ALEX KAN - MacOS Error -5000 still persistent on v8 for G5

Questions and Answers : Macintosh : ALEX KAN - MacOS Error -5000 still persistent on v8 for G5

Author	Message
Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 677176 - Posted 13 Nov 2007 13:56:34 UTC
	Hopefully Alex Kan will read this or have it relayed to him. Just recently, I've discovered that some of my Tasks have been processed without the annoying MacOS Error -5000 message. The error seems to have completely stopped on my G4 http://setiathome.berkeley.edu/results.php?hostid=3734525 But not on my G5 Dual core http://setiathome.berkeley.edu/results.php?hostid=2881519 To recap., here is the offending message, always duplicated:- <core_client_version>5.10.29</core_client_version> <![CDATA[ <stderr_txt> MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G5-optimized v8-nographics) V5.13 by Alex Kan The G4 shows no Error message:- <core_client_version>5.10.29</core_client_version> <![CDATA[ <stderr_txt> OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G4-optimized v8-nographics) V5.13 by Alex Kan My assumption is that Alex Kan has very kindly cured the problem in v8 for the G4, but has not yet amended in v8 for the G5. Keith
	ID: 677176 \|

Odysseus Volunteer tester Send message Joined: Jul 26 99 Posts: 1615 Credit: 1,985,646 RAC: 1,437	Message 677507 - Posted 14 Nov 2007 2:20:23 UTC - in response to Message 677176.
	To recap., here is the offending message, always duplicated:- It might be happening each time the app starts crunching. Do these tasks take between 1x and 2x your switch-between interval? Does the executable file have a green-dish icon, or the generic terminal? My assumption is that Alex Kan has very kindly cured the problem in v8 for the G4, but has not yet amended in v8 for the G5. It must be something to do with the security sandbox. My v8-for-G5 app here shows the green dish and doesnÕt get the errorsÑbut IÕm still running BOINC v5.4.9. OTOH my partnerÕs iMac is also doing fine with v8-for-Core2 under BOINC v5.8.17. ____________
	ID: 677507 \|

Alex Kan Volunteer developer Send message Joined: Dec 4 03 Posts: 127 Credit: 29,269 RAC: 0	Message 677580 - Posted 14 Nov 2007 7:47:09 UTC - in response to Message 677176.
	I've checked the G5 archive I posted, and it looks like the icon was correctly baked into the app. However, from reexamining the source code for the section that sets the icon (which was part of the original source code I checked out from CVS once upon a time), I think I understand what you're running into. Odysseus was right to ask about the security sandbox and file permissions, as the app actually tries to set its icon regardless of whether or not it has already been set. Therefore, anytime the app doesn't have write access to itself with the permissions it's running with, it fails to set its own icon and displays the error you were seeing. My suggestion for fixing the issue is to check the permissions for the application itself, and make sure that the permissions for both the owner and group (under Ownership & Permission > Details, if you're using Get Info in the Finder) are both set to "Read & Write."
	ID: 677580 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 677876 - Posted 14 Nov 2007 21:33:42 UTC
	Alex It seems as though it is still a mystery. I've checked the Info for the green icon for the Setiathome optimized app and it shows as: Owner boinc_mster Read & Write Group boinc_project Read & Write others Read Only sah_40.png which is the black logo has the same permissions, which is not surprising, as all contents of the folder had been set (by me from the folder info) to the same permissions on BOINC installation. There is also a file stat_icon, which is a text file relating to a soft link to sah_40.png. This would appear to be the only one accessed today according to the Info. What is puzzling is that the G5 and the G4 have been treated identically each with their appropriate optimized version of the setiathome. And both are working perfectly, but for that annoying message on the G5. Odysseus I am not familiar with "security sandbox" unless this is just a technical term for permissions. As I mentioned above, the executable file does have the green dish in both G4 and G5. I assume you mean "switch between applications" in processor usage. Both set to 60 mins (unaltered). Presumably this is academic as both computers run setiathome alone with no other project involved? You are running earlier versions. So was I previously and always had had the -5000 error messages. It is only just recently I have noticed that my G4 desktop and G4 titanium laptop are now both cured. However they both always used to show the errors previously (I assume that that was up to the change from v7.1 to v8, but am not certain). Let me know what either or both of you would like me to do to help find the root of this G5 problem. Thanks, Alex for all your efforts in getting the Optimized apps up and running. Much appreciated. Keith P.S. I'll do a repair permissions with Disk Utility, just in case that might help.
