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

Re: Darwin-dev Digest, Vol 2, Issue 112
by Alexey Proskuryakov 16 May '05

16 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Microsoft-Entourage/11.1.0.040913 On 16.05.2005 20:49, "Justin Walker" <justin(a)mac.com> wrote: >> FWIW, last August I reported how a file can be deleted via SMB, and still >> be seen as existing on the same (!) local system (rdar://3749257). It's >> still not fixed, so looks like Apple is not very interested in such >> issues. > > It seems like a stretch to say that, because your favorite bug isn't > fixed, Apple isn't interested. I've always found them to be > interested, but often there are higher priority issues to deal with. Well, perhaps I'm too afraid to imagine a file system issue worse than this in a shipping system :). Seriously, I thought that I implied the same by saying "isn't very interested", rather than "isn't interested". - WBR, Alexey Proskuryakov _______________________________________________ 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: Darwin-dev Digest, Vol 2, Issue 112
by Justin Walker 16 May '05

16 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 16, 2005, at 8:50, Alexey Proskuryakov wrote: On 16.05.2005 16:24, "Dan Smith" <d_smith(a)ecrm.com> wrote: If this is truly a bug, and one that is likely to bite more than a very small number of people, our best strategy is to wait for Apple to fix it. Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large Institute for General Semantics -------- Some people have a mental horizon of radius zero, and call it their point of view. -- David Hilbert -------- _______________________________________________ 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 was hoping that someone on this list might know more about what's going on "under the hood" and might have a plausible explanation for what changed in 10.4 and how it is possible to get a sequence of events in which a) a file is deleted on the remote system, b) is SEEN as not existing any more on the local system, but c) gets an error when the local system attempts to rename a file to the now-available name of the deleted file. FWIW, last August I reported how a file can be deleted via SMB, and still be seen as existing on the same (!) local system (rdar://3749257). It's still not fixed, so looks like Apple is not very interested in such issues. It seems like a stretch to say that, because your favorite bug isn't fixed, Apple isn't interested. I've always found them to be interested, but often there are higher priority issues to deal with. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Darwin-dev Digest, Vol 2, Issue 112
by Alexey Proskuryakov 16 May '05

16 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Microsoft-Entourage/11.1.0.040913 On 16.05.2005 16:24, "Dan Smith" <d_smith(a)ecrm.com> wrote: > I was hoping that someone on this list might know more about > what's going on "under the hood" and might have a plausible > explanation for what changed in 10.4 and how it is possible to get > a sequence of events in which a) a file is deleted on the remote system, > b) is SEEN as not existing any more on the local system, but > c) gets an error when the local system attempts to rename a > file to the now-available name of the deleted file. > If this is truly a bug, and one that is likely to bite more > than a very small number of people, our best strategy is > to wait for Apple to fix it. FWIW, last August I reported how a file can be deleted via SMB, and still be seen as existing on the same (!) local system (rdar://3749257). It's still not fixed, so looks like Apple is not very interested in such issues. In my case, looks like that it's case-insensitivity of SMB that doesn't work properly; however, your issue is probably different. - WBR, Alexey Proskuryakov _______________________________________________ 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: Darwin-dev Digest, Vol 2, Issue 112
by Dan Smith 16 May '05

16 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com darwin-dev(a)lists.apple.com writes: >ontent-Type: text/plain; charset=ISO-8859-1 > >On 5/13/05, Justin Walker <justin(a)mac.com> wrote: >> >> On May 13, 2005, at 13:00, Dan Smith wrote: >> >> > The test application mentioned below is available for download from: >> > >> > http://world.std.com/~dpbsmith/CrossNetworkRenameBug.app.zip >> >> If this is supposed to be a bug report, you won't get much satisfaction >> here (it's not an official channel to Apple). I did report it to Apple. That's why the message begins "Entered as bug 4118699." > >And even if it isn't, high-level applications are outside the scope of >this list. It's not a problem with a high-level application. I supplied a high-level application that _demonstrates_ the problem. I was hoping that someone on this list might know more about what's going on "under the hood" and might have a plausible explanation for what changed in 10.4 and how it is possible to get a sequence of events in which a) a file is deleted on the remote system, b) is SEEN as not existing any more on the local system, but c) gets an error when the local system attempts to rename a file to the now-available name of the deleted file. If this is truly a bug, and one that is likely to bite more than a very small number of people, our best strategy is to wait for Apple to fix it. If it's just a case of a change in undocumented behavior that we shouldn't be depending on, then we need to figure out how to work around it. That requires a better understanding of what's happening than I now have. If there's a better mailing list for discussing issues with file system API's used with remote SMB files under OS X 10.4, by all means point me to it. > _______________________________________________ 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: Tiger speed issues...
by Stéphane Sudre 15 May '05

