[ase-users] Python 3 and PickleTrajectory
Jens Jørgen Mortensen
jensj at fysik.dtu.dk
Wed Mar 18 08:20:04 CET 2015
On 03/03/2015 02:44 PM, Peterson, Andrew wrote:
> This is a great idea. Thanks!
>
> Andy
>
> On Wed, Feb 25, 2015 at 1:32 AM, Jens Jørgen Mortensen
> <jensj at fysik.dtu.dk <mailto:jensj at fysik.dtu.dk>> wrote:
>
> Hi!
>
> As some of you may have noticed, ASE-trunk now works with both
> Python 2 and 3! More specifically Python 2.6, 2.7 and 3.4.
>
> It's not that we all need to start using Python 3 right now, but
> some day (in two years maybe) we'll have to. When I say that it
> works, I mean our test-suite passes and our Buildbots will check
> this for every new change made to the code. Some things that are
> currently not tested may not work:
>
> * ase-gui won't work (interface to GTK completely changed for
> Python 3)
> * Many calculator interfaces need to be fixed
> * Reading of PickleTrajectory files will not work if they contain
> certain types of constraints
>
> The first two can be fixed some day, but I feel we need to worry
> about the PickleTrajectory files now.
>
> * Writing/reading constrints is very fragile - the slightest
> rearrangement of constraints.py will break things
> * Reading constraints from Python 3