	ID: 677876 \|

Odysseus Volunteer tester Send message Joined: Jul 26 99 Posts: 1615 Credit: 1,985,646 RAC: 1,437	Message 678077 - Posted 15 Nov 2007 3:07:59 UTC - in response to Message 677876.
	I am not familiar with "security sandbox" unless this is just a technical term for permissions. I was referring to the ownership of the BOINC folder by the boinc_master & boinc_project user & group you mention above, which are part of a tighter security regime in the recent Mac versions. With earlier versions of BOINC all the files belong to the user who installed it. See Sandbox design on the BOINC site. In general a ÒsandboxÓ is a restricted running environment that prevents programs from accessing data outside their domain. Should BOINC somehow download a Trojan-horse application, any consequential damage or security breaches would be limited to the contents of the BOINC Data folder, as the OS will block any of its attempts to open files that donÕt belong to the BOINC pseudo-user.
	ID: 678077 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 678168 - Posted 15 Nov 2007 12:02:26 UTC
	Odysseus O.K. I should have done a search on "sandbox", and didn't do so. But, having said that, I am still not a lot wiser as to how to help. With a little guidance I am willing to do whatever may be required on the G5 to track down the reason for this -5000 error warning. Keith
	ID: 678168 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 681561 - Posted 20 Nov 2007 17:16:47 UTC Last modified: 20 Nov 2007 17:21:47 UTC
	Is nobody else getting the Mac error -5000 message that I always get on my G5? Running without error message on G4 and on Macbook Pro, all on OSX 10.4.11 and BOINC 5.10.30. All are each running their own version 8 of Alex Kan's Enhanced apps. That's the only difference. Permissions appear to be correct on all Macs. What tests can I make to find reason for the annoying error messages? Details of recent task completed as follows:- ============================================= Task ID 660396647 Name 20fe07ag.30859.72642.8.6.233_0 Workunit 180005948 Created 16 Nov 2007 12:48:14 UTC Sent 16 Nov 2007 21:26:19 UTC Received 20 Nov 2007 9:50:09 UTC Server state Over Outcome Success Client state Done Exit status 0 (0x0) Computer ID 2881519 Report deadline 25 Nov 2007 13:46:19 UTC CPU time 3817.450523 stderr out <core_client_version>5.10.30</core_client_version> <![CDATA[ <stderr_txt> MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G5-optimized v8-nographics) V5.13 by Alex Kan Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 1.669327 Flopcounter: 5369287543084.746094 Spike count: 4 Pulse count: 0 Triplet count: 1 Gaussian count: 0 </stderr_txt> ]]> Validate state Valid Claimed credit 17.7276261121097 Granted credit 17.7276261121097 application version 5.27 ============================================= Keith
	ID: 681561 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 682763 - Posted 22 Nov 2007 21:39:45 UTC Last modified: 22 Nov 2007 22:38:10 UTC
	Now, I don't know if anyone is "listening" but I'm doing my best to solve this problem. I've installed BOINC Manager 5.4.9 to see if (as with Odysseus on his G5) the mysterious -5000 message is gone. But, no. Here is the log report:- ========================================================== stderr out <core_client_version>5.4.9</core_client_version> <stderr_txt> MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G5-optimized v8-nographics) V5.13 by Alex Kan Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.383378 OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G5-optimized v8-nographics) V5.13 by Alex Kan Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.383378 Flopcounter: 22679468585106.753906 Spike count: 0 Pulse count: 0 Triplet count: 0 Gaussian count: 0 </stderr_txt> Validate state Valid