[Fink-devel] Faster installation of fink systems - a proposal
Dear Fink Developers, I think it is far too tedious to manually select all packages everytime fink. Also, I prefer installings this in one go, rather than continously finding a package that I may need (which could happen at a bad occasions, e.g. when Internet it is not availabe, while at a conference, in a train...). Therefore a suggestion: Why not add a virtual package called stable- all or just stable-base or similar that installs ALL fink packages. I suggest having several virtual packages in fact: - stable-minimal - stable-medium-with-x11 - stable-medium-without-x11 - stable-medium-gnome - stable-medium-kde - stable-maximal etc. I myself would do a nightly fink install stable-maximal once and for all. Best Wishes, Johan Glimming --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Faster installation of fink systems - a proposal
On May 22, 2005, at 2:14 PM, Johan Glimming wrote: Dear Fink Developers, I think it is far too tedious to manually select all packages everytime fink. Also, I prefer installings this in one go, rather than continously finding a package that I may need (which could happen at a bad occasions, e.g. when Internet it is not availabe, while at a conference, in a train...). Therefore a suggestion: Why not add a virtual package (Technically, what you're proposing isn't a virtual package. It's a bundle package--meaning that it depends on many other packages but doesn't install any files of its own.) called stable-all or just stable-base or similar that installs ALL fink packages. (It is completely impossible to install the entire Fink distribution because some packages conflict with each other) I suggest having several virtual packages in fact: - stable-minimal - stable-medium-with-x11 - stable-medium-without-x11 - stable-medium-gnome Is this basically like installing bundle-gnome + bundle-gnome-office (and possibly all of the other GNOME-based apps)? - stable-medium-kde This one's mostly covered by bundle-kde(-ssl) plus all of the other KDE-based apps that aren't part of the main distribution. - stable-maximal etc. I myself would do a nightly fink install stable-maximal once and for all. Umm...my reading of the above makes me think this is more troublesome than helpful. To force a bundle to update dependent packages requires it to depend on explicit versions of the dependencies, and therefore when any of these gets changed the dependency in the bundle has to be changed, even for minor updates. On the other hand, the way the existing bundle-* package work is much simpler. You have a bundle, which installs whatever the latest versions of the dependencies in the particular tree (stable/unstable) happen to be. If there's an update, the usual fink selfupdate; fink update-all sequence brings you the newer version without messing with the bundle. People who install the bundle later get the later versions automatically still. In any case it's not too hard to set up bundle packages. You could do it yourself on your own system and if it works then submit what you've done. Best Wishes, Johan Glimming -- Alexander Hansen Fink Documentarian [Day Job] Levitated Dipole Experiment http://psfcwww2.psfc.mit.edu/ldx/ --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] Re: fink in unstable is unstable
Martin, I don't think it is wise to remove the BuildConflicts for freeglut and glut. The current freeglut and glut packaging has freeglut built as libfreeglut rather than libglut to avoid the binary incompatibility of the libraries. However to avoid requiring any package that wants to use freeglut instead of glut requiring changes to its configure and Makefiles, I have symlinks in the freeglut package to the glut based header and library names...that is in /sw/include/GL we have... lrwxr-xr-x 1 root admin 10 May 20 08:50 glut.h - freeglut.h lrwxr-xr-x 1 root admin 14 May 20 08:50 glut_ext.h - freeglut_ext.h lrwxr-xr-x 1 root admin 14 May 20 08:50 glut_std.h - freeglut_std.h and in /sw/lib we have... lrwxr-xr-x 1 root admin 13 May 20 08:50 libglut.a - libfreeglut.a lrwxr-xr-x 1 root admin 19 May 20 08:50 libglut.dylib - libfreeglut.3.dylib I could do a careful check to make sure the freeglut build isn't leaky and won't build against the wrong headers and libraries but it would be best if fink knew how to properly cope with BuildConflicts. Jack --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Faster installation of fink systems - a proposal
