February 2006 - Darwin-dev - lists.apple.com

Re: Intel source code release
by André-John Mas 18 Feb '06

18 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com André On 17-Feb-06, at 23:26 , Andrew Pinski wrote: This is a joke really because it was rushed by management and the software is no where near ready. Look at the compiler for an example, Apple did not update to a newer version of GCC during that time which means they are behind in speed in general anyways. Projects got pushed back because of this switch. An example is the move to dwarf2 debugging and multiple other things. Nine months is not enough time to do a proper release of any kind of big software as Mac OS X. -- Pinski A GCC developer which gets little credit from Apple for keeping GCC working on Darwin (even on x86 lately). _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/ajmas%40sympatico.ca _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I am not sure that it was only nine months. MacOS X has been compiled in parallel as an Intel binary since its introduction. I suspect that the reason they didn't use the latest GCC was for reasons of stability. I have already had to program for Solaris during the time they were going from their non-ANSI C++ compiler to their ANSI C++ compiler, and it cause many headaches. With such a large project it makes more sense to stick to something that works, to get things working and then concentrate on upgrading the technology afterwards. Trying to develop with too many unknowns is recipie for disaster. Nine months of really hard work by the folks at Apple, who rightly decided that it was more important to deliver a kick-ass x86 operating system than to spend forever trying to package the sources for third-party consumption. This email sent to ajmas(a)sympatico.ca This email sent to site_archiver(a)lists.apple.com

1 0

Re: Intel source code release
by Andrew Pinski 18 Feb '06

18 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com This is a joke really because it was rushed by management and the software is no where near ready. Look at the compiler for an example, Apple did not update to a newer version of GCC during that time which means they are behind in speed in general anyways. Projects got pushed back because of this switch. An example is the move to dwarf2 debugging and multiple other things. Nine months is not enough time to do a proper release of any kind of big software as Mac OS X. -- Pinski A GCC developer which gets little credit from Apple for keeping GCC working on Darwin (even on x86 lately). _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… Nine months of really hard work by the folks at Apple, who rightly decided that it was more important to deliver a kick-ass x86 operating system than to spend forever trying to package the sources for third-party consumption. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Intel source code release
by Mike Smith 18 Feb '06

18 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Feb 17, 2006, at 12:03 PM, darwin-dev-request(a)lists.apple.com wrote: From: "Peter O'Gorman" <peter(a)pogma.com> Subject: intel source code release To: darwin-dev(a)lists.apple.com, hackers(a)opendarwin.org You might search for a simpler reason. Date: Fri, 17 Feb 2006 20:44:23 +0100 From: St?phane Sudre <ssudre2(a)wanadoo.fr> Subject: Re: intel source code release To: darwin-dev(a)lists.apple.com = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I assume that the idea is to limit the source code availability to those who are attempting to steal Mac OS X and use it on systems not built or approved by Apple. I can understand and applaud the goal, but not the methods. By limiting published source code to that which is "infected" by the GPL, Apple is, in my honest opinion, scoring an own goal. What would be interesting to see in the Intel source version is the explanation why one week before the announcement of Mac OS X for Intel at WWDC, the x86 version of Darwin was said to be a dying 3-leg horse from a performance point of view with no hope of recovery and now, since the release of the Intel iMac Core Duo, it is (or stated to be) a stalion. Nine months of really hard work by the folks at Apple, who rightly decided that it was more important to deliver a kick-ass x86 operating system than to spend forever trying to package the sources for third-party consumption. As to your comments about performance, it was, and it is, and I hope that when you get your hands on an Intel-powered Macintosh you're as happy as the rest of us at how it turned out. This email sent to site_archiver(a)lists.apple.com