Claimed credit 74.8271796291752 Granted credit 74.8271796291752 application version 5.27 ========================================================== So, what next? Are there any other G5s out there with the same -5000 messages? Or, are there any without the -5000 messages? Or what else might be causing the problem on my G5? Maybe I can do some delving to see if this "sandbox" is where I must look? (How and where?) Keith P.S. Having a look around in version 5.4.9 (re-installed as mentioned above), it makes me appreciate (from what was missing there) what a terrific job has been done in improving the BOINC Manager in the time up to version 5.10.30. So much is so easily taken for granted. I would encourage anyone else to do a temporary backward install and they will then, too, really appreciate the good work that has been done since then. P.P.S. Just seen another task completed in version 5.4.9 and guess what ---- NO -5000 MESSAGES!!!!!! It turns out that this was on the first task STARTED on 5.4.9, whereas the 2 that were RESUMED still had the 2 -5000 errors reported, I've just got to go back to 5.10.30 to see if it's cured (and we never know why), or see if it returns. Here is the latest "clean" log:- =================================================== <core_client_version>5.4.9</core_client_version> <stderr_txt> OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G5-optimized v8-nographics) V5.13 by Alex Kan Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.432258 Flopcounter: 15823255312929.169922 Spike count: 0 Pulse count: 0 Triplet count: 2 Gaussian count: 0 </stderr_txt> =====================================================
	ID: 682763 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 683562 - Posted 24 Nov 2007 13:46:33 UTC
	No good. I've reinstalled 5.10.30 in place of 5.4.9 and It's back again on every task completed. What can I do next? ========================================================== <core_client_version>5.10.30</core_client_version> <![CDATA[ <stderr_txt> MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G5-optimized v8-nographics) V5.13 by Alex Kan Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 1.387897 Flopcounter: 5877930596537.412109 Spike count: 0 Pulse count: 1 Triplet count: 0 Gaussian count: 0 </stderr_txt> ]]> ======================================================== I cannot believe I am the only sufferer of this error message. And why only on my G5? I did notice that permission under 5.4.9 showed as my own name rather than boinc_master and that may have something to do with it. I did try removing the stat_icon file from the setiathome.berkeley.edu folder, which had the effect of showing a yellow "B" icon on the Simple View screen instead of the black & white Icon. The stat_icon file was however reinstated and showed the black & white icon again at the next task update. Now, I don't know how much of this will be useful, but hope that it may help Alex Kan track down the anomaly, which I intuitively think it must have it's answer in seti_enhanced-ppc-v8-g5-nographics or in some file I have somehow got installed in the BOINC Data folder, but i do not know where to start looking here. There are no error messages on my G4 computers, nor on the MacBook Pro, even though they appear to be set up identically (but for each appropriate Alex Kan app). The -5000 error message had appeared on all computers prior to Alex's version 8 being installed. Has anyone got any ideas? There must be a simple answer there somewhere. Keith
	ID: 683562 \|

Alex Kan Volunteer developer Send message Joined: Dec 4 03 Posts: 127 Credit: 29,269 RAC: 0	Message 684302 - Posted 25 Nov 2007 19:28:36 UTC - in response to Message 683562.
