Problems with COCO 3
Moderator: jasper
Problems with COCO 3
I assumed that I could switch from COCO 2 to COCO 3 without a translation time penalty. If I have to consume time translating COCO 2 fsd files for use in COCO 3, then perhaps I can save time by just continuing to use COCO 2. Why move to COCO 3? and face a translation time penalty?
Re: Problems with COCO 3
I have no yet found a problem with any file between COCO 2 and COCO 3. You are the first to report it. It would help if you can send me the file so that I can assess what the problem is.
Problems with COCO 3 (6)
OK, I now understand your position. Thanks for the explanation. I will do my best to provide you with a shareable fsd file exhibiting the translation problem. I have no idea when or if this will occur. Until then, thanks for your patience.
Problems with COCO 3 (7)
With this posting, I am pleased to report different, better results from my original posting. I am running COFE 2.7.0.23 (C2) along side COFE 3.0.0.19 (C3). The main difference is that FS4.3.1.fsd (FS) solved in both series of runs.
The testing consists of two series of runs. The first test began with a fresh computer start. FS was loaded and solved twice in C2, saved, then FS was loaded into C3 and saved with a different name, then FS was solved twice in C3. The results are reported below.
In summary, I only have two issues, Item 2 and Item 4, so far. Sorry for the previous erroneous posting, but this posting should correct it.
The testing consists of two series of runs. The first test began with a fresh computer start. FS was loaded and solved twice in C2, saved, then FS was loaded into C3 and saved with a different name, then FS was solved twice in C3. The results are reported below.
- 1) FS solved in C2 in 2s, 453ms. In the second run, FS solved in C2 in 2s, 63ms. FS was saved with all streams GREEN.
2) FS was loaded and an attempt was made to change its name to that of an existing file in the folder. Stream names were duplicated in FS so that was corrected, but during the attempted save, C3 deleted the existing file during a [Save As] attempt. Can this be fixed as it is not good behavior?
3) The streams were all GREEN when FS was loaded into C3. FS solved in C3 in 6min, 21s, 0ms. In the second run, FS solved in C3 in 22s, 688ms.
4) When FS was saved in C3, the streams were all BLACK except those not used in the simulation, those remained GREEN. Is BLACK the new GREEN? Can this be fixed?
In summary, I only have two issues, Item 2 and Item 4, so far. Sorry for the previous erroneous posting, but this posting should correct it.
Re: Problems with COCO 3
2) I cannot reproduce this behaviour: it should not delete the file until the new file has been succesfully saved and merely needs to be copied in place. I will need a recipe to reproduce this.
4) Black means fully specified, but unsolved. Green means solved. If the streams to do not turn green, this is a bug. Can you send me some document that reproduces this?
Also if the solution time is larger than in COCO 2 I would like to investigate. Solution time is typically considerably smaller in COCO 3. I suspect an issue with the problem statement may be at the root of it. Please also try when setting the amount of threads to 1 (Edit -> Preferences) to avoid the overhead of creating the PMCs in all threads (should not be anywhere near 22 s though) and out of memory conditions.
4) Black means fully specified, but unsolved. Green means solved. If the streams to do not turn green, this is a bug. Can you send me some document that reproduces this?
Also if the solution time is larger than in COCO 2 I would like to investigate. Solution time is typically considerably smaller in COCO 3. I suspect an issue with the problem statement may be at the root of it. Please also try when setting the amount of threads to 1 (Edit -> Preferences) to avoid the overhead of creating the PMCs in all threads (should not be anywhere near 22 s though) and out of memory conditions.
Re: Problems with COCO 3 (8)
Here is my list of problems which is to be updated periodically.
Re: Problems with COCO 3
It would really help if you can send me a document or recipe to reproduce the issues.
Problems with COCO 3 (9)
Hopefully the 105+ Forum members will contribute their time to provide files to use to help complete COCO 3 functionality. The members should chose an interesting problem from the list, then work with it until it can be resolved. Maybe a contest would be a good idea to interest members in participating. No reason for a few members to have all of the fun!
Re: Problems with COCO 3
I get problem reports and functionality requests on a weekly basis. Typically with fsd or recipe to reproduce, and typically these issues are fixed within a day.
And when it comes to an update the magnitude of going from COCO 2 to COCO 3: can't make an omelet without breaking eggs. More than 10 thousand lines of new solver code. There is bound to be a few bugs in there. I fix them pretty much as fast as they are identified (and reproducible).
And when it comes to an update the magnitude of going from COCO 2 to COCO 3: can't make an omelet without breaking eggs. More than 10 thousand lines of new solver code. There is bound to be a few bugs in there. I fix them pretty much as fast as they are identified (and reproducible).
Re: Problems with COCO 3
Dear All,
I'm a forum member during a longer time, and contribute to it on a regular basis.
Keeping in mind this is a voluntary contribution of Jasper to all of us, I want to thank Jasper for this once again.
Jasper, keep up the good work!
For a while however I am pretty much annoyed and struck by the rude and demanding way nrgeng is complaining about a free and in my opinion excellent program, and the swift way Jasper is still trying to solve all of his (and our) issues.
So: nrgeng: please behave, and provide Jasper with the reasonable things he is asking. Or otherwise: stop using the program and pay for a commercial product.
Thanks,
Henk Fikkert
I'm a forum member during a longer time, and contribute to it on a regular basis.
Keeping in mind this is a voluntary contribution of Jasper to all of us, I want to thank Jasper for this once again.
Jasper, keep up the good work!
For a while however I am pretty much annoyed and struck by the rude and demanding way nrgeng is complaining about a free and in my opinion excellent program, and the swift way Jasper is still trying to solve all of his (and our) issues.
So: nrgeng: please behave, and provide Jasper with the reasonable things he is asking. Or otherwise: stop using the program and pay for a commercial product.
Thanks,
Henk Fikkert