Umm...my reading of the above makes me think this is more troublesome than helpful. Hi Again, Contrary to open source culture I cannot do this myself. But, I disagree that this is a bad idea. :-) The very idea of having this sort of bundle packages is * like a Linux distribution, Fink could not come with single packages, but also with tested combinations useful for many systems. The idea is PRECISELY to avoid dependency problems, and make a selection of useful packages based on best practice. Installing a stable bundle will give no compilation errors or dependency problems, in the best of all worlds. * like a Linux distribution, most users have no time to select and learn exactly what are the dependencies of certain packages etc. For us, it is much easier to just install a pre-selected bunch of packages known to work. For such bundles, there could be maintainers: preferrably system groups at some univerrsity etc could absorb such a responsibility. For this I suggest to open a Fink Foundation which awards such system administrators with a some funds to replace portion of their salary. Maybe Apple will donate? :-) This could become a network of university system admins or such like. Best Wishes, Johan Glimming --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Re: fink in unstable is unstable
[EMAIL PROTECTED] (Jack Howarth) wrote: [snip] but it would be best if fink knew how to properly cope with BuildConflicts. Sadly, fink doesn't know how to properly cope with BuildConflicts, nor is this likely to change without a complete rewrite of fink's dependency engine. (This rewrite is something whose need we are well aware of, but nobody currently working on fink has the time to do it.) I hope you are willing to work within the limitations of fink as it currently exists. -- Dave --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Faster installation of fink systems - a proposal
On May 23, 2005, at 8:52 AM, Johan Glimming wrote: Umm...my reading of the above makes me think this is more troublesome than helpful. Hi Again, Contrary to open source culture I cannot do this myself. But, I disagree that this is a bad idea. :-) For clarity--the only part I thought might be annoying would be if there were need for frequent updates of the placeholders. If these are stable snapshots, then I withdraw that. The very idea of having this sort of bundle packages is * like a Linux distribution, Fink could not come with single packages, but also with tested combinations useful for many systems. The idea is PRECISELY to avoid dependency problems, and make a selection of useful packages based on best practice. Installing a stable bundle will give no compilation errors or dependency problems, in the best of all worlds. Agreed. * like a Linux distribution, most users have no time to select and learn exactly what are the dependencies of certain packages etc. For us, it is much easier to just install a pre-selected bunch of packages known to work. Also agreed. This was the philosophy behind the existing bundles. For such bundles, there could be maintainers: preferrably system groups at some univerrsity etc could absorb such a responsibility. There would be maintainers in any case. The question that comes up is the chain of command--who is responsible for which package in the nest of dependencies. For this I suggest to open a Fink Foundation which awards such system administrators with a some funds to replace portion of their salary. Work _is_ underway to form a Fink nonprofit organization. Maybe Apple will donate? :-) Fat chance. ;-) This could become a network of university system admins or such like. Best Wishes, Johan Glimming -- Alexander Hansen Fink Documentarian [Day Job] Levitated Dipole Experiment http://psfcwww2.psfc.mit.edu/ldx/ --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] Re: Faster installation of fink systems - a proposal
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I know of something used in Debian GNU+Linux that sounds is somewhat (it is a little of a stretch) what you are talking about (called taskselect). If lets you decide what you want to do with the computer, like being a web server or a dial-up server. Is that more of what you would want? Johan Glimming wrote: Dear Fink Developers, I think it is far too tedious to manually select all packages everytime fink. Also, I prefer installings this in one go, rather than continously finding a package that I may need (which could happen at a bad occasions, e.g. when Internet it is not availabe, while at a conference, in a train...). Therefore a suggestion: Why not add a virtual package called stable- all or just stable-base or similar that installs ALL fink packages. I suggest having several virtual packages in fact: - stable-minimal - stable-medium-with-x11 - stable-medium-without-x11 - stable-medium-gnome - stable-medium-kde - stable-maximal etc. I myself would do a nightly fink install stable-maximal once and for all. Best Wishes, Johan Glimming - -- Benn Newman E-mail: [EMAIL PROTECTED] | Maintainer of the Fink Developer Map Jabber: [EMAIL PROTECTED] | http://finkproject.org/map/ Please avoid sending Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (Darwin) iEYEARECAAYFAkKSIQwACgkQFE65lPR8xrESqQCeOdeVT+K75cr0BY9Qz6bM5K6L 76gAnRNlFtAZS6R6cgtqqYWnS6BwiWtE =EO4m -END PGP SIGNATURE- --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Re: Faster installation of fink systems - a proposal
