RE: flipdcd corrupting dcd files

From: peter.schmidtke_at_fr.netgrs.com
Date: Fri Jun 15 2012 - 00:58:16 CDT

Dear Joshua,

Thanks for your reply, indeed I missed that part where flipping dcd files is not reversible.
I am aware of the fact that VMD can read either endiness, but the reason why I need this now is as you said, some home grown code. I'll try to integrate the molfile plugin to avoid such problems, but in order to make a proof of concept I want to avoid spending too much time in programming ;)

Thanks again for your reply!

Peter

-----Message d'origine-----
De : Joshua D. Moore [mailto:joshuadmoore_at_gmail.com]
Envoyé : jeudi 14 juin 2012 20:40
À : SCHMIDTKE Peter IDRS
Cc : namd-l_at_ks.uiuc.edu
Objet : Re: namd-l: flipdcd corrupting dcd files

According to the website

http://www.ks.uiuc.edu/Development/MDTools/flipdcd/

"This conversion is not reversible so you may wish to make a backup copy of your file(s)."

VMD can read either endiness now on any endiness machine, so unless you are trying to read the file with your home grown code, I'm not sure why you even need to use flipdcd anymore?

On Thu, Jun 14, 2012 at 7:30 AM, <peter.schmidtke_at_fr.netgrs.com> wrote:
> Dear NAMD mailing list readers,
>
>
>
> I'm experiencing some problems with flipping endianess of my
> trajectories
> (32 bit DCD format written by NAMD 2.9 CUDA in little endian) using
> the flipdcd program coming with the NAMD binaries.
>
> When I do the flipdcd going to big endian and back again to little and
> so on, after some time my trajectory file gets corrupted (VMD can only
> read some frames in the beginning, but not the end anymore). It's
> mostly the end of the trajectory that becomes "unreadable".
>
>
>
> Are these issues known using flipdcd? My dcd files are 2.1Gb big, by
> the way.
>
>
>
> Thanks in advance for your reply.
>
>
>
> Best regards.
>
>
>
> Peter Schmidtke

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:40 CST