Re: Where is the source code for the versions of emc and axis on bdi 4.30 and 4.38?
On Monday 16 January 2006 22:14, [email protected] wrote:
Where is the source code for the versions of emc and axis that appear on bdi 4.30 and 4.38?
Jeff apt-get source emc Will get you the most recent tarball. Check with viewcvs at SF for the last tag - This will be most recent checkout from which the emc package was compiled from. Regards, Paul. -- From the Klingon book of C: Klingon function calls do not have 'parameters' - they have 'arguments' - and they ALWAYS WIN THEM.
On Tue, Jan 17, 2006 at 11:16:57PM +0000, Paul wrote:
On Monday 16 January 2006 22:14, [email protected] wrote:
Where is the source code for the versions of emc and axis that appear on bdi 4.30 and 4.38?
Jeff
apt-get source emc
Will get you the most recent tarball. Check with viewcvs at SF for the last tag - This will be most recent checkout from which the emc package was compiled from.
Hi. Thanks for these instructions. However, I was still not able to get the source for the versions of emc and axis that appear on bdi 4.38. I was able to get the source for a different version of emc, but I wasn't able to get any source for axis. Before I could get any source, I had to add several lines to /etc/apt/sources.list (manually transcribed, so I may have introduced errors): deb-src http://ftp.debian.org/debian sarge main contrib deb-src http://mirrors.neutron.com/emc-bdi/apt srge updates extras deb-src http://solaris.cs.utt.ro/emc-bdi/apt srge updates extras Then, I had to run 'apt-get update' again. How, 'apt-get source emc' downloads 'emc_1.0-29.tar.gz' and 'emc_1.0-29.dsc'. However, 'dpkg -l emc' reports that the version on bdi 4.38 is actually '1.0-38'. And 'apt-get source axis' (hey, I was going to try the obvious here) reports that it is 'Unable to find a source package for axis'. This message is widely copied again, as I still haven't seen my messages appear on bdi4emc-help, and I know there are some who are only subscribed to emc-users who are interested in the answer as well. Please copy emc-users on reply, or I'll be happy to continue forwarding copies to them. Jeff PS I normally sign my messages with gpg, but just in case the problem with the bdi4emc-help list is the signature, I'm omitting it from this message.
On Tue, Jan 17, 2006 at 06:38:06PM -0600, [email protected] wrote:
[N]ow, 'apt-get source emc' downloads 'emc_1.0-29.tar.gz' and 'emc_1.0-29.dsc'. However, 'dpkg -l emc' reports that the version on bdi 4.38 is actually '1.0-38'.
Following up my earlier message, I tried to buid emc_1.0-29 according to the instructions in the debian apt-howto (http://www.debian.org/doc/manuals/apt-howto/ch-sourcehandling.en.html) As that page suggests, I ran apt-get -b source emc which eventually fails: ... dpkg-buildpackage: host architecture is i386 deb/rules clean /usr/bin/dpkg-buildpackage: line 175: debian/rules: No such file or directory Build command 'cd emc-1.0 && dpkg-buildpackage -b -uc' failed. E: Child process failed The 'rules' file is listed as 'required stuff under debian/' in the Debian New Maintainers Guide, so I can only assume that the source archive that 'apt-get' downloaded has somehow become damaged. Thanks for any pointers you can give me. Jeff
On Wednesday 18 January 2006 00:56, [email protected] wrote:
apt-get -b source emc which eventually fails: ... dpkg-buildpackage: host architecture is i386 deb/rules clean /usr/bin/dpkg-buildpackage: line 175: debian/rules: No such file or directory Build command 'cd emc-1.0 && dpkg-buildpackage -b -uc' failed. E: Child process failed The 'rules' file is listed as 'required stuff under debian/' in the Debian New Maintainers Guide, so I can only assume that the source archive that 'apt-get' downloaded has somehow become damaged.
dpkg-buildpackage will always fail to build the emc-1.0 package. To successfully build the emc-1.0-xx.deb requires (amongst other things), m-a, dh_helper, and a few extra packages from the experimental branch. The emc-modules package is dependant on a specific kernel package version and can only be built along side that one kernel. May be when m-a has been fully debugged and migrated through to "stable" we can look at fixing this minor problem. You can still compile emc by switching to the emc-1.0/src directory, running configure, followed by make as per the usual instructions. This will result in a fully functional set of binaries that are identical to the packaged version. Regards, Paul. -- From the Klingon book of C: Klingon function calls do not have 'parameters' - they have 'arguments' - and they ALWAYS WIN THEM.
On Wed, Jan 18, 2006 at 11:24:33AM +0000, Paul wrote:
dpkg-buildpackage will always fail to build the emc-1.0 package. To successfully build the emc-1.0-xx.deb requires (amongst other things), m-a, dh_helper, and a few extra packages from the experimental branch. The emc-modules package is dependant on a specific kernel package version and can only be built along side that one kernel. May be when m-a has been fully debugged and migrated through to "stable" we can look at fixing this minor problem.