1 0

Re: intel source code release
by Andrew Gallatin 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Graham J Lee writes: > On 17/2/06 19:44, Stéphane Sudre wrote: > > > > > > What would be interesting to see in the Intel source version is the > > explanation why one week before the announcement of Mac OS X for Intel > > at WWDC, the x86 version of Darwin was said to be a dying 3-leg horse > > from a performance point of view with no hope of recovery and now, since > > the release of the Intel iMac Core Duo, it is (or stated to be) a stalion. > > > > This could help me for turf related stuff :) > > > I think (or at least, it appears to be) because "Darwin/x86" and "Darwin > on OS X for Intel" are probably different beasties. I'm surprised > there's not been more activity over on darwin-x86@ actually... They've already broken the IOKit driver ABI between Darwin and MacOSX on Intel by removing padding from critical C++ classes so that modules built on Darwin will not load on MacOSX, and vice-versa. So there is at least one difference. Drew _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: intel source code release
by Jonas Maebe 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 17 Feb 2006, at 21:49, Jonas Maebe wrote: Jonas _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… (although Apple's license is a free software license recognised by the GPL) I meant "by the FSF", of course (and I'm talking about the APSL 2.0, earlier versions were not free software licenses by the FSF definition). This email sent to site_archiver(a)lists.apple.com

1 0

Re: intel source code release
by Graham J Lee 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Thunderbird 1.5 (Macintosh/20051201) This could help me for turf related stuff :) -- Graham Lee UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342 _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… On 17/2/06 19:44, Stéphane Sudre wrote: What would be interesting to see in the Intel source version is the explanation why one week before the announcement of Mac OS X for Intel at WWDC, the x86 version of Darwin was said to be a dying 3-leg horse from a performance point of view with no hope of recovery and now, since the release of the Intel iMac Core Duo, it is (or stated to be) a stalion. I think (or at least, it appears to be) because "Darwin/x86" and "Darwin on OS X for Intel" are probably different beasties. I'm surprised there's not been more activity over on darwin-x86@ actually... This email sent to site_archiver(a)lists.apple.com

1 0

Re: intel source code release
by Jonas Maebe 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 17 Feb 2006, at 21:36, Derick Centeno wrote: Jonas _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… If what has been reported here is true and Apple is changing or has removed standard open source components to it's purposes (whatever that may be), which further limits other current and potentially workable open source projects -- also reported, it might be time to consider the fact that Apple did enter into a open source agreement of a particular unique nature which has aspects of the GPL standard but is not completely standard open source. Actually, Apple's open source license is based on the modified BSD license and is pretty much standard in that respect (if they would have used the "standard modified BSD license", as it is used by Open/ Net/FreeBSD, it wouldn't change anything to the current situation). You are confusing open source software and free software (although Apple's license is a free software license recognised by the GPL), and in particular GPL-compatible and GPL-incompatible free software licenses. Apple's free software license is by no means the only GPL- incompatible free software license out there, nor is it particularly insidious. You can find the whole list at http://www.fsf.org/ licensing/licenses/index_html#GPLIncompatibleLicenses Since they are the only copyright holders (I assume) on their APSL- licensed stuff, even if they would have used the GPL for older versions, they still could keep future versions completely closed. In the end, this has nothing to do with GPL vs anything, and licenses are just a red herring. It's simply about convincing Apple that keeping this stuff open will help them more than it will harm them. This email sent to site_archiver(a)lists.apple.com