	Optimized apps compiled for different architectures don't differ in any way that would cause them to behave differently with regard to setting their icons, and as I discovered before, apps always try to set their icons every time they run, regardless of whether or not they've already done so in the past. That the messages went away with 5.4.9 and returned with 5.10.30 still points to the problem being a file permissions issue. One thing I can suggest at this point is to revert to the stock SETI app via BOINC, then make the permissions on the optimized app match the stock app exactly. This is actually what I've done the last few times I've installed the app on a machine. Do you know how to use the Terminal and commands like ls and chmod? You'll have more control over things (like execute permissions) this way. Perhaps I should look into whether or not I can have the icon-setting code fail silently. It seems like people are more likely to notice an error message than a missing icon. :P
	ID: 684302 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 684437 - Posted 26 Nov 2007 1:11:40 UTC - in response to Message 684302. Last modified: 26 Nov 2007 1:15:10 UTC
	Optimized apps compiled for different architectures don't differ in any way that would cause them to behave differently with regard to setting their icons, and as I discovered before, apps always try to set their icons every time they run, regardless of whether or not they've already done so in the past. That the messages went away with 5.4.9 and returned with 5.10.30 still points to the problem being a file permissions issue. One thing I can suggest at this point is to revert to the stock SETI app via BOINC, then make the permissions on the optimized app match the stock app exactly. This is actually what I've done the last few times I've installed the app on a machine. Do you know how to use the Terminal and commands like ls and chmod? You'll have more control over things (like execute permissions) this way. Perhaps I should look into whether or not I can have the icon-setting code fail silently. It seems like people are more likely to notice an error message than a missing icon. :P Alex I don't know if this is what you require to see why the G5 is different from the G4 (which has no problem). The only difference I can see is in the .DS_Store file where it is the G4 that has my "personal" permission instead of boinc_master. Maybe that should be my personal permission on the G5, but it cannot be changed from the GetInfo dialogue. The 2 listings are as follows: Last login: Sun Nov 25 22:48:25 on ttyp1 Welcome to Darwin! You have new mail. KLS-G4:~ keithscott$ ls -l -R -a /Library/Application\\ Support/BOINC\\ Data/projects/ total 16 drwxrwxr-x 4 boinc_ma boinc_ma 136 Nov 18 22:22 . drwxrwxr-x 31 boinc_ma boinc_ma 1054 Nov 25 23:07 .. -rw-rw-r-- 1 boinc_ma boinc_pr 6148 Nov 18 22:22 .DS_Store drwxrwxr-x 27 boinc_ma boinc_pr 918 Nov 25 18:00 setiathome.berkeley.edu /Library/Application Support/BOINC Data/projects//setiathome.berkeley.edu: total 9800 drwxrwxr-x 27 boinc_ma boinc_pr 918 Nov 25 18:00 . drwxrwxr-x 4 boinc_ma boinc_ma 136 Nov 18 22:22 .. -rw-rw-r-- 1 keithsco boinc_pr 6148 Nov 23 23:25 .DS_Store -rw-rw-r-- 1 boinc_ma boinc_pr 375287 Nov 25 06:42 01ja07aa.19394.8661.8.6.120 -rw-rw-r-- 1 boinc_pr boinc_pr 21944 Nov 25 18:37 01ja07aa.19394.8661.8.6.120_1_0 -rw-rw-r-- 1 boinc_ma boinc_pr 375290 Nov 24 18:29 22fe07ac.5713.19704.16.6.75 -rw-rw-r-- 1 boinc_pr boinc_pr 21313 Nov 25 12:26 22fe07ac.5713.19704.16.6.75_2_0 -rw-rw-r-- 1 boinc_ma boinc_pr 375294 Nov 24 21:21 22fe07ad.16308.22976.12.6.209 -rw-rw-r-- 1 boinc_pr boinc_pr 21316 Nov 25 12:28 22fe07ad.16308.22976.12.6.209_1_0 -rw-rw-r-- 1 boinc_ma boinc_pr 375290 Nov 24 16:11 22fe07ad.29008.17659.5.6.128 -rw-rw-r-- 1 boinc_pr boinc_pr 21313 Nov 24 16:17 22fe07ad.29008.17659.5.6.128_1_0 -rw-rw-r-- 1 boinc_ma boinc_pr 375280 Nov 25 15:33 22mr07ad.22531.17659.4.6.109 -rw-rw-r-- 1 boinc_ma boinc_pr 414 Oct 7 23:42 AUTHORS -rw-rw-r-- 1 boinc_ma boinc_pr 17988 May 27 2006 COPYING -rw-rw-r-- 1 boinc_ma boinc_pr 793 May 27 2006 COPYRIGHT -rw-rw-r-- 1 boinc_ma boinc_pr 181 May 27 2006 EXTRA_NOTE -rw-rw-r-- 1 boinc_ma boinc_pr 1721 Oct 7 23:46 README -rw-rw-r-- 1 boinc_ma boinc_pr 354 Oct 7 23:50 app_info.xml -rw-rw-r-- 1 boinc_ma boinc_pr 52779 Nov 18 23:07 arecibo_181.png -rw-rw-r-- 1 boinc_ma boinc_pr 2536 Nov 18 23:07 sah_40.png -rw-rw-r-- 1 boinc_ma boinc_pr 25488 Nov 18 23:07 sah_banner_290.png -rw-rw-r-- 1 boinc_ma boinc_pr 35399 Nov 18 23:07 sah_ss_290.png -rwxrwxr-x 1 boinc_ma boinc_pr 2808464 Oct 7 23:45 seti_enhanced-ppc-v8-g4-nographics -rw-rw-r-- 1 boinc_ma boinc_pr 71 Nov 25 18:00 slideshow_setiathome_enhanced_00 -rw-rw-r-- 1 boinc_ma boinc_pr 72 Nov 25 18:00 slideshow_setiathome_enhanced_01 -rw-rw-r-- 1 boinc_ma boinc_pr 75 Nov 25 18:00 slideshow_setiathome_enhanced_02 -rw-rw-r-- 1 boinc_ma boinc_pr 67 Nov 25 18:00 stat_icon KLS-G4:~ keithscott$ Last login: Fri Nov 23 14:37:15 on ttyp1 Welcome to Darwin! You have new mail. KLS-G5:~ keithscott$ ls -l -R -a /Library/Application\\ Support/BOINC\\ Data/projects/ total 16 drwxrwxr-x 4 boinc_ma boinc_ma 136 Nov 13 14:53 . drwxrwxr-x 33 boinc_ma boinc_ma 1122 Nov 25 23:02 .. -rw-rw-r-- 1 boinc_ma boinc_pr 6148 Nov 13 14:56 .DS_Store drwxrwxr-x 54 boinc_ma boinc_pr 1836 Nov 25 22:59 setiathome.berkeley.edu /Library/Application Support/BOINC Data/projects//setiathome.berkeley.edu: total 31008 drwxrwxr-x 54 boinc_ma boinc_pr 1836 Nov 25 22:59 . drwxrwxr-x 4 boinc_ma boinc_ma 136 Nov 13 14:53 .. -rw-rw-r-- 1 boinc_ma boinc_pr 6148 Nov 23 22:28 .DS_Store -rw-rw-r-- 1 boinc_ma boinc_pr 375304 Nov 25 12:48 01ja07aa.10440.3344.11.6.205 -rw-rw-r-- 1 boinc_ma boinc_pr 375284 Nov 25 12:46 01ja07aa.11132.3753.14.6.221 -rw-rw-r-- 1 boinc_ma boinc_pr 375406 Nov 23 11:39 21fe07ab.17874.127828.16.6.66 -rw-rw-r-- 1 boinc_ma boinc_pr 375293 Nov 22 05:20 22fe07aa.17135.8248.7.6.130 -rw-rw-r-- 1 boinc_pr boinc_pr 21317 Nov 25 19:58 22fe07aa.17135.8248.7.6.130_1_0 -rw-rw-r-- 1 boinc_ma boinc_pr 375290 Nov 22 08:08 22fe07aa.17349.11929.9.6.119 ....................... -rw-rw-r-- 1 boinc_ma boinc_pr 375303 Nov 25 21:35 22mr07ad.3603.14387.10.6.250 -rw-rw-r-- 1 boinc_ma boinc_pr 375292 Nov 25 18:46 22mr07ad.3603.4980.10.6.30 -rw-rw-r-- 1 boinc_ma boinc_pr 375293 Nov 25 22:59 22mr07ad.4306.19704.11.6.186 -rw-rw-r-- 1 boinc_ma boinc_pr 414 Oct 7 23:42 AUTHORS -rw-rw-r-- 1 boinc_ma boinc_pr 17988 May 27 2006 COPYING -rw-rw-r-- 1 boinc_ma boinc_pr 793 May 27 2006 COPYRIGHT -rw-rw-r-- 1 boinc_ma boinc_pr 181 May 27 2006 EXTRA_NOTE -rw-rw-r-- 1 boinc_ma boinc_pr 1721 Oct 7 23:46 README -rw-rw-r-- 1 boinc_ma boinc_pr 354 Oct 7 23:50 app_info.xml -rw-rw-r-- 1 boinc_ma boinc_pr 52779 Nov 22 22:49 arecibo_181.png -rw-rw-r-- 1 boinc_ma boinc_pr 2536 Nov 23 14:56 sah_40.png -rw-rw-r-- 1 boinc_ma boinc_pr 25488 Nov 22 22:49 sah_banner_290.png -rw-rw-r-- 1 boinc_ma boinc_pr 35399 Nov 22 22:49 sah_ss_290.png -rwxrwxr-x 1 boinc_ma boinc_pr 2789348 Oct 7 23:45 seti_enhanced-ppc-v8-g5-nographics -rw-rw-r-- 1 boinc_ma boinc_pr 71 Nov 25 22:59 slideshow_setiathome_enhanced_00 -rw-rw-r-- 1 boinc_ma boinc_pr 72 Nov 25 22:59 slideshow_setiathome_enhanced_01 -rw-rw-r-- 1 boinc_ma boinc_pr 75 Nov 25 22:59 slideshow_setiathome_enhanced_02 -rw-rw-r-- 1 boinc_ma boinc_pr 67 Nov 25 22:59 stat_icon KLS-G5:~ keithscott$ Previously, whenever installing the optimized app, I had got a message to say that permissions were incorrect, and that I should reinstall BOINC Manager. On doing so, all was O.K. with no further problem (except for the -5000 message on the G5, of course). Keith
	ID: 684437 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 689221 - Posted 6 Dec 2007 10:18:30 UTC - in response to Message 684302.