I should not need to remind anyone of what the GNU GPL says, or that anaconda, the Linux kernel, rtai, emc, and axis (among other packages you modify and distribute) are licensed by the GNU GPL, but I will anyway (emphasis mine, text from GPL v2 section 3): The source code for a work means *the preferred form of the work for making modifications to it*. For an executable work, complete source code means all the source code for all modules it contains, *plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable*. The "works" you are distributing are emc_1.0-38_i386.deb, python2.3-axis_1.1+cvs20051221_i386.deb, etc, and so the "complete source" you must supply includes not only the source files you have finally begun to put online, but the "associated interface definition files" (e.g., the required headers), and the "scripts used to control compilation and installation of the executable" (e.g., the contents of the debian/ subdirectory, and any other scripts you use to perform the build) Until it is possible for anyone to download the source for these packages, and rebuild them from that source into an equivalent binary package, the requirements of the GPL have not been satisfied. It is a red herring to talk about whether the build process is experimental, fragile, or the like. Failure to comply with the GPL isn't just a "minor problem", either. When you fail to comply, your rights under the GPL are terminated, and nothing else but the license grants you the right to modify or distribute the programs. Jeff
On Wednesday 18 January 2006 13:40, [email protected] wrote:
Until it is possible for anyone to download the source for these packages, and rebuild them from that source into an equivalent binary package, the requirements of the GPL have not been satisfied.
In accordance with sections 1 and 3.b. of the GPL (Ver.2), the original source tarballs for both axis and emc have been added to the primary repository. These tarballs include *ALL* make files and build scripts to reproduce the binaries. Should anyone require the sources for any of the Debian supplied packages, I will, for a nominal fee, provide on floppy, CD, or DVD, the original tarballs (which is above and beyond the requirement of section 3.c). To quote the final paragraph of Sec. 3: If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code. I will of course, cease and desist forthwith, any redistribution of axis and it's associated sources if you (as the co-author of axis) so desire. Regards, Paul. -- From the Klingon book of C: Klingon function calls do not have 'parameters' - they have 'arguments' - and they ALWAYS WIN THEM.
On Wed, Jan 18, 2006 at 04:43:39PM +0000, Paul wrote:
In accordance with sections 1 and 3.b. of the GPL (Ver.2), the original source tarballs for both axis and emc have been added to the primary repository.
A URL or guidance on setting up sources.list would help here. I looked around on ourproject.org and mirrors.neutron.com and came up empty. I'm not seeing anything new with 'apt-get source', either, even after an 'apt-get update'. [...]
I will of course, cease and desist forthwith, any redistribution of axis and it's associated sources if you (as the co-author of axis) so desire.
I'm not asking you to do that. I'm asking you to fulfill the terms of the GPL when you do redistribute a modified version of AXIS. Jeff
On Thursday 19 January 2006 04:37, [email protected] wrote:
I'm not seeing anything new with 'apt-get source', either, even after an 'apt-get update'.
I am not reponsible for either of the two servers you mentioned in an earlier email - As I said then, it will take a day or two for the cron jobs to update. i.e. 24-48 hours. If the files have not appeared by Monday, a polite reminder will be sent out.
I'm not asking you to do that. I'm asking you to fulfill the terms of the GPL when you do redistribute a modified version of AXIS.
No doubt you will publicly remind the www.thecooltool.com people in an equally public manner of their obligations. Regards, Paul. -- From the Klingon book of C: Klingon function calls do not have 'parameters' - they have 'arguments' - and they ALWAYS WIN THEM.
apt-get source emc cd emc-1.0/src ./configure && make apt-get source python-axis cd python-axis-1.1+cvs.20051221 EMCROOT=../emc-1.0 python setup.py build Both tasks complete without fatal errors. Subject closed. Regards, Paul. -- From the Klingon book of C: Klingon function calls do not have 'parameters' - they have 'arguments' - and they ALWAYS WIN THEM.
On Wednesday 18 January 2006 00:38, [email protected] wrote:
Before I could get any source, I had to add several lines to /etc/apt/sources.list (manually transcribed, so I may have introduced errors): deb-src http://ftp.debian.org/debian sarge main contrib deb-src http://mirrors.neutron.com/emc-bdi/apt sarge updates extras deb-src http://solaris.cs.utt.ro/emc-bdi/apt sarge updates extras Then, I had to run 'apt-get update' again.
Yup, you'll need the deb-src lines (with the typos corrected) before apt-get will download the sources...
How, 'apt-get source emc' downloads 'emc_1.0-29.tar.gz' and 'emc_1.0-29.dsc'. However, 'dpkg -l emc' reports that the version on bdi 4.38 is actually '1.0-38'.
OK.. Looks like the 1.0-38 sources failed to find their way to the mirrors. Have reseeded the primary repository and the updates will appear over the next day or two.
And 'apt-get source axis' (hey, I was going to try the obvious here) reports that it is 'Unable to find a source package for axis'.
Always assumed people interested in the current or latest version of Axis would go straight to the authors or download from http://axis.unpy.net/index.cgi/downloads - Again, added the original tarball to the primary repository, so this will also apear sometime over the next few days. Regards, Paul. -- From the Klingon book of C: Klingon function calls do not have 'parameters' - they have 'arguments' - and they ALWAYS WIN THEM.
participants (2)
-
jepler＠unpythonic.net -
Paul