1 0

Re: intel source code release
by Derick Centeno 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com If what has been reported here is true and Apple is changing or has removed standard open source components to it's purposes (whatever that may be), which further limits other current and potentially workable open source projects -- also reported, it might be time to consider the fact that Apple did enter into a open source agreement of a particular unique nature which has aspects of the GPL standard but is not completely standard open source. In short, it is a form of "open source" as Apple has redefined as open source in it's own terms, which Apple feels comfortable with; this is the kind of "open source" definition which Apple developers engage together with Apple in employing. Apple is not the only company to do so, and probably won't be the last. The specific details discussing this nuance appeared a long time ago -- perhaps on /. or elsewhere, I'm not sure. It was one of those articles of interest to lawyers and persons who follow, use and attend to the standard GPL. I believe I recall a statement in the same article of a key founder of the GPL concept commenting on the nature of the kind of "open source" clause which Apple has engaged upon and employed. If my memory serves me correctly, that comment foresaw the situation reported upon here, these many years later. I recommend that others evaluating this situation who do not care to "reason why" this or that with Apple or any similar corporate entity consider returning to the standard GPL fold, where there remain many good and solid projects and companies abiding to the spirit and nature of the standard GPL while simultaneously engage in commercial production. There remain many GPL based Linux distributions, and the strongest amongst those supporting the PPC remains Yellow Dog Linux (YDL) (www.yellowdoglinux.com) which adheres closely to Red Hat Linux. Apple is not the only entity that can redefine what is done or how; so can you. If you are looking to keep your different computer architectures going and yet remain within the GPL standard -- without odd and unpleasant surprises here/there or now/then, remember there is no problem running Red Hat Linux on Intel boxes (which the new Macs are just like any other Intel box) and run YDL on the PPCs you have and the projects you have will be done -- YOUR -- way! By the way, for all the preachy news regarding speed of the new Intel systems over PPC; some will be surprised to determine for themselves performance statistics when Intel machines and PPC machines are running Red Hat Linux or YDL. Best wishes, respectfully to all ... On Fri, 2006-02-17 at 20:44 +0100, Stéphane Sudre wrote: > On vendredi, février 17, 2006, at 04:32 PM, Graham Jouee wrote: > > > On 17 Feb 2006, at 14:38, Peter O'Gorman wrote: > > > >> Hi, > > [...] > >> > >> By limiting published source code to that which is "infected" by the > >> GPL, Apple is, in my honest opinion, scoring an own goal. > > [...] > > > > I agree, but then I think that a large slice of the people you just > > sent your message to (especially on hackers@) are going to agree also. > > I for one have filed an Enhancement Request with Apple asking for > > greater access to the Darwin/intel sources and outlining what I use it > > for, and why I won't be able to get all of my work done if I went to > > an intel-only environment with the current source availability. IOW, > > I explain why I'm not about to rush out and upgrade all of my PPC Macs > > ;-). I don't know how successful that will be, but source-code > > availability of the core UNIX is a reason that I and some of my > > colleagues have given to justify going with OS X; not having that > > makes other BSDs, Linux and Slowlaris[!] more attractive than OS X for > > some purposes. > > What would be interesting to see in the Intel source version is the > explanation why one week before the announcement of Mac OS X for Intel > at WWDC, the x86 version of Darwin was said to be a dying 3-leg horse > from a performance point of view with no hope of recovery and now, > since the release of the Intel iMac Core Duo, it is (or stated to be) a > stalion. > > This could help me for turf related stuff :) > _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: intel source code release
by Stéphane Sudre 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On vendredi, février 17, 2006, at 04:32 PM, Graham J Lee wrote: On 17 Feb 2006, at 14:38, Peter O'Gorman wrote: [...] This could help me for turf related stuff :) _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… Hi, [...] By limiting published source code to that which is "infected" by the GPL, Apple is, in my honest opinion, scoring an own goal. I agree, but then I think that a large slice of the people you just sent your message to (especially on hackers@) are going to agree also. I for one have filed an Enhancement Request with Apple asking for greater access to the Darwin/intel sources and outlining what I use it for, and why I won't be able to get all of my work done if I went to an intel-only environment with the current source availability. IOW, I explain why I'm not about to rush out and upgrade all of my PPC Macs ;-). I don't know how successful that will be, but source-code availability of the core UNIX is a reason that I and some of my colleagues have given to justify going with OS X; not having that makes other BSDs, Linux and Slowlaris[!] more attractive than OS X for some purposes. What would be interesting to see in the Intel source version is the explanation why one week before the announcement of Mac OS X for Intel at WWDC, the x86 version of Darwin was said to be a dying 3-leg horse from a performance point of view with no hope of recovery and now, since the release of the Intel iMac Core Duo, it is (or stated to be) a stalion. This email sent to site_archiver(a)lists.apple.com

1 0

Re: intel source code release
by Graham J Lee 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 17 Feb 2006, at 14:38, Peter O'Gorman wrote: [...] Cheers, Graham. -- Graham Lee GPG Key ID: 01D5B9D8 UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342 01865 273450 _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com Hi, [...] By limiting published source code to that which is "infected" by the GPL, Apple is, in my honest opinion, scoring an own goal. I agree, but then I think that a large slice of the people you just sent your message to (especially on hackers@) are going to agree also. I for one have filed an Enhancement Request with Apple asking for greater access to the Darwin/intel sources and outlining what I use it for, and why I won't be able to get all of my work done if I went to an intel-only environment with the current source availability. IOW, I explain why I'm not about to rush out and upgrade all of my PPC Macs ;-). I don't know how successful that will be, but source-code availability of the core UNIX is a reason that I and some of my colleagues have given to justify going with OS X; not having that makes other BSDs, Linux and Slowlaris[!] more attractive than OS X for some purposes. smime.p7s

1 0