June 2005 - Darwin-dev - lists.apple.com

Re: we were opensource, once
by Gary 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com User-agent: Mozilla Thunderbird 1.0.2 (Macintosh/20050317) Michael L Torrie wrote: Have you contacted any redhat developers about this issue? Under the terms of the GPL, you have legal access to the patch that fixes this bug. Not sure what you mean by that last statement. If Apple doesn't owe the open source community anything, then does the community owe apple anything? Ultimately, like I said, Apple is a cathedral, and I expect cathedral-like behavior from them. I had hoped, based on experiences with other OSS vendors, that it would be different. That was all I was really trying to say. -Gary _______________________________________________ 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 can download their kernel source for any major releases they make. So that's not the issue, really. I am curious, however, as to why they'd bother to obfuscate it in bugzilla. No, you said it quite well. And I imagine my response to that portion of the discussion was more to the original poster. Your cathedral example is a perfect one -- it's an historical analogy mentioned several times in Steven Levy's Hackers re IBM vs. DEC per their corporate culture. IBM was more like a priesthood in which their secrets were guarded closely with an almost religious-like zeal. DEC, on the other hand, invited feedback and was more friendly to hardware and OS hacking. This email sent to site_archiver(a)lists.apple.com

1 0

Re: we were opensource, once
by Finlay Dobbie 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=UvMiFSXlTmouiMSbeUaQXahJgR1StqAqXi95Wg/SARSKKjoES/kTByfG4pq8V6ByTQgoAiJ1AkXEOVbOvylcEi+SjFX6sgycJwwl2eOwUluIvmu++4PruZv1DK6C7bh57CQFsdP7DRcPsFi+4wX4V/Ckm6zSOIdKvxmwPEHu0T8= On 6/27/05, Markus Hitter <mah(a)jump-ing.de> wrote: > > Having clear patches to core code such as samba and openldap > > (opendirectory) > > would also go a long ways. Finally, shipping a self-hosting > > development > > environment (out of the box) would be a requirement to truly bring > > darwin back to where you feel it once was. > > There's nothing stopping the Darwin community to develop such things. > But how would you build a strong community around a closed, deaf and > mute cathedral? Another major problem with the self-hosting thing is that a lot of Darwin projects have moved over the past few years to being first PBX and now Xcode-based. Xcode is not open-source. There is no real way to build an Xcode project without Xcode, and creating the tools to do so is a fairly unattractive proposition because: - Xcode is a moving target (project file format changes in a non-backwards compatible way with some regularity - just see how many different filename extensions we've had in the pat 4 years: pbxproj, pbproj, xcode, xcodeproj) - the audience that would benefit from such tools is small - Apple engineers may or may not be working on their own solution to this problem in secret (although unlikely, Xcode has always had far bigger problems which undoubtedly keep the team fully occupied) - divining the steps taken by Xcode to build a given project is fairly laborious and uninteresting I did take some steps towards this goal a while back, but in the end I assigned other, more personally interesting projects a higher priority. And Xcode is heavily dependent on closed source technologies. Cocoa, for starters. -- Finlay _______________________________________________ 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: we were opensource, once
by Markus Hitter 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Am 27.06.2005 um 21:00 schrieb Dustin Lang: No, it isn't. Especially if you can't rebuild a modified ISO. Markus - - - - - - - - - - - - - - - - - - - Dipl. Ing. Markus Hitter http://www.jump-ing.de/ _______________________________________________ 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… Maybe I'm spoiled by my linux experience (make menuconfig; make; make modules_install), but I found the Darwin build experience somewhat less than satisfactory (read: I couldn't get it to work). Maybe the 'preferred' route is to just grab the ISO, but that's not really an open-source experience, is it? The currently preferred way to build Darwin is to use darwinbuild: It grabs source tarballs as well as patches for them and build off the resulting sources. I've yet to hear a success story on Darwin 8.x, however. This email sent to site_archiver(a)lists.apple.com

