[Gmsh] Socket port parsing bug ?

[Pierre Boulenguez]

Hi Christophe,

Thank you for this precision ! I was not able to locate this brand new doc
section but it might be related to my browser cache.

While we are at it, I think I found two bugs in the gmsh remote code:

+ first, it seems to me that the GMSH_MERGE_FILE remote command is
interpreted as a 'load file' command, which clears the 3D model even if
what is merged are data to be applied on that model (in contrast with the
GUI 'merge file' menu item).

+ second, in the header parsing function:
    int ReceiveHeader(int *type, int *len, int *swap) /* GmshSocket.h, 2
files */
the byte swapping logic to parse the message length is at present:
    if(*len < 0) return 0;
    if(*swap) _SwapBytes((char*)len, sizeof(int), 1);
which works well as long as 'len' is less than 128 :) It should probably
read:
    if(*swap) _SwapBytes((char*)len, sizeof(int), 1);
    if(*len < 0) return 0;
(This change may also be applied to the message type parsing a few lines
before although it is here less critical as the MessageType enum max value
is 104).

With kind regards,
And thank you again for gmsh.
Pierre







2013/8/20 Christophe Geuzaine <cgeuzaine at ulg.ac.be>

>
> Hi Pierre - it's normal (but the documentation is indeed not very
> explicit...). The socket name is actually a "base name", to which the
> unique client ID is appended. The doc has been fixed.
>
> Thanks,
>
> Christophe
>
> On 14 Aug 2013, at 00:54, Pierre Boulenguez <pierre.boulenguez at gmail.com>
> wrote:
>
> > Typo : "localhost:10000" gives gmsh listening on ip 100002
> > On 14/08/2013 00:49, Pierre Boulenguez wrote:
> >> Precision: I use ubuntu 13.04
> >>
> >> On 14/08/2013 00:44, Pierre Boulenguez wrote:
> >>> Hello,
> >>>
> >>> A '2' is suffixed to the socket ip number of gmsh: for instance,
> configured with "localhost:10000" (through the GUI or the  Gmsh Option
> File) gmsh will in fact listen to the port 10002.
> >>>
> >>> I could not find information regarding this behavior in the
> documentation so I suspect it is a bug. I observed the phenomenon with gmsh
> 2.6.2 and 2.8.2.
> >>>
> >>> With kind regards,
> >>> Pierre
> >>
> >
> >
> > _______________________________________________
> > gmsh mailing list
> > gmsh at geuz.org
> > http://www.geuz.org/mailman/listinfo/gmsh
>
> --
> Prof. Christophe Geuzaine
> University of Liege, Electrical Engineering and Computer Science
> http://www.montefiore.ulg.ac.be/~geuzaine
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20130825/861673ff/attachment.html>