FreeFem++-cs
an integrated environment for FreeFem++
Antoine Le Hyaric, Jacques-Louis Lions Laboratory, Pierre and Marie Curie University.

 

Frequently Asked Questions

Is there more than one FreeFem++-cs?

No, there is only one. FreeFem++-cs development started in 2004. It was first presented on the FreeFem++ web page, and then on this web site. But it is still basically the same program.

is FreeFem++-cs slower than FreeFem++?

No. Since version 3.0, FreeFem++ uses inter-process communication to display graphics, so FreeFem++ and FreeFem++-cs now have exactly the same speed characteristics. If you have a different experience on your platform, this is a bug, please report it on the bugs page.

Do I need to download both FreeFem++ and FreeFem++-cs?

No, you only need to download FreeFem++-cs to run EDP scripts. But experienced users who compile their own shared libraries with the command "ff-c++" need to download the FreeFem++ package as well.

Can FreeFem++-cs use the DLLs generated with the ff-c++ script?

Yes, FreeFem++ and FreeFem++-cs are fully compatible, even at the DLL level. Please note that Windows DLLs do not share an output stream, so the output of "printf()" and "cout" in a DLL may vary.

Why is FreeFem++-cs not included in the FreeFem++ package anymore?

FreeFem++-cs was previously part of the FreeFem++ package. But FreeFem++ changes too often to guaranty that FreeFem++-cs stays compatible with every new version. So I have created a separate, more stable package for FreeFem++-cs.

Is FreeFem++-cs a "fork" of FreeFem++?

No. FreeFem++-cs is a shell around FreeFem++. The version of FreeFem++ included in the FreeFem++-cs archive is the one developed by Frédéric Hecht.

Which OS does FreeFem++-cs work on?

I have been able to compile FreeFem++-cs on Microsoft Windows, Linux, and MacOS X. But if you want to try it on another platform, I would be very happy to help and/or to include your patches.

Is it possible to use FreeFem++-cs with any version of FreeFem++?

Unfortunately, this is not possible anymore because of technical reasons. Every new version of FreeFem++ needs to be patched to communicate with FreeFem++-cs through a TCP socket. In most cases the patches do not apply cleanly from one version of FreeFem++ to the next.

Is it possible to call FreeFem++-cs in batch mode, eg from a shell script?

From version 11.14, when called with command-line option "-Batch" and a FreeFem++ program file name, FreeFem++-cs automatically executes the FreeFem++ program and returns control to the calling shell script when the program ends. FreeFem++-cs still needs to open windows, but this can also be hidden (at least in Linux) with xvfb.

What should I do if a virus scanner reports that FreeFem++-cs contains a security threat?

To the best of my knowledge, FreeFem++-cs does not contain any virus or other security threat. The development machines are protected with a virus detector and they do not report any threat. If your virus scanner reports otherwise, it is very likely that it is a "false positive" (ie the scanner uses a "heuristic" algorithm that is not 100% reliable). Please check the suspected file through other scanners (see for instance the free online service http://www.virustotal.com) to get more information and let me know the result.

My question is not answered here. How could I let you know?

Please use the comment form below or contact me.



Add your comments

Name :

Comments :


Please copy the following numbers (sorry, this is for computer-generated spam prevention) :





Previous comments (newest first)
anonymous 22/05/12 04:41
Hi Antoine,

thanks again for the reply and sorry if I get back to you so late. Anyway yes, probably that's the problem. My domain is a pretty big rectangle, about 900x300. I will try to see how it works for a smaller domain (even though I need a pretty big one for my problem) and see how it works. Anyway please let me know if you will be able to fix the bug. Thank you again,

Tommy
antoine 10/05/12 10:30
Hi Tommy,

Does your computational domain use large values of X and Y? (of the order of 1000). If so, you have unfortunately hit a known bug. VTK seems to be clipping big objects out of the scene when displaying the picture for the first time, and I don't know how to change that yet. I am currently working on another bug (tetgen under Windows) but I will try and solve that one again just after that.

A.
anonymous 10/05/12 03:49
Hi Antoine,

thank you for the quick reply. I think I'm almost there but still there's a little problem. That is, I cannot actually see the pictures. I mean that all I see is the white background and figure's number shown in the tab increasing, but I don't see any plot. It seems that the only way I have to show the plot is to pass the cursor on the graphic window and left click where the plot is supposed to be. In this way the graphic actually appears. Then if move away the cursor, the tab number start rising again quickly (sometimes I see actually new tabs appearing, even if for just a moment) but it everything becomes white again (besides for the color bar which is always visible).
What am I missing now? Thank you very much again for your help and your work!