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On May 23, 2005, at 2:29 PM, Benn Newman wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I know of something used in Debian GNU+Linux that sounds is somewhat (it is a little of a stretch) what you are talking about (called taskselect). If lets you decide what you want to do with the computer, like being a web server or a dial-up server. Is that more of what you would want? That is exactly what he wants. - -chris zubrzycki - - -- PGP public key: http://homepage.mac.com/beren/publickey.txt ID: 0xA2ABC070 Fprint: 26B0 BA6B A409 FA83 42B3 1688 FBF9 8232 A2AB C070 A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail? -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (Darwin) Comment: Please sign reply-http://www.gnupg.org iEYEARECAAYFAkKSS+UACgkQ+/mCMqKrwHAxSACgio6P9pYWTFImJwmqBZa77NRm nfsAoNan5+575rsv9jnXAh7tth4zhP2p =kK/h -END PGP SIGNATURE- --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] revised freeglut/glut submitted
I have submitted revised version of the glut (3.7-25) and the freeglut (2.2.0-3) packages for fink unstable into the tracking system. The only changes made were the removal of the Build-Conflicts line which currently isn't handled properly under fink. This change will prevent fink from deinstalling glut if freeglut was installed and freeglut if glut was installed allowing the builds that require them to succeed. As I mention in the tracker reports, both packages pass fink validate for their info files and resulting debs. Jack --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] new pymol-py package
I have resubmitted the pymol-py package as new package on the tracker since I was having no luck whatsoever in getting this into unstable. Once again, this pymol-py package replaces the current pymol package and allows builds against python2.2, python2.3 or python2.4 with the resulting pymol-py22, pymol-py23 and pymol-py24 packages being built. As before, the info and debs all pass fink validate. I would REALLY appreciate some feedback here if there is a reason why this package is being delayed. Jack --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] new pymol-py package
Which tree should they go into? I tested them on 10.3 and it works fine. If I know which trees and if there are no other objections I will commit them. Is the glut/freeglut issue that was discussed a bit ago solved? - Koen. On May 23, 2005, at 7:41 PM, Jack Howarth wrote: I have resubmitted the pymol-py package as new package on the tracker since I was having no luck whatsoever in getting this into unstable. Once again, this pymol-py package replaces the current pymol package and allows builds against python2.2, python2.3 or python2.4 with the resulting pymol-py22, pymol-py23 and pymol-py24 packages being built. As before, the info and debs all pass fink validate. I would REALLY appreciate some feedback here if there is a reason why this package is being delayed. Jack --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] new pymol-py package
Koen, The pymol-py package should go in both the 10.3 and 10.4-transitional trees (unstable). Thanks in advance. Jack --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] new pymol-py package
On May 23, 2005, at 10:25 PM, Jack Howarth wrote: While you are adding pymol-py, please update the glut and freeglut packages in the 10.3 and 10.4-transitional unstable trees as well. Ok done. I assume they build and validate fine, I did not check that. - Koen. --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] new pymol-py package
On Mon, May 23, 2005 at 10:25:54PM -0400, Jack Howarth wrote: Koen, While you are adding pymol-py, please update the glut and freeglut packages in the 10.3 and 10.4-transitional unstable trees as well. http://sourceforge.net/tracker/index.php?func=detailaid=1207413group_id=17203atid=414256 http://sourceforge.net/tracker/index.php?func=detailaid=1207415group_id=17203atid=414256 These are identical to the current versions of the packages except the Build-Conflicts field has been removed from each of the info file. Currently fink is unable to properly handle Build-Conflicts so that if you build pymol-py which Build-Depends on freeglut with glut installed, fink will build, install and then deinstall freeglut in favor of glut just before it tries to build pymol-py. Removal of the Build-Conflicts eliminates this glitch which has been discussed in the thread entitled fink in unstable is unstable. The goal is clean build first and foremost; *then* worry about avoiding buildlock deadlock. If there's leakage from an installed pkg into the build, please do specify the BuildConflicts...that's why we created the field no? dan -- Daniel Macks [EMAIL PROTECTED] http://www.netspace.org/~dmacks --- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_id=7412alloc_id=16344op=click ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-devel