1 0

Re: we were opensource, once
by Markus Hitter 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Am 27.06.2005 um 18:46 schrieb Michael L Torrie: Have fun, Markus - - - - - - - - - - - - - - - - - - - Dipl. Ing. Markus Hitter http://www.jump-ing.de/ _______________________________________________ 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 would have been perfectly willing to dive into the code and try to identify the problems I had and help Apple fix them, but I was unable to do this. IMHO, the really disencouraging thing is, you can fix bugs to some extents, just to find them with the next OS X release fixed independently by an Apple engineer as well. I was initially encouraged by Apple's embracing of open source, but I have come to realize that Apple is just as much a cathedral as Microsoft is, only they happen to use (fork) open source code in a legal and license-abiding way. The funny thing these days is, Apple prefers to work lots more with other open source teams, like e.g. FSF gcc, than with it's own community. If you want to know about and participate in the future of the Apple version of gcc, looking an Apple Darwin or OpenDarwin is moot, you have to discuss on gcc's mailing lists instead. An open, read-only bug tracking system would be a great start. When the interested public was locked out of Apple's cvs and out of Radar and OpenDarwin was founded instead, I told my fears about the effective end of community work with Darwin, but was flamed. Hmmm. Having clear patches to core code such as samba and openldap (opendirectory) would also go a long ways. Finally, shipping a self-hosting development environment (out of the box) would be a requirement to truly bring darwin back to where you feel it once was. There's nothing stopping the Darwin community to develop such things. But how would you build a strong community around a closed, deaf and mute cathedral? For me, Darwin is read-only source and OpenDarwin is fine for building add-ons at the CLI level: XPostFacto, Darwine, ... you name it ... This email sent to site_archiver(a)lists.apple.com

1 0

Re: we were opensource, once
by Michael L Torrie 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com On Mon, 2005-06-27 at 11:15 -0700, Gary wrote: > Michael L Torrie wrote: > > Red Hat is just as guilty of hiding their dirty laundry. Take this > for example: > > [PATCH] read() sometimes crashes in the second pthread_set_canceltype > call (https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=130926) > > This is a major 2.4 linux kernel patch that fixes a serious flaw in the > POSIX threading library in RH Enterprise Linux 2.1 -- an OS that's been > around for a few years now. Strangely enough, even when logged in to my > Red Hat bugzilla account, I'm told that I am not authorized to access > that particular bug ID. Tis true. However none of Apple's bug system is open at all to my account. RedHat does need to get their act together on this also. Apple is not the only guilty one, just more guilty. Have you contacted any redhat developers about this issue? Under the terms of the GPL, you have legal access to the patch that fixes this bug. > You are forgetting that Apple, HP, IBM, SGI, Sun Microsystems, Intel, > nVidia, ATI, VIA, Red Hat, Novell, etc. are all businesses that sell > highly competitive products and will continue to do anything to stay > in business and make money for themselves and shareholders -- even > if it is under the guise of allegedly contributing to the open > source community. I can't deny that some of these companies have > contributed significiantly to the open source community -- some more > than others -- but one shouldn't be surprised when they hold their > cards close to their chest. I would imagine that Sun is going to > face similar challenges with Open Solaris. Who's going to want to > give freely to a project that is unlikely to give much back? Not sure what you mean by that last statement. If Apple doesn't owe the open source community anything, then does the community owe apple anything? Ultimately, like I said, Apple is a cathedral, and I expect cathedral-like behavior from them. I had hoped, based on experiences with other OSS vendors, that it would be different. That was all I was really trying to say. > > -Gary > _______________________________________________ > Hackers mailing list > Hackers(a)opendarwin.org > http://www.opendarwin.org/mailman/listinfo/hackers -- _______________________________________________ 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: we were opensource, once
by Gary 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com User-agent: Mozilla Thunderbird 1.0.2 (Macintosh/20050317) Michael L Torrie wrote: Red Hat is just as guilty of hiding their dirty laundry. Take this for example: [PATCH] read() sometimes crashes in the second pthread_set_canceltype call (https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=130926) This is a major 2.4 linux kernel patch that fixes a serious flaw in the POSIX threading library in RH Enterprise Linux 2.1 -- an OS that's been around for a few years now. Strangely enough, even when logged in to my Red Hat bugzilla account, I'm told that I am not authorized to access that particular bug ID. You are forgetting that Apple, HP, IBM, SGI, Sun Microsystems, Intel, nVidia, ATI, VIA, Red Hat, Novell, etc. are all businesses that sell highly competitive products and will continue to do anything to stay in business and make money for themselves and shareholders -- even if it is under the guise of allegedly contributing to the open source community. I can't deny that some of these companies have contributed significiantly to the open source community -- some more than others -- but one shouldn't be surprised when they hold their cards close to their chest. I would imagine that Sun is going to face similar challenges with Open Solaris. Who's going to want to give freely to a project that is unlikely to give much back? -Gary _______________________________________________ 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… Unlike fellow enterprise competitors like RedHat, Apple's darwin bug tracking system is not open and I was not able to see if my problems had already been reported. ... Had this been RedHat, I could have had dialogs with the RedHat engineers and developers over this. These are the things I have come to expect and rely on in my traditional Linux world. This email sent to site_archiver(a)lists.apple.com

1 0

Re: geteuid() while debugging
by Jim Ingham 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Jim On Jun 26, 2005, at 12:04 PM, darwin-dev-request(a)lists.apple.com wrote: 2005/06/25 v 21:58, Justin C. Walker: You have to run gdb as root to debug processess running as root. (If you think about it, it makes sense.) I don't know whether there's any way for Xcode to run gdb as root but you can do it yourself (using sudo) if you don't depend on Xcode's GUI. I think you have to log in as root to do this; as I recall a change was made to Mac OS X a while back to prohibit launching GUI apps as root while logged in as not-root (or, perhaps, just prohibiting launching GUI apps under a different ID than that of the logged-in user; I forget). Not really. If one really needs to run a GUI application as root it's quite easy using sudo, e.g.: sudo /Applications/TextEdit.app/Contents/MacOS/TextEdit & There were just some minor issues in earlier versions of Mac OS X and in Tiger it seems to work perfectly. But this is not what I was talking about: For those who like the Xcode debugger GUI it might be nice (and should be technically possible AFAIK) to run Xcode as an ordinary user (to avoid potential risks) and gdb as root "inside" XCode, but there's no such option in Xcode. -- Adam Nohejl Loki Software mailto:adam@lokisw.com http://lokisw.com _______________________________________________ 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… There's a feature request for running or debugging the target app as another user in the Xcode database. However, feel free to file one yourself; the more noise we get about a particular feature, the higher priority it tends to get... This email sent to site_archiver(a)lists.apple.com

1 0

Re: we were opensource, once
by Michael L Torrie 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com On Mon, 2005-06-27 at 23:52 +0900, Peter O'Gorman wrote: > Hi, > First a note that this is a whine. I know it is. I really appreciate your comments. As a lurker who has never participated in Darwin development, I think your comments very accurate portray the impression I have had for some time now. Apple appears to be open source, and indeed is in many respects (IE source code is available), but the process is definitely no open. This is the letter of the open source law, but not the spirit, and it is very disheartening to me. For example, I have reported several bugs to Apple regarding some serious issues I had with Panther Server. These bugs mostly centered on OpenDirectory and were likely in OpenLDAP and how it interacted with PasswordService (which is closed source). Unlike fellow enterprise competitors like RedHat, Apple's darwin bug tracking system is not open and I was not able to see if my problems had already been reported. Nor had I any idea if Apple's engineers were making any progress in tracking down these bugs. Had this been RedHat, I could have had dialogs with the RedHat engineers and developers over this. Fortunately, almost two years later, the problems I experienced seemed to be fixed, so when I speak about my LDAP problems, I'm talking more about the process than the actual bug itself. Of course, the source code to the pieces like OpenLDAP is available on the opendarwin site as well as Apple's site. However, though the source code is available, it is not buildable. There are developer-only dependencies (header files, special frameworks for building) that are difficult (if only because I have no idea how to obtain them) for us mere users to get a hold of. I would have been perfectly willing to dive into the code and try to identify the problems I had and help Apple fix them, but I was unable to do this. Besides the problem of unbuildable source tarballs, Apple does not cleanly separate their addons and patches from the original sources. This makes tracking bugs and problems outside of the Apple support system virtually impossible. I would have loved, for example, to try the newer OpenLDAP 2.2 code under Panther Server to see how it performed. However without patches, such an endeavour is very difficult. So basically I have to consider all source code from darwin to be a complete fork. This seems to be counter productive because it cuts off darwin and apple users from the community support that already exists outside of Apple. The OpenLDAP developers, for example, cannot do anything to help with openldap problems, even if the bug was actually in their own code and not Apple's extensions. I was initially encouraged by Apple's embracing of open source, but I have come to realize that Apple is just as much a cathedral as Microsoft is, only they happen to use (fork) open source code in a legal and license-abiding way. Apple has clearly been advantaged because of this and I don't have a problem with that, but I would like to see things a little more open and transparent as you have suggested. An open, read-only bug tracking system would be a great start. Having clear patches to core code such as samba and openldap (opendirectory) would also go a long ways. Finally, shipping a self-hosting development environment (out of the box) would be a requirement to truly bring darwin back to where you feel it once was. These are the things I have come to expect and rely on in my traditional Linux world. Michael -- _______________________________________________ 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: Early lowlevel startup of user processes (crt1.o ... )??
by Rob Ballantyne 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Mozilla Thunderbird 1.0 (Macintosh/20041206) It's more an attempt to figure out what's required-- just educational. Ok, I'll try adding that. I was assuming (stupidly probably) that since I wasn't using threads that that wasn't required. Is it really true that what would normally be 'libc' requires pthreads? Thanks for the help! Shantonu On Jun 26, 2005, at 4:13 PM, Rob Ballantyne wrote: Hi, I'm trying to understand the early startup process. In particular I'm trying to build a standalone assembly language program (no crt?.o) that will link against libSystem.dylib. I've managed to link correctly against it for simple routines (atoi for example) but I've run into a problem attempting to use printf. The system cores on me. Witness: (gdb) run Starting program: /Users/ballanty/devel/AsmTest/test3 Testing...1,2, Program received signal EXC_BAD_ACCESS, Could not access memory. Reason: KERN_PROTECTION_FAILURE at address: 0x00000000 0x90003464 in malloc () (gdb) where #0 0x90003464 in malloc () #1 0x90110d08 in localeconv_l () #2 0x9000fc30 in __vfprintf () #3 0x900fa088 in vfprintf_l () #4 0x90022318 in printf () #5 0x00001f88 in start () at test3.s:28 #6 0x00000000 in ?? () Cannot access memory at address 0x0 Cannot access memory at address 0xa Cannot access memory at address 0x0 Cannot access memory at address 0x0 #7 0x00000000 in ?? () Cannot access memory at address 0x0 Cannot access memory at address 0x0 Cannot access memory at address 0xa Previous frame inner to this frame (corrupt stack?) (gdb) It looks like it is linked correctly but I can't figure out why malloc is core-ing (looks like it's dereferncing a null pointer to me). I suspect there is some startup routing (initialization?) that must be called before some stuff from the C library can be called. Can anyone confirm that? In crt.c it appears the only significant thing that gets called is: __dyld_make_delayed_module_initializer_calls. The comment in the code seems to indicate it's C++ specific though!!! Can anyone confirm/deny my guess? I've examined: start.s, crt.c, and dyld.s. This appears it might be the problem. Many Thanks for any help! Rob _______________________________________________ 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/ssen% 40opendarwin.org This email sent to ssen(a)opendarwin.org _______________________________________________ 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… Shantonu Sen wrote: Why are you doing this? Why not start with a working crt1.o and remove what you don't want? Lots of things get called from crt.c, among them (*_cthread_init_routine)() in libSystem.B.dylib, which is responsible for initializing pthread state for the default thread. Without calling this routine, you're not going to get very far. Rob !DSPAM:42bf9892144684465477762! This email sent to site_archiver(a)lists.apple.com

1 0

we were opensource, once
by Peter O'Gorman 27 Jun '05

27 Jun '05

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com User-agent: Mozilla Thunderbird 1.0 (Macintosh/20041206) Peter -- Peter O'Gorman - http://www.pogma.com _______________________________________________ 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, First a note that this is a whine. I know it is. But I find myself looking at webkit <http://webkit.opendarwin.org> and it is making me remember the day... Webkit is making use of opendarwin's bugzilla, even for bugs reported by team members. Webkit has a live cvs repository available to all. Webkit developers discuss things openly on open mailing lists. The project, at the moment is truly open. This is good stuff. But, then I remember. Darwin used to have an open bug system where everyone could view opensource bugs. Darwin used to have a mostly live cvs repository that everyone could use to check out almost-up-to-date sources. Darwin used to even have non-Apple committers. Now, of course, we get source tarballs. We do get them on-time, and mostly complete, which is good. But we don't get any input into the development of the operating system. We do not get any opportunity to help. There are a number of features included in mac os x 10.4 that I would have liked to have seen argued/discussed, but there was no forum to do so, so the discussion remained internal to Apple, and the result, while good, could have been better. There is currently no reason for any non-Apple person to start hacking on darwin. External opensource projects are included in the OS, it is true. But many of these external projects have to look at the opensource tarballs to see how to get their projects building/running on a new darwin version. This is hardly an opensource developers dream. Oh well. I'm told that writing this will accomplish nothing. I'll simply be flamed. Whatever. I'm happy to see that at least some people at Apple are considering opensource. Thanks webkit. This email sent to site_archiver(a)lists.apple.com

1 0