Some time back I was working with my 64 bit version and it shut itself down and I was unable to get it to open the model again. So I switched over to the 32 bit and after several tries the modle was revived. Recently I started getting error messages again. one said that I might want to try to debug. I hit the OK. The part of my model that converted daily data to monthly averages was horribly modified. These changes occurred in seveal places within the model as well as previous versions and in a word document that coatained the code. I corrected the bad code in one place and restructure my external data sets. Saved everything and noew I am having problems opening the model. I have attached the error messages that I get below. I'm really frustrated and puzzled. Can you tell what these mean and if you understand my problem?
Main menu
You are here
Error messages
Tue, 06/09/2011 - 15:43 — John E. Rogers04
The compiler raised a problem with the code generated for this model.
It may help to try the 'Debug' option.
The error was:
C:\DOCUME~1\jroger04\My Documents\My Simile files\sim813000\seagrasstesta5lconditional4>set PATH=C:\Program Files\Simile5.8\System\bin
C:\DOCUME~1\jroger04\My Documents\My Simile files\sim813000\seagrasstesta5lconditional4>g++ -Wno-trigraphs -m32 -c -o objtmp.o -I"C:\PROGRA~1\Simile5.8\Run" -I"C:\Program Files\Simile5.8\System\include\mingw" model.cpp
C:\DOCUME~1\jroger04\My Documents\My Simile files\sim813000\seagrasstesta5lconditional4>g++ -shared -o model3.dll objtmp.o
model.cpp:7818:37: warning: unknown escape sequence '\l'
model.cpp:7818:37: warning: unknown escape sequence '\T'
model.cpp:7818:37: warning: unknown escape sequence '\M'
model.cpp:7818:37: warning: unknown escape sequence '\w'
model.cpp:7818:37: warning: unknown escape sequence '\B'
model.cpp:7839:28: warning: unknown escape sequence '\l'
model.cpp:7839:28: warning: unknown escape sequence '\T'
model.cpp:7839:28: warning: unknown escape sequence '\M'
model.cpp:7839:28: warning: unknown escape sequence '\w'
model.cpp:7839:28: warning: unknown escape sequence '\B'.
Press "Help" to display a relevant page from Simile's documentation.
and
Simile encountered an unexpected problem:
invalid command name ".progress.message"
The error was:
invalid command name ".progress.message"
while executing
".progress.message configure -text [eval [list format $msgs($key)] $lits]"
(procedure "FillProgressBox" line 4)
invoked from within
"FillProgressBox pl_sort {}"
invoked from within
"catch $args retVal"
Press "Help" to display a relevant page from Simile's documentation.
Reporter:
John E. Rogers04
Created:
Tue, 06/09/2011 - 15:43
Updated:
Fri, 09/03/2012 - 07:19
Comments
#1
Hi Jon,
It is a very long time since I heard of anyone losing model data due to a crash -- the auto restore feature works very well. Don't know why it failed on this occasion.
The 64- and 32-bit versions should be exactly the same in terms of saved file formats. Are you saying that you now cannot run the 64-bit version at all?
If an error message suggests you try the 'debug' option, hitting OK does not do anything. You have to select the Debug option from the Model menu to use this feature. It is most useful when a model is producing SIGSEGV errors.
I don't understand how Simile can cause changes in a Word document. It cannot open one. Is that really what you are saying?
These are interesting error messages. I guess they happen every time you try to run the model. It would be useful to have a copy of the model to see what is causing them. Can you send one to me at support please?
--Jasper
#2
My models are to large for this format. We used a different mechanism of transfer some time ago. I would be glad to use that.
#3
Please mail to support@simulistics.com
--Jasper
#4
OK, I received the model, and I have found the cause of the compilation bug. The submodel "Boundry Conditions" and the variable "TSS array" have comments containing Windows-style directory paths with backslashes. Simile makes all data, including comments, available in the executable model, and the backslashes are causing syntax errors in the compilation stage. This is a bug in Simile and will be fixed in the next release.
In the mean time you can work around the problem by changing the directory paths in comments to Unix style with the directories separated by forward slashes. These do not cause problems for the compiler.
I will now have a look at the behaviour of the model.
#5
Yesterday I copied the code for month days to a Word file and then using find and replace exchanged PHYTON_D for SALINE_D and then inserted it in salinity monthly total. When I tried to compile the system hit a point where error appeared (something like seperation failure for the prolog code and the system crashed and closed down. I will check on your recent comment, thanks.
#6
Making your suggested changes allowed me to compile and correct the code for flow in PHYTON submodel. It now runs correctly. So when I was making changes in the code and thought it was compiling it wasn't which led me to thinking something was wrong with the way it was running. It just wasn't compiling so the same code would run without changes. I think this was meant to drive me crazy.
Now to the copy and paste problem. The errors that appear are "segmentation violation and "file 16ead 90" cannot find chanel named.
can not find channel named "file16ead90"
can not find channel named "file16ead90"
while executing
"eof $plPipe(stream)"
(procedure "send_pl_cmd" line 8)
invoked from within
"send_pl_cmd call:$plCmd"
(procedure "prolog" line 5)
invoked from within
"prolog tk_abandon"
(procedure "AbandonObj" line 2)
invoked from within
"AbandonObj"
(command bound to event)
#7
I have fixed all the current problems. Before when I had not fixed the comments and was trying to compile a comment would come up saying a coding error had been detected and suggested that I might like to debub. I said OK and the drop down went away i then hit the grey triangle to compile and the model appeared to be doing just. But it wasn't so any changes that I made were not compiled into the model and it always ran on old code. Therfore I was trying every thing to fix the problem and nothing worked. Now that we have changed the comment sections and I made some corrections everything is going fine at the moment. So thank you very much for your help in the matter. John
#8