	Optimized apps compiled for different architectures don't differ in any way that would cause them to behave differently with regard to setting their icons, and as I discovered before, apps always try to set their icons every time they run, regardless of whether or not they've already done so in the past. That the messages went away with 5.4.9 and returned with 5.10.30 still points to the problem being a file permissions issue. One thing I can suggest at this point is to revert to the stock SETI app via BOINC, then make the permissions on the optimized app match the stock app exactly. This is actually what I've done the last few times I've installed the app on a machine. Do you know how to use the Terminal and commands like ls and chmod? You'll have more control over things (like execute permissions) this way. Perhaps I should look into whether or not I can have the icon-setting code fail silently. It seems like people are more likely to notice an error message than a missing icon. :P Alex I've used Terminal to execute the chmod and sudo commands exactly as recommended at front of Number Crunching board to "correct' the permissios, and the result, after making sure the cache held had been completely cleared and new tasks had been started, is disappointingly as follows: --------------------------------------------------------------------- <core_client_version>5.10.30</core_client_version> <![CDATA[ <stderr_txt> MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 MacOS Error -5000 occured in /Users/alexkan/seti/boinc/api/mac_icon.C line 107 OS X optimized S@H Enhanced application by Alex Kan Version info: OS X Altivec (G4/G5, G5-optimized v8-nographics) V5.13 by Alex Kan Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.400294 Flopcounter: 16433634041671.234375 Spike count: 4 Pulse count: 0 Triplet count: 0 Gaussian count: 1 </stderr_txt> ]]> ------------------------------------------------------------------- Why on Earth should this only affect my G5,and not either G4 or the MacBook Pro? All have been treated in the same way but for the differing app versions required by each. Quite apart from my annoyance that this problem is still there, I am willing to do whatever you may feel is helpful in tracking down the reason why my computer should (apparently) be unique in this respect. Should I set up a debug command (in a cc.config xml file) and, if so, which? Keith
	ID: 689221 \|

Alex Kan Volunteer developer Send message Joined: Dec 4 03 Posts: 127 Credit: 29,269 RAC: 0	Message 690822 - Posted 12 Dec 2007 7:20:04 UTC
	Alright, so after looking at your error messages more carefully and revisiting my build settings, I think I see what happened: before I released v8, I built updated 5.10.x-based BOINC libraries in an attempt to fix issues people were seeing with running at less than 100% CPU usage. In updating the Xcode projects I use to build the apps, I somehow missed the G5 build target, which means that although the G5 science app code is technically no different from the others, it uses an older version of the libraries, which are apparently prone to the issue that you were seeing. In my opinion, because this issue doesn't affect the correctness of returned results (unlike the flaws that made v7.2mb1 necessary), it doesn't merit releasing a new set of binaries, especially if doing so will cause the G5 version number to be out of sync with the others. In any case, I've already fixed my own Xcode projects to ensure that future releases should link against the right libraries.
	ID: 690822 \|

Keith Send message Joined: May 19 99 Posts: 419 Credit: 742,270 RAC: 906	Message 690890 - Posted 12 Dec 2007 17:22:20 UTC - in response to Message 690822. Last modified: 12 Dec 2007 17:32:40 UTC
	Alright, so after looking at your error messages more carefully and revisiting my build settings, I think I see what happened: before I released v8, I built updated 5.10.x-based BOINC libraries in an attempt to fix issues people were seeing with running at less than 100% CPU usage. In updating the Xcode projects I use to build the apps, I somehow missed the G5 build target, which means that although the G5 science app code is technically no different from the others, it uses an older version of the libraries, which are apparently prone to the issue that you were seeing. In my opinion, because this issue doesn't affect the correctness of returned results (unlike the flaws that made v7.2mb1 necessary), it doesn't merit releasing a new set of binaries, especially if doing so will cause the G5 version number to be out of sync with the others. In any case, I've already fixed my own Xcode projects to ensure that future releases should link against the right libraries. Alex Many thanks for taking the trouble to reply with the explanation. I had been doing a little delving and had already found that others were having the same messages, and was going to report that I had done so. Now that the reason is known, and I do not have a "problem" with my own computer, I can rest easy, and just look forward to your updated version as soon as it is available. And there was I, thinking I had not followed the instructions on fixing permissions correctly!! Double, triple and quadruple checking everything!! Keith
	ID: 690890 \|

Questions and Answers : Macintosh : ALEX KAN - MacOS Error -5000 still persistent on v8 for G5