[gschaider]

Look: try 'which as' to check whether the binutils are correctly installed.

BTW: it's much easier to use OpenFOAM on Linux (everything you need is already installed)

[look]

Dear Bernhard Gschaider:

Thank for your help.
I kown what is wrong.

Look

[liuzhw]

Hello,
Thank OP groups for the nice CFD codes.
I install the foam-1.3 on the cygwin operating system, and it is run well.
now, I want to run the complied solver on the compute without installint
cygwin by copy some dll files into one fold. there are some thing wrong,

/*---------------------------------------------------------------------------*\
| ========= | |
| \ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \ / O peration | Version: 1.3 |
| \ / A nd | Web: http://www.openfoam.org |
| \/ M anipulation | |
\*---------------------------------------------------------------------------*/
fileName::stripInvalid<> called for string /cygdrive/c/DocumentstandSettings/Administrator
3 [sig] water2D 1856 open-stackdumpfile: Dumping strack trace to water2D.exe.sackdump

who can help me, thanks advance!

[pvita]

You need to set some enviromental variables as in Linux.

[irc]

Brooks has provided a comprehensive set of patches etc., with the help of Petr & Niklas, for v1.3. How simple would it be to update these to a new beta version of OF? Were the patches created semi-automatically (i.e. relatively easily), or is it a question of editing each patch manually (i.e. long-winded!)?

This is in reference to Hrv's development version of the code, which includes some additional dynamic mesh goodies that aren't present in v1.3.

Ian: Unfortunately, they were only created semi-automatically insofar as my text editor has a "search and replace on disk" feature that I used for some of the file renaming. After I finished the 1.3 port, I put "write scripts to do this automatically" on my to-do list, but it's still there.

The vast bulk of the work was involved in taking care of the case-sensitivity issues. This required renaming the relevant files (a couple of dozen or so, if you ignore the non-ported FoamX stuff) on a Linux machine, copying the tree to a Windows machine, and going through and search-and-replacing all the #include lines for those files. As I recall, that took about three or four hours, including some trial-and-error finding files I'd neglected to rename the first time.

After that, there was porting Petr's build system over, which was rather straightforward (and should be identical in the new version, I'd think, so you can probably reuse the old patch).

Then, many of the Make/options files in the source need to be updated to reference additional libraries to build on Windows. This, again, was rather a bit of a trial-and-error process, taking about a day (but much of it doing other things while waiting for things to build).

If the new development version hasn't changed too much in the overall structure, though, it may be relatively easy to take the diff files for the v1.3 and development code, and edit them a little bit to apply to the ported v1.3 code -- that's certainly how I'd try to do it! The only thing that really should need to be changed are the filenames in #include lines and occasional Make/files files, and the changelog.cygwin file in the port has a complete list (unless the dev code adds more). I couldn't do that going from v1.2 to v1.3 because so much of the core OpenFOAM source was moved into the FiniteVolume library, which affected everything.

Feel free to email me if you'd like, or ask more questions here. Also, the changelog.cygwin file has a bit more explanation of exactly what changes were made, which may be useful.

[hjasak]

The things is that the development version has got the same structure as 1.3 but there is lots of new stuff. This includes bug fixes (e.g. moving wall calculations don't work properly with the release), new libraries of moving meshes, topo changes and mesh deformation free surface solvers, examples of top-level codes together with tutorials, new boundary conditions and lots of bug fixes.

I know Ian needs this (been in touch) but I don't do any work on the Windows (allergical reaction) :-) If anyone wants to have a go, I'll point you to the snapshot + provide background info as requested. I don;t think there are additional name clashes and a set of scripts that would do the job would be very welcome.

Hrv

[irc]

Sounds like a project for when I have a little more time. Thanks for the offers Brooks & Hrv; I will be back in touch about this next month when hopefully things have quietened down a little!

[waynezw0618]

dear everyone,
when i run ./Allrun in IcoFoam,i have problem like:
block mesh:not found
runing icoFoam on cavity
IcoFoam:not found
case already cloned:remove case directory Cavity fine to clone
sed:cannot read cavitifine/constant/polyesh/Blocakmeshdict:no file or directory
mv cannot regular file "cavitifine/constant/polyesh/Blocakmeshdict":no file or directory
...
...
what is wrong with it?
how could i do?
thanks

[luca_g]

I'm using the cygwin port of OpenFOAM 1.3.
It seems to me that solvers (as well as other OpenFOAM executable) run at most to 50% cpu usage whereas a completely different application (compiled with gcc) can run up to 100% cpu.

Does anybody have an idea why ?