log in
1) Questions and Answers : Wish list : Fix to scheduler_reply needed (Message 1068)
Posted 27 Oct 2008 by John McLeod VII
THe scheduler reply XML file has the following invalid XML in it:

[quote]
<file_name>mm_ProjectIcon_BETA_01.png</filename>

please note that <file_name> does not match </filename> and causes the BOINC clients to crash.

I am not certain where the scheduler reply is generated. I am certain that this needs to be fixed.
2) Questions and Answers : Windows : Screen Saver Problem. (Message 775)
Posted 29 May 2008 by John McLeod VII
For some reason, I am getting the following Debug Assertion Failure in the 6.2.4 BOINC screen saver when Mind Modeling is supposed to start displaying graphics.



“Debug Assertion Failed

Program …

Dbgdel.cpp

Line 52



Expression _BLOCK_TYPE_IS_VALID(pHead->nBlockUse)”



This occurs when the “BOINC Screen Saver” message ends, and the screen saver is supposed to start.



Currently running Projects:



Mind Modeling – ACT-R cognitive modeling environment 3.00

Climate Prediction – UK Met Office HadSM3 Slab Model 5.06

A different time that had the same diagnostic error:

Mind Modeling

Milky Way.

A third time that did NOT have the diagnostic error

Milky Way.
CPDN.
3) Message boards : Number crunching : Migrating form 5.10.45 to 6.2.1 (Message 716)
Posted 8 May 2008 by John McLeod VII
I am on the Alpha test team for BOINC. I noticed that Mind Modeling does not seem to behave very well on the upgrade from 5.10.45 to 6.2.1. In particular, it seems that the lisp subsystem has some dependency on where it is installed (it can be instlled anywhere, but cannot be moved). BOINC 6.2.1 splits the BOINC executables from the BOINC projects and data. The projects and data move to a new location on the hard disk. Mind modeling refuses to run until you delete the lisp directories in the Mind Modeling project directory and abort any running Mind Modeling tasks.

I left Mind Modeling running for hours with no progress after the move from 5.10.45 to 6.2.1. I would hope that some way of getting Mind Modeling to self correct would be possible.

Please note that BOINC 6.2.1 is not ready for distribution yet as we are still chasing some bugs out.

John,
Thanks for the insight on this.

I will take a look at this before I release the new version and hopefully modify things in such a way that it can operate in both environments.

Cheers,
Jack

It works in both environments. The problem is the transition. However, I may have mis-spoken. I got the second transition to work merely by aborting all running tasks. The following ones then started correctly.
4) Message boards : Number crunching : Migrating form 5.10.45 to 6.2.1 (Message 712)
Posted 7 May 2008 by John McLeod VII
I am on the Alpha test team for BOINC. I noticed that Mind Modeling does not seem to behave very well on the upgrade from 5.10.45 to 6.2.1. In particular, it seems that the lisp subsystem has some dependency on where it is installed (it can be instlled anywhere, but cannot be moved). BOINC 6.2.1 splits the BOINC executables from the BOINC projects and data. The projects and data move to a new location on the hard disk. Mind modeling refuses to run until you delete the lisp directories in the Mind Modeling project directory and abort any running Mind Modeling tasks.

I left Mind Modeling running for hours with no progress after the move from 5.10.45 to 6.2.1. I would hope that some way of getting Mind Modeling to self correct would be possible.

Please note that BOINC 6.2.1 is not ready for distribution yet as we are still chasing some bugs out.




Main page · Your account · Message boards


Copyright © 2020 MindModeling.org