From: John Stone (johns_at_ks.uiuc.edu)
Date: Thu Jan 26 2006 - 10:41:41 CST

Axel,
  I agree, this has become a common issue of late.
Rather than documenting the limitation in the numerous places that would
have been required, I spent a couple of days writing and testing
Windows-specific large-file-capable I/O code which is now used by the new
DCD plugin, and so hopefully this limitation has now been eliminated for
Windows users. Most of the other file formats will still have the limitation
since they still use stdio calls to read their respective file formats,
but the DCD plugin should now work with huge files. This new code
will be available to Windows users when I post the next test version
of VMD 1.8.4.

If Windows users have encountered any other file size limitations with other
plugins, let me know and I may be able to adapt those plugins to use the
new I/O code as well.

  John Stone
  vmd_at_ks.uiuc.edu

On Thu, Jan 26, 2006 at 08:09:22AM -0500, Axel Kohlmeyer wrote:
> On Wed, 25 Jan 2006, Vineet Pande wrote:
>
> VP> Thanks John-
> VP> I am running on Windows XP, trajectory is a merged and
>
> while VMD itself is only limited by the available memory,
> the i/o interface of windows imposes a 2GB filesize
> limit. if you split your trajectories into smaller
> chunks, it should work.
>
> btw.: this _has_ been discussed here _numerous_ times.
> john, perhaps it would make sense to document it in
> the manual in the sections mentioning reading files
> (via GUI or script).
>
> regards,
> axel.
>
> VP> recentered CHARMM output with 2000 frames which i
> VP> think is making the problem; because when I load the
> VP> same trajectory but with less frames (i.e. around 20
> VP> frames by jumping more picoseconds of trajectory per
> VP> frame), it works fine. So its either the "file size
> VP> problem" or as you mention the "platform problem". Do
> VP> we expect to load very many frames with solvent for a
> VP> large system like mine in any other platform like
> VP> Linux etc.? I am dealing with a system of total 107295
> VP> ATOMS and the trajectory is a .cor (binary) CHARMM
> VP> file, which is read correctly as a "CHARMM; NAMD;
> VP> XPLOR, DCD" type in VMD.
> VP>
> VP> Regards..
> VP> Vineet Pande
> VP> Karolinska Institute, Stockholm
> VP>
> VP>
> VP> --- John Stone <johns_at_ks.uiuc.edu> wrote:
> VP>
> VP> >
> VP> > Hi,
> VP> > What file format is your CHARMM trajectory written
> VP> > in?
> VP> > Is it being loaded by VMD as a DCD file, or some
> VP> > other format (COOR?)?
> VP> > If it is a DCD file, the next question is what
> VP> > platform you're
> VP> > running VMD on? The specific error message you
> VP> > mention below
> VP> > doesn't sound like anything in the DCD reader, thus
> VP> > my initial
> VP> > question about which file format your trajectory was
> VP> > written in.
> VP> > If the trajectory is written as a COOR file (read in
> VP> > VMD by corplugin),
> VP> > then the problem could be one of several things. I
> VP> > just fixed a couple
> VP> > of bugs in corplugin, specifically affecting the
> VP> > extended COOR format
> VP> > for structures with more than 99,999 atoms. It
> VP> > could be that you've
> VP> > just run into one of the problems I recently fixed.
> VP> > The next test version
> VP> > I post on the web site will include the new
> VP> > corplugin, and may fix your
> VP> > problem if it's an issue with file formatting. If
> VP> > your problem is due
> VP> > to the file size, then the solution will depend on
> VP> > which platform you're
> VP> > running VMD on.
> VP> >
> VP> > John Stone
> VP> > vmd_at_ks.uiuc.edu
> VP> >
> VP> > On Wed, Jan 25, 2006 at 04:21:53AM -0800, Vineet
> VP> > Pande wrote:
> VP> > > Hi,
> VP> > > I am trying to load a CHARMM trajectory (after
> VP> > merging
> VP> > > smaller chunks and recentering solvent), but VMD
> VP> > > (version 1.8.4b6) gives an error message saying-
> VP> > Error
> VP> > > opening file: improperly formatted line. The
> VP> > number of
> VP> > > atoms in my system are 107295 and the size of the
> VP> > PBC
> VP> > > trajectory is around 2GB. When I load similar file
> VP> > > without solvent and PBCs, it works fine. Does it
> VP> > have
> VP> > > anthing to do with size limits?
> VP> > >
> VP> > > Thanks in advance,
> VP> > > Vineet Pande
> VP> > > Karolinska Institute, Stockholm
> VP> > >
> VP> > >
> VP> > > __________________________________________________
> VP> > > Do You Yahoo!?
> VP> > > Tired of spam? Yahoo! Mail has the best spam
> VP> > protection around
> VP> > > http://mail.yahoo.com
> VP> >
> VP> > --
> VP> > NIH Resource for Macromolecular Modeling and
> VP> > Bioinformatics
> VP> > Beckman Institute for Advanced Science and
> VP> > Technology
> VP> > University of Illinois, 405 N. Mathews Ave, Urbana,
> VP> > IL 61801
> VP> > Email: johns_at_ks.uiuc.edu Phone:
> VP> > 217-244-3349
> VP> > WWW: http://www.ks.uiuc.edu/~johns/ Fax:
> VP> > 217-244-6078
> VP> >
> VP>
> VP>
> VP> __________________________________________________
> VP> Do You Yahoo!?
> VP> Tired of spam? Yahoo! Mail has the best spam protection around
> VP> http://mail.yahoo.com
> VP>
>
> --
> =======================================================================
> Axel Kohlmeyer akohlmey_at_cmm.chem.upenn.edu http://www.cmm.upenn.edu
> Center for Molecular Modeling -- University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.

-- 
NIH Resource for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
Email: johns_at_ks.uiuc.edu                 Phone: 217-244-3349
  WWW: http://www.ks.uiuc.edu/~johns/      Fax: 217-244-6078