[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [FEMM] Updated development version


--- Dave Squires <djsquires@xxxxxxxx> wrote:

I found a bug in the code. [...] For some reason FEMMVIEW uses the original geometry from the first run. You don't see the change in FEMMVIEW.

The way that the multiple document interface program that you get out of Visual C++ works is that if you try to open a document that has the same name as one that's already open, it just brings up the window of the open document instead of reloading the document from disk. Now, the program really did the run and has the right results on disk, it just hasn't loaded them. You don't have to shut down the program and start it back up to get the results--just select File|Reload from the main menu of the "stale" postprocessor window, and that forces a reload.

I _thought_ that I had pretty much eliminated this behavior on the latest-n-greatest development version (01Apr04). I fixed it so that clicking on the "glasses" icon in the preprocessor forces a reload if there is already some version of the answer already open. However, I hadn't really thought about the entry on the MRU files list--I ought to see if I can force a reload if an entry off the list is selected. However, I think that I want to keep it so that if you select a window from the Window menu entry, that it doesn't automatically force a reload.
Wijono wrote:

Dear David,
In addition to Dave Squires problem, if we close
(exit) the FEMM, it exits without any question.
Does it mean that FEMM saves everything automatically.
If not, is the changing (in Dave problem) not saved by
FEMM (just saved in buffer)?
Many thanks.

Regards,
Wijono


In Dave S's problem, the file does get saved, and the answer new answer is there on disk -- it just didn't get loaded until the user manually force a reload with File|Reload (or by manually closing the .ans window and opening it up again from the MRU list).

I know that the program doesn't prompt you to save your work before you quit, and that it saves your work to disk without prompting during a mesh or an analysis. This isn't really "standard" behavior for a Windows program, and I probably ought to get around to change this behavior to forestall the occasional questions about it.....

Dave.
--
http://femm.foster-miller.com/~dmeeker