15 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com dimanche 15 mai 2005, à 02:57 AM, Peter Bierman a écrit : Since it's doing less, that's not surprising. o No more display of StartupItem being launched. o No more display of Safe Boot until the Login Window is displayed. The last 2 points are a bit a problem IMHO. _______________________________________________ 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… At 9:51 AM +1000 5/15/05, wadeslists(a)mac.com wrote: Is this due to the fact the Tiger kernel is 64 bit based and G4 processors are not? Or that I have 256MB RAM instead of 1GB? Bingo - the RAM, that is. Tiger is, as with every new incarnation of MacOS X, even more RAM hungry than the previous version. Actually, if you take some measurements, you will find that Tiger uses LESS memory to get from boot to the Finder than 10.3.9 does. o More StartupItems being removed to use other more On-Demand mechanisms. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Tiger speed issues...
by j o a r 15 May '05

15 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 15 maj 2005, at 12.31, Jonas Maebe wrote: Don't guess, measure! Use Shark! j o a r _______________________________________________ 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 There a lot more swapping (or at least disk activity, athough spotlight has finished indexing quite a while ago) going on than before when having a few apps active at the same time. smime.p7s

1 0

Re: Tiger speed issues...
by Jonas Maebe 15 May '05

15 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 15 May 2005, at 02:57, Peter Bierman 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… Actually, if you take some measurements, you will find that Tiger uses LESS memory to get from boot to the Finder than 10.3.9 does. There is some change though, I experience exactly the same slowdown on my iBook G4/1.2 GHz with 256MB ram going from 10.3.9 to 10.4. There a lot more swapping (or at least disk activity, athough spotlight has finished indexing quite a while ago) going on than before when having a few apps active at the same time. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Tiger speed issues...
by Sean McBride 15 May '05

15 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Warwick Hall (phosgate(a)unwired.com.au) on 2005-05-15 19:44 said: >I have a iBook G4 1GHz 256MB computer as my development machine. >Recently I installed Tiger. Before my machine was running smoothly. >Now it is noticeably slower. Waking up and application switching >takes toooooo long for instance. What does Activity Monitor say? What does Shark say? What does this have to do with Darwin? Maybe the Spotlight indexer is running for the first time and scanning all your disks? -- "Have you ever been out some place, recognizing everybody's face, until you realized that there was no one there you knew?" - The Offspring _______________________________________________ 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: Tiger speed issues...
by Peter Bierman 15 May '05

15 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -pmb _______________________________________________ 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… At 9:51 AM +1000 5/15/05, wadeslists(a)mac.com wrote: Is this due to the fact the Tiger kernel is 64 bit based and G4 processors are not? Or that I have 256MB RAM instead of 1GB? Bingo - the RAM, that is. Tiger is, as with every new incarnation of MacOS X, even more RAM hungry than the previous version. Actually, if you take some measurements, you will find that Tiger uses LESS memory to get from boot to the Finder than 10.3.9 does. But yes, 256 MB is the suggested minimum, and as such will feel sluggish compared to larger amounts. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Tiger speed issues...
by Tim Seufert 15 May '05

15 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 14, 2005, at 4:57 PM, Justin Walker wrote: _______________________________________________ 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… What you are experiencing is, I think, due to the small size of memory. I think (although I've not been motivated to track this down) that there is more paging going on now, possibly because apps and frameworks are bigger. However, I have to say, once an app is up front and in control, my overworked Pismo feels way snappier than earlier versions of Mac OS X. As confirmation of the RAM theory, I offer the fact that my PowerBook G4 (with 1.5GB RAM) feels faster just about everywhere under 10.4 than 10.3, not just when one app is in front and stays there. I have not yet observed 10.4 to do any paging on this machine. 10.3 would page things out occasionally, even though it didn't really need to with this much RAM. Probably 10.4's VM system is better at reclaiming buffer cache pages when under memory pressure, or something like that. (There, I've mentioned part of the Darwin kernel to help keep this on topic. :) This email sent to site_archiver(a)lists.apple.com

1 0