[Freedos-kernel] Broken VERSION= in 2035 etc. - too many kernels
Hi, after a while we found that only the Lucho / Arkady kernel has working VERSION=, which reminds me that we have too many versions: - outdated 2035 - Jeremy has added an unknown amount of patches to CVS HEAD 2035a, but I have no idea if / how well those patches got reviewed and which bugs they fix - Tom has added only the best few lines of the Lucho / Arkady patches, those which OBVIOUSLY fix bugs, but will have missed several bugfixes which were hidden between optimizations - Lucho / Arkady produced a big amount of patches which mainly contain optimizations (hopefully not too compiler specific), but also important bugfixes, as well as hard to justify variable renamings, indendation changes, macro changes, etc.!? So I suggest: Jeremy, please create a list of improvements / changes between 2035 and newest CVS kernel, and a way to download all patches separately (or in one big tgz/zip), along with short descriptions of them. Tom, are you planning to publish your kernel variant? In either case, could you help us to discuss the feature / bugfix differences between your and Jeremies kernel? I think it should be possible to sort the recent patches into important useful patches vs. tuning / optimization / experimental patches, to create a combined Tom's and Jeremies best pick of all recent patches kernel on which all parties could settle for a stable 2035b official release for freedos.sourceforge.net ... Finally Lucho and Arkady can comment on that by describing the advantages of the patches which did not made it into 2035b, and by pointing out which bugfixes should be added before 2035b would be published officially on SF.net ... In either case a list of patches, in a file, not spread over dozens of mails, would greatly improve the overview over the available improvements, dependency between patches, give a guide which helps to find relevant 2035 / 2035b differences in processing bug reports for either (!) of the two, give information which patches are considered to be stable and which are more obscure and therefore not yet reviewed well, and so on. There must be a lot of useful stuff in all those patches, but we might lose overview and miss all the goodies if the patch authors and kernel maintainers do not start to help mortal programmers like me to catch up with development and figure out what ingredients will be in 2035b and why 2035b will be quite a bit better than 2035. So please take some time and start with that effort - as patch authors you should be the best source of information and help towards a general consent about 2035b. Finally we will have good old 2035 Tom / Jeremy / list reviewed stuff which will become 2035b Arkady / Lucho / others (Eduardo, Tom, Bart...) patches waiting for 2035c or in other words only THREE different kernels, not four-or-more. Hopefully this is no too far fetched vision / suggestion. Thanks for reading :). Eric. --- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php ___ Freedos-kernel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-kernel
Re: [Freedos-kernel] Broken VERSION= in 2035 etc. - too many kernels
Hello Eric, - Tom has added only the best few lines of the Lucho / Arkady patches, those which OBVIOUSLY fix bugs, but will have missed several bugfixes which were hidden between optimizations that might have happened. But for sure I missed some bew bugs hidden between optimizations. - Lucho / Arkady produced a big amount of patches which mainly contain optimizations (hopefully not too compiler specific), but also important bugfixes, which ones ? Tom, are you planning to publish your kernel variant? www.drivesnapshot.de/freedos/ke2035bt.zip as of today In either case, could you help us to discuss the feature / bugfix differences between your and Jeremies kernel? bugfixes - yes. There must be a lot of useful stuff in all those patches, but we might lose overview and miss all the goodies if the patch authors and kernel maintainers do not start to help mortal programmers like me to catch up with development why should kernel programmers care if mortal programmers can't follow? you won't follow anyway. tom --- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php ___ Freedos-kernel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-kernel
Re: [Freedos-kernel] Broken VERSION= in 2035 etc. - too many kernels
