On Fri, 2005-06-03 at 22:19 -0400, Lamar Owen wrote: > On Friday 03 June 2005 08:48, Johnny Hughes wrote: > > I am a little touchy on this subject ... being that I hear it so often. > > I most assuredly understand this. I reply to your message in this thread > because I find you quite level-headed. > > > I disagree ... the work that RedHat put into RHEL is significant. It is > > a great distro, and worthy to win this award and many others ... but so > > is CentOS. > > > BUT ... they didn't put any work into CentOS. All their work was put > > into RHEL. They are compensated fairly well for that work too, I might > > add. > > Um, I don't if I can agree with that statement. CentOS is built upon the > source RPMs of RHEL with sometimes substantial modifications, right? So, out > of the changelog entries in the various spec files in CentOS what percentage > of persons making changelog entries have an @redhat.com address? I have some > work in CentOS in the PostgreSQL set, but it is quite small compared to the > rest of the work. But Red Hat developers (some of who are upstream > developers paid to work on the upstream package by Red Hat) have put a > substantial amount of nontrivial work into the packages that through the > CentOS team's efforts became CentOS. > > I do understand what you are saying, why you are saying it, and the > distinction you are attempting to make, but it comes across at least to me as > understating the work Red Hat people have done. Let's not overstate Red > Hat's contribution, by all means; but let's ot understate it either. > Let me try to say it another way then, I think my meaning is not clear. Red Hat did their work, not at all for CentOS, but only for RHEL. They release their work for others to use (and they are very good about erring on the side of open source if releasing a specific file is questionable). CentOS uses that released work to build a distro ... and we do so meeting all their requirements (even go further than they require in some instances). So ... CentOS depends upon RHEL to release their code ... but RedHat doesn't do the work FOR the CentOS project ... they do the work for themselves. I am not suggesting that their work is not significant ... quite the opposite is true. It is a huge amount of work. CentOS could not be published as it is without the excellent work done by Red Hat. They released it as open source software and released a set of rules to use it ... which CentOS follows. However, they do not release everything required to build all the SRPMS, nor do they tell you what combinations of programs are required to be installed to get the SRPMS to build correctly. There are many -devel packages that must be built from code to get the SRPMS provided to build. The build hosts have to be prepared in certain ways, with certain specific software installed (some of which is not provided by RedHat at all). They do not provide instructions to do this (nor would I expect them to). > > the items they publish from somewhere else. They are required, > > therefore, to make their source code public. They take that requirement > > seriously, and they do an outstanding job of publishing their source > > code openly. They should be commended for that. I do it every chance I > > get :) > > They are not required to make the source for non-GPL code public (unless the > license requires it; the PostgreSQL BSD license for instance does not require > it), nor are they required to distribute any spec files or special > initscripts they may use. They are not required to distribute source to > anyone but the receipients of the binary code, and they cannot restrict said > recipient's right to redistribute the source under the GPL. But GPL code is > not the majority of the Red Hat dist, is it? Anyone have a count of bytes > under GPL versus other licenses in CentOS? 1005/1494 is GPL/LGPL 146 is BSD I did not do a detailed comparison on the BSD works ... if the BSD products are totally separate, then they don't have to be redistributed. If they use GPL libraries, then the source might need to distributed. > If not, I'll do the math, > tomorrow. So Red Hat could make the job much more difficult by distributing > the source as simple tarballs with no specs, no initscripts, no notes. Like > you say, they have done an outstanding job releasing things they don't have > to release as source in SRPM format, making it not as difficult as it could > be to rebuild a trademark-free build. Not that it is trivial or even easy; > but it could be more difficult. > I disagree. I think the below clause requires that they release their spec file and SRPMS (at least for GPL/LGPL items): "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." That is, to people who they distribute their executables to. Not to the public as a whole. BUT, if someone had a legitimate AS license and they met all the requirements to redistribute the code, then they could redistribute the SRPMS and RPMS they built even if full public access to the SRPMS was removed by RedHat. (Not that Red Hat would remove public access ... but that is an argument that is sometimes heard) > > Using your logic, The only people who would get credit are the > > programmers who write the code for the parent projects. > > Many of whom work for Red Hat. Some of whom could not afford to work fulltime > on their project were it not for Red Hat's employment. > > > Or the Fedora > > Core volunteers who actually package and test probably 95% of the stuff > > that get into RHEL. > > Most of the Fedora Core packagers work for Red Hat, with notable exceptions. > Red Hat just distanced themselves from DIRECT control of Fedora to try and get more volunteers to become part of the project: http://informationweek.com/story/showArticle.jhtml?articleID=16430008 http://news.com.com/Red+Hat+lets+go+of+Fedora+Linux/2100-7344_3-5730931.html?tag=nefd.top http://www.eweek.com/article2/0,1759,1823403,00.asp http://business.newsforge.com/article.pl?sid=05/06/03/1729211&tid=18 > > So no, I don't think that you should vote for RHEL in a place where > > CentOS is also listed > > I agree wholeheartedly with your conclusion, but not with all your reasons. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: <http://lists.centos.org/pipermail/centos/attachments/20050604/c9c38fd5/attachment-0005.sig>