sponsor for this packages. It exists in ITP list (#297961). It can be downloaded at : http://www.enseirb.fr/~braurede/deb_dev/gaim-themes/ I can sponsor this if you like. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFS: mocp -- music on console player
On Fri, Jan 14, 2005 at 12:15:08PM +0100, Michal Jeczalik Jr wrote: I have already got a sponsor, sorry. No need to be sorry, glad you have one. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFS: mocp -- music on console player
On Mon, Jan 10, 2005 at 03:06:35PM +0100, Michal Jeczalik Jr wrote: You can find the package on: http://www.orora.org/michalj/debian/mocp/ I have been using it for half a year. Also a few friends of mine use it. It has useful 'detach' option and supports themes. mocp-2.1.4 is a stable version (2.2.x is development). Now I am looking for a sponsor. Feedback would also be appreciated. It's very nice and I can see that the latest appears to have all the Debian kinks sorted out now. I can sponsor it if you like and you haven't already got a sponsor. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: maint-guide update (chapt 6-9, appendix)
On Tue, Jan 11, 2005 at 01:44:51AM +0100, Osamu Aoki wrote: I need critical review again. New maint-guide should show up on the web soon as the *11* January 2005 version at http://www.debian.org/doc/maint-guide/ I am thinking to upload around 23rd based on this version. Joy or anyone, if you do not like any contents here, please tell me so. I will accommodate your request. You know it would of been a good idea to at the very least show the two URLs of the before and after document. A colour-coded diff (ala viewcvs and friends) would also of helped. I'm not sure which one you are talking about now and it would of been nice to see considering how... vocal I am sometimes about the Debian documents. - Craig -- Craig Small GnuPG:1C1B D893 1418 2AF4 45EE 95CB C76C E5AC 12CA DFA5 Eye-Net Consulting http://www.enc.com.au/ MIEE Debian developer csmall at : enc.com.au ieee.org debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]