Eric Auer wrote: Hi, after a while we found that only the Lucho / Arkady kernel has working VERSION=, ... If it is broke in stable and fixed in dev, then it should be fixed in head soon. Please people, the kernels are not Lucho/Arkady vs Tom/..., they are stable and development (unstable). ... which reminds me that we have too many versions: There is exactly one (1) kernel; in cvs there are two (2) branches and many versions (from the original Pat release up to current cvs versions). There are are also forks I'm sure, Tom for one has one of these, but as he makes available his kernel, any bugfixes in his should show up in sourceforge one, time permitting. - outdated 2035 - Jeremy has added an unknown amount of patches to CVS HEAD 2035a, but I have no idea if / how well those patches got reviewed and which bugs they fix See history.txt in the docs directory. Also see http://www.fdos.org/kernel/head2unstable.diff if you are curious of the difference (minus new files) between stable dev. - Tom has added only the best few lines of the Lucho / Arkady patches, those which OBVIOUSLY fix bugs, but will have missed several bugfixes which were hidden between optimizations - Lucho / Arkady produced a big amount of patches which mainly contain optimizations (hopefully not too compiler specific), but also important bugfixes, as well as hard to justify variable renamings, indendation changes, macro changes, etc.!? Please note, you can easily ignore identation changes with a few diff options, that is largely why I don't care if they are made or not. So I suggest: Jeremy, please create a list of improvements / changes between 2035 and newest CVS kernel, and a way to download all patches separately (or in one big tgz/zip), along with short descriptions of them. As best I know (though haven't double checked through my email) all patches are integrated in dev branch, so its history.txt should tell you everything that has changed (recent additions are mostly organized by person then file), and a cvs diff can get you the patch. Tom, are you planning to publish your kernel variant? In either case, Tom has said on several occasions his kernel is available upon request. [Note: I see looking ahead he has provided a URL.] could you help us to discuss the feature / bugfix differences between your and Jeremies kernel? I think it should be possible to sort the recent patches into important useful patches vs. tuning / optimization / experimental patches, to create a combined Tom's and Jeremies best pick of all recent patches kernel on which all parties could settle for a stable 2035b official release for freedos.sourceforge.net ... what happened to 2035a? :-) Finally Lucho and Arkady can comment on that by describing the advantages of the patches which did not made it into 2035b, and by pointing out I'm still reviewing the patches in the dev branch and merging into stable. Any that make it in that others strongly disagree with can always be reverted (such as the recent thread started by Bart). I'm trying to be conservative in what goes into stable, but limiting the divergence at the same time. which bugfixes should be added before 2035b would be published officially on SF.net ... In either case a list of patches, in a file, not spread over dozens of mails, would greatly improve the overview over the available improvements, dependency between patches, give a guide which helps to find relevant 2035 / 2035b differences in processing bug reports for either (!) of the two, give information which patches are considered to be stable and which are more obscure and therefore not yet reviewed well, and so on. ... Eric. comments/reviews/discussions (not arguments) on patches/bugs/regressions are of course always welcome. Please be patient with me though, I have but a few hours a week (squashed between sleep and my jobs) to work on projects. Jeremy :-) --- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php ___ Freedos-kernel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-kernel
Re: [Freedos-kernel] Broken VERSION= in 2035 etc. - too many kernels
Hi! 18--2004 16:50 [EMAIL PROTECTED] (Eric Auer) wrote to [EMAIL PROTECTED]: EA Jeremy, please create a list of improvements / changes between 2035 EA and newest CVS kernel, and a way to download all patches separately EA (or in one big tgz/zip), along with short descriptions of them. All my patches (at least, which have seen effect, like changed MENU behavior) was described. and this list should be seen through history.txt. --- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php ___ Freedos-kernel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-kernel
Re: [Freedos-kernel] Broken VERSION= in 2035 etc. - too many kernels
Hi! 18--2004 20:32 [EMAIL PROTECTED] (tom ehlert) wrote to Eric Auer [EMAIL PROTECTED]: - Lucho / Arkady produced a big amount of patches which mainly contain optimizations (hopefully not too compiler specific), but also important bugfixes, te which ones ? See the subject for one fixed bug. BTW, this bugfix by itself is very easy, only two lines to change os_setver_m*or instead os_m*or. --- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php ___ Freedos-kernel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-kernel
Re: [Freedos-kernel] Broken VERSION= in 2035 etc. - too many kernels
Hi! 19--2004 02:34 Arkady V.Belousov wrote to [EMAIL PROTECTED]: - Lucho / Arkady produced a big amount of patches which mainly contain optimizations (hopefully not too compiler specific), but also important bugfixes, te which ones ? AVB See the subject for one fixed bug. I mean: one from many. AVB BTW, this bugfix by itself is very AVB easy, only two lines to change os_setver_m*or instead os_m*or. --- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php ___ Freedos-kernel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-kernel