Tommy
antoine 09/05/12 11:00
Hi Tommy,

It is quite a reasonable request to see evolutionary problems as films. What happens at the moment is that every new figure appears as a new "tab". Because all the tabs are in the same window, each figure does appear on top of the previous one. But you still see a growing number of new tabs. If you want the previous tabs to be just forgotten, you can set "Options/Other Options/Number of Pictures" to 1.

Actually, FreeFem does the same thing: it keeps up to 10 previous pictures below the current one. But it does not display tabs so they are simply invisible :-)

Thank you for your interest in FF and FFCS.

Regards,

Antoine.
Tommy 09/05/12 10:41
Hi Antoine,

and first of all congratulations for the work you're doing with Freefem++-cs. I have a question which might sound trivial, but I just can't find the answer (also, I'm pretty new to FreeFem++ and numerical simulations in general).

Anyway.. is it possible to get the plots always in the same figure? I'm running a simulation for a reaction-diffusion system and I'd like to get the plots to be shown consecutively one over the other. So to get a sort of "smooth video" of what's going on let's say. Actually, this is just what happens when I use simply FreeFem++, and I don't understand why I can't find a way to do the same in FreeFem-cs. The possibility to get every plot in different figures is nice, but I'd like to know if there's a way to do what I've just said above.
I thought the answer might lie in the button "direct freefem++ plots to this window", but it doesn't work.
I hope the answer is not: "just use freefem++"...

Thank you, and sorry if the question sounds (or is) too stupid.

Tommy

Ps. I'm using MAC OS if it makes any difference.
antoine 13/04/12 13:53
@Luigi:

Thank you for your interest in FreeFem++!

FreeFem++ does not translate EDP scripts to a binary file because it is an interpreter (ie not a compiler). But there are several ways to run FF scripts on machines that do not accept a graphical interface. The easiest is to install the plain (text-based) FF from http://www.freefem.org/ff++ on the requested machine. Or you can run FFCS in remote server mode (the client runs on a graphics-capable machine - usually your laptop, and the server runs on the text-only machine).

Please let me know if you need any further help choosing one of these options.

Antoine.
Luigi 13/04/12 13:37
Hi Antoine.

I am a student and I am writing my thesis using your software. To implement the programs which I have written on the university hardware, I have been asked the "binary file generated by my program" (as it not possible to run them through the graphic interface but only from the executable file). Is it possible to generate it? How should I do? Thank you
Luigi
anonymous 10/02/12 23:15
Hi Antoine,
I am trying to do anisotropic mesh adaptation with bamg for file with less nombre of triangles it works, but for the file m.nopo which has the following characteristics nb triangles 3261 size of the domain 50 x 50
when i do the following command
bamg -b m.nopo -coef 1.0 -maxsubdiv 2 -AbsError -NoRescaling -NbJacobi 5 -hmax 20.0 -hmin 0.25 -anisomax 100 -ratio 0.0 -nbv 7000 -v 1 -err 0.001 -Mbb val.bb -onopo m2.nopo
I get the following error message
Fatal error in meshgenerator 113
curent line =-1
Meshinig error :Bamg
please do i have to modify the source code bamg.cpp parameters ?
Thanks
antoine 09/02/12 14:30
Hi Shyam,

The "-Batch" command-line option should work fine in version 12.1. If it does not, please let me know.

Regards,

Antoine.
anonymous 08/02/12 14:51
Hi Antoine,

The batch command is enabled in the new version 12.1 for Ubuntu 10.04.
Is there any problem with this version or is it ready for use, I meant the
precompiled version?

Thanks a lot.
Regards,
Shyam
Tomita 19/01/12 02:13
Hi Antoine,

Thank you for your quick reply.
If you have items that I can try, Please specify.
Best regards,
Tomita
anonymous 18/01/12 14:07
Hi Antoine,

Thank you for your reply.
Yes, indeed I would be very happy if you could send me the precompiled version of ffcs 11.14 Ubuntu 10.04 LTS. Also could you explain how the "\-Batch\" command works?

Thanks a lot.
Regards,
Shyam
antoine 18/01/12 11:45
Hi Tomita,

I am not sure what causes the problem that you describe. Let me try the same test as you on my own Windows system. I will get back to you as soon as possible.

Antoine.
antoine 18/01/12 11:43
Hi Shyam,

The \"-Batch\" command-line option, that forces FFCS to return after a script has been completed, has only been introduced in version 11.14. Version 11.13 will keep waiting indefinitely without returning to the calling script. Do you want me to compile version 11.14 for your system?

A Very Happy New Year to you too!

Regards,

Antoine.
anonymous 17/01/12 16:47
Hi Antoine,

I am using the precompiled version of FFCS version 11.13.
Is it possible in this version to know exactly when an FFCS program has finished running so as to kill the program in a bash script rather than using the "sleep" command as suggested by you earlier.

Thanks and have a great year ahead.

Regards,
Shyam
Tomita 17/01/12 06:45
Thank you for developing useful FreeFem++-cs.
I was using the FFCS from version 11.7, and recently upgraded to FFCS11.14, then tetgen does not work.
For example, cylinder.edp works on 11.7, and does not work on 11.13 & 11.14.
........
tetgen: before tetrahedralize( , &in, &out);
numberof regions 1
numberof hole 0
........ ---> no reply

What causes this? Japanese Windows System? I need a Windows 7?

---- My PC conidition:
FFCS & FF++ binary packages on Windows XP (Japanese System)
C:/tools/freeFEM/FreeFem++ (FreeFem++3.17)
C:/tools/freeFEM/FreeFem++-cs-11.7
C:/tools/freeFEM/FreeFem++-cs-11.13
C:/tools/freeFEM/FreeFem++-cs-11.14
FFCS is Sequential Server mode

And I feel the same behavior on other my Vista PC.
antoine 01/12/11 14:58
Hi Shyam,

I would be very happy if I could be of any help to you. See you next week.

Best regards,

Antoine.
anonymous 29/11/11 10:59
Hi Antoine,

I would be attending the workshop on FreeFem++ which is to be held during this December5-7. Will it be possible if I could meet you and discuss some problems that I have with you in person during sometime that is convenient as it would be a great help for my PhD?

With sincere gratitude,
Shyam
anonymous 22/11/11 13:29
Hi Antoine,

Thanks for the help. It does work provided I give a good guess
for the time needed by the computation to finish.

Anyways, I would be waiting to hear from you when you have implemented
the option that returns control to the script as that would be more efficient.

Thanks.
Shyam
antoine 17/11/11 16:41
Hi Shyam,

Unfortunately, there is no easy way for the bash script to know when the FFCS computation has ended. So I need to implement an option for FFCS to return control to the script by itself when it is finished. I'll let you know as soon as it is ready. As a workaround, what you can do now is:

${FREEFEM} check.edp > check.out & (<- notice the '&' sign)
sleep N
kill %%

where N is the maximum time that FFCS takes to run in seconds (of course it is not ideal since all FFCS runs do not take the same amount of time).

Regards,

Antoine.
anonymous 17/11/11 15:50
Hi Antoine,

Thanks for the tip regarding UMFPACK. I have one question:
If i launch a ffcs simulation using a bash script, say for example
FreeFem++-cs check.edp, even when the program is done, the bash script doesn\'t
understand that the program is done and stays with the same window even when the ffcs script is finished running, i.e, even after Bien: On a fini Normalement. How can I tell the bash script that the program is over and you can close the window and go on to the next computation?

I hope I have been clear with my question.
The bash script that I\'m using is

FREEFEM=/home/shyam7sunder/Desktop/FreeFem++-cs10.16/FreeFem++-cs
date
echo \"now doing: $FREEFEM check.edp\"
${FREEFEM} check.edp > check.out
echo \"done\"
date

It always stays on the line ${FREEFEM} check.edp > check.out.

Thanks in advance.

Regards,
Shyam Sunder
antoine 17/11/11 10:40
Hi Shyam,

I am happy to know that it works for you. I believe that Mumps and other external libraries will compile better with FreeFem++ 3.17, so I will try and include it in FFCS as soon as possible.

Umfpack requires a noticeable amount of extra memory compared to iterative methods like CG and GMRES. If your computer is short of RAM, you may have to change solvers. If your computer has lots of RAM and Umfpack still does not run properly, you can first try to use the UMFPACK64 load module from FreeFem.

bash scripting: if you specify a full path to the FFCS executable (eg \"/home/user/Desktop/FreeFem++-cs 11.12/FreeFem++-cs\"), FFCS should run without any interference with any other version of FF or FFCS. Please let me know if it doesn\'t. AutoRun is not available at the moment, but I will add a command-line option for it. And you should already be able to get rid of the FFCS window with xvfb (although I am not sure about OpenGL). Would you be interested to try?

Thank you for your interest in FFCS.

Regards,

Antoine.
anonymous 16/11/11 18:21
Hi Antoine,

Thank you very much for the precompiled version. The MPI Parallel Solver
works fine. I guess in the precompiled version some libraries like
Mumps, Pastix etc. are missing but the important thing is that the mpi is working.

I have a couple of questions to ask you and would be happy if you could help me.
1. When I do stability calculations, for some parameter conditions I get this as the output:
umfpack_zi_numeric failed
current line = 109
Exec error : umfpack_zi_numeric failed
-- number :1
Exec error : umfpack_zi_numeric failed
-- number :1
err code 7 , mpirank 0
ffc: error: FreeFem++ returned error 7

If I reduce the size of the matrices involved (by reducing the grid resoultion) this seems to work. At the sametime, for the same grid resolution with a different paramter setting it doesn't work. Could you tell what is happening with this?

2. Secondly, if I have to run the FreeFem++-cs simulations for a large paramter range using /bin/bash scripting, I need to call FreeFem++-cs from the shell. But how can I pick the correct version?? Also, is it possible to do 'Auto Run' without having a window open as I have large number of calculations.

Thanks a lot in advance and thank you once again for the version 11.12.

Regards,
Shyam
antoine 14/11/11 16:40
Hi Shyam,

A precompiled version of FFCS with MPI for Ubuntu 10.04 64 is now available at

http://www.ljll.math.upmc.fr/lehyaric/ffcs/news/20111114-1549.php

Could you let me know if it works for you?

Thank you in advance,

Antoine.
anonymous 10/11/11 15:16
Hi Antoine,

Thanks again for the reply. I would wait for the source to be released officially by you in this website (hopefully before the end of next week). I am not in a hurry and I am currently happy with my precompiled freefem++-cs. My version of Ubuntu is 10.04 64 bit LTS. I hope it would be stable this time around.

Thanks a lot for freefem++-cs and the answers.

Shyam Sunder
antoine 10/11/11 15:05
Hi Shyam,

Parallel computing has been disabled in many precompiled version of FFCS because of problems with multiple MPI versions (basically, one needed to choose one version of MPI before compiling FFCS). But I hope to enable MPI by default for the latest Ubuntu (11.10). The regression tests have passed on Debian. Now I just need to run them on Ubuntu and create a package.

Sorry for all the waiting. If you are interested, I could give you access to the current source for you to compile and check if it is stable enough for you.

Best regards,

Antoine.
anonymous 10/11/11 14:46
Hi Antoine,

Is there some news with the new version???

And in a precompiled Freefem++-cs how can I implement parallel
computing???.....I would wait for the new version in any case but I just
wanted to know how it can be done.

Thanks a lot.

Shyam Sunder
antoine 02/11/11 11:44
Hi Shyam,

As is often the case, the new version was delayed because of a bug that I had quite a difficult time to solve. But yes, the new version is on its way. At the moment, it passes 33 regression tests out of 400, so it is not quite ready yet. Sorry for the unexpected delay.

Best regards,

Antoine.
anonymous 02/11/11 11:11
Respected Sir,

Is there any news on the version 11.12 yet?

Thank you.

Shyam Sunder
Shyam Sunder 24/10/11 10:03
Bonjour Antoine,

Thank you for your message and I am very grateful to you for taking your time off for replying. I am using a Linux Ubuntu 10.04 64-bit platform. When I use the pre-compiled version of 11.7, I still cannot choose the Parallel-Solver from the Options Menu. Is there something that I can do about it? Or should I compile 11.7 from the source too?

With sincere gratitude,

Shyam Sunder
antoine 21/10/11 20:00
Hi Shyam,

You do not disturb me at all. I need input from FFCS users.

The main difference between 11.7 and 11.12 will be an up-to-date FreeFem++ (version 3.16). 11.12 may integrate MPI more easily, but I cannot be sure yet (MPI is very tricky on the Windows platform).

So my advice would be to stick with an easy-to-install precompiled 11.7 until next week, and then try 11.12.

I will get back to you as soon as I know more about the MPI status.

Best regards,

Antoine.
Shyam Sunder 21/10/11 16:28
Hi Antoine,

Thanks for your quick reply. I guess I would install the version 11.7, ffcs 11.7.

Do you recommend that or it is better to wait for 11.12?...

Thank you once again and I am sorry for disturbing you so much.

Regards,
Shyam
antoine 21/10/11 15:44
Hi Shyam,

Sorry for all the trouble. Version 11.12 will be tested on all platforms so it should be a lot more stable than 11.11 (which was only tested on one platform). Would it be ok for you to wait for the new version? (hopefully it will be out by the end of next week)

Best regards,

Antoine.
Shyam Sunder 21/10/11 14:57
Hi Antoine,

I am still stuck with upgrading to ffcs 11.11.
When I run the make file, it shows
error: freeyamslib.h: No such file or directory.

And from the onwards, it shows some other errors, but I guess it is related to the
previous one. Could you throw some light on this regard please.

I thank you once again for the previous message. Now at least I can download the
libraries from the server.
With sincere gratitude,

Shyam Sunder
antoine 20/10/11 13:45
Hi Shyam,

Thank you for your message. It often happens that some required libraries change versions, and the previous version then fails to download properly. I will check that now and create a new package.

Regards,

Antoine.
Shyam Sunder 20/10/11 11:08
Hi Antoine,

Sorry to bother you again. When I compile ffcs 11.11 from the source file,
and then do the 'make' command, it needs to download the libraries like Freeyams, mmg3d etc from the upmc.fr server. I guess the server is down cause its not able to download the libraries. Could you please check that?...

Is it possible to use the libraries from an older version to get around this or I wait for the server to start functioning again?...

Thanks a lot and have a good day.

Regards,
Shyam Sunder
Shyam Sunder 18/10/11 19:11
Dear Antoine,

I have been using FreeFem++-cs 10.16 and I recently updated to FreeFem++ 11.7.
I want to parallelize my code and for that in the 'Options' menu, I need to select Parallel SOlver instead of Sequential Solver right?...

But I am not able to select that option and hence all the mpi commands are not understood by the FreeFem++-cs11.7 when I run a program.

I would be grateful if you could help me with this problem.

With sincere gratitude,

Shyam Sunder
antoine 24/03/11 11:58
BAMG is not included in FFCS at the moment because I have not had a request for it yet. Please let me know how you want to use it (do you want to call it from EDP scripts or in a separate command window?), and I will package it.
smail 23/03/11 18:42
please i try to excute BAMG from Freefem++cs but it does not work
please how it cal work?
antoine 07/12/10 12:24
@ed:

It looks like border point c6(t=5) does not match c7(t=0) but c7(t=5). It could explain why FF says that one border is in the wrong way ("sens" in French). Please let me know if this solves your question.
ed 06/10/10 22:08
Hello there,

First of all thank you for developing such a valuable freeware-tool!
It looks like I encounter a problem which is not described anywhere.

I actually have a problem with the borders of my configuration. They are not recognized properly by the mesh generator, although this kind of script seems to be working on other computers. The error message is following:

------------------------------------------
Error in the def of sub domain 2 form border 6/8: Bad sens 16 1
------------------------------------------

and also :

------------------------------------------
Meshing error: Bamg
number : 777,
err code 4 , mpirank 0
ffc: error: FreeFem++ returned error 4
ffc: error: feedbuffer(): zero-sized chunk
------------------------------------------

FYI the borders are defined as follows (no overlapping exept at end points):
------------------------------------------
border c1(t=0,5){ x=0; y=t ; label=1 ;};
border c2(t=0,2.5){ x=t; y=5 ; label=2 ;}; // h = 15
border c3(t=2.5,5){ x=2.5; y=t ; label=1 ;};
border c4(t=2.5,2.7){ x=t ; y=2.5 ;label=1 ;};
border c5(t=2.5,5){ x=2.7; y=t ;label=1 ;};
border c6(t=2.7,5){ x=t ; y=5 ;label=3 ;}; // h=0
border c7(t=0,5){ x=5 ; y=t ;label=1 ;};
border c8(t=0,5){ x=t ; y=0 ;label=1 ;};

mesh Th = buildmesh(c1(6)+c2(6)+c3(6)+c4(4)+c5(4)+c6(4)+c7(6)+c8(6));
plot(Th, wait=1);
------------------------------------------

I can't understand why it does not work. Do you possibly have an idea? My OS is Windows Vista 32bit.
Thank you for your answer!
antoine 16/07/10 21:03
@anonymous:

No, sorry, MPI is not included in FFCS yet. It will hopefully be done sometime this autumn.
anonymous 16/07/10 20:10
Is MPI feature implemented ?
vinhvl 06/07/10 08:47
ko la ji
antoine 25/06/10 23:32
@Robert:

Here are the file sizes and checksums for the latest version:

>du -b ffcs-10.13*
24544393 ffcs-10.13-Linux32.tgz
25981155 ffcs-10.13-Linux64.tgz
28198010 ffcs-10.13-Mac64.zip
27003318 ffcs-10.13-MacIntel.zip
63322770 ffcs-10.13.tgz
25360127 ffcs-10.13-Ubuntu104.tgz
29420301 ffcs-10.13-Ubuntu804-64.tgz
27229853 ffcs-10.13-Ubuntu804.tgz
17362307 ffcs-10.13-Windows.exe

>md5sum ffcs-10.13*
67389b83ba8c10aab312b13aa2ee932e ffcs-10.13-Linux32.tgz
3132285ddc9ae7b6262a2e15ba81a9ea ffcs-10.13-Linux64.tgz
5cbb7f56b27300254f36243ca09af899 ffcs-10.13-Mac64.zip
6f3cedfc563c94ceca947839cff39352 ffcs-10.13-MacIntel.zip
2fde94c0e1f5cf025179eb9b13a21b65 ffcs-10.13.tgz
345126e646d176b8faa7a195564ac882 ffcs-10.13-Ubuntu104.tgz
ea7dcfda6ee999b65dcb14e74b01a409 ffcs-10.13-Ubuntu804-64.tgz
55931b83121a416855cac5c639d2d67b ffcs-10.13-Ubuntu804.tgz
28c56a0f23998ab136b6574b3e5fb658 ffcs-10.13-Windows.exe
Robert in Oxford 23/06/10 23:57
My installation does not work - crashes as the windows are forming. The first possibility is that the file didn't download perfectly. Could you privide exact file size (number of bytes) and MD5 sums for download validation.
antoine 09/06/10 23:20
@anonymous 9/6/10 21:48

Unfortunately, if it says "cannot execute binary file" on mandriva 2008, it means that you need to download the source package and recompile it. Everything should be fine after that. FreeFem++-cs runs independently, whether or not FreeFem++ is installed on the machine. If there are more requests for mandriva, I will try and create a specific binary download for it.
anonymous 09/06/10 21:48
I run FreeFem++-cs, but send me the message: cannot execute binary file. This for mandriva 2008, with mandriva 2010, all ok! I need install something? I have install FreeFem++ in both systems. And I install FreeFem++-cs in mandriva 2008, but no work either.
anonymous 13/05/10 23:09
Yes, it does. Thank you very much for your quick replay.
antoine 13/05/10 23:01
@anonymous:

Thank you for the bug report.

For some reason that I do not fully understand yet, some Windows machines complain about a missing mingwm10.dll, and some don't. I will look into it as soon as I can. In the meantime, as a workaround, you can try to copy the file "mingwm10.dll" from "c:\Program Files\FreeFem++-cs 10.7\Contents\Windows" to "c:\Program Files\FreeFem++-cs 10.7".

Please let me know if it solves the problem for you.
anonymous 13/05/10 22:24
When I start FreeFem++cs-10.7 on Windows XP I get message:
"This application has failed to start because mingwm10.dll was not found. Re-installing the application may fix this problem." I have already installed FreeFem++cs-9.11. and I haven't had any problem.
antoine 04/01/10 10:37
@anonymous:

To uninstall and reinstall FreeFem+-cs, please follow the guidlines from http://www.ann.jussieu.fr/~lehyaric/ffcs/news-20090518-1620.php and let me know if log and 3d still pose problems after that.

Regards,

Antoine.
anonymous 01/01/10 16:10
Could any one guide please.
log file is not appearing after running a FreeFem++ program on linux ubuntu why ??


How can I un-install freefem++ from ubuntu and reinstall again.

How can i get support for proper 3D inimation on ubuntu 9.04, my processor in intell 32 bit.
antoine 26/10/09 15:29
@Vincent Laude:

Thank you for the bug report. Could you please try and run the command "xlogo" in the same way that you run "FreeFem++-cs" and let me know if you see the same error message?

Regards,

Antoine.
Vincent Laude 07/10/09 18:09
I have tried the 9.10 version with my OpenSuSE 11.1 / gnome linux. When running FreeFem++-cs I get the error:
>No protocol specified
>Can't open display: :0.0
and the program crashes.
This is solved by typing
> xhost +
before running the program, but you may want to correct this in a forthcoming version?
Thanks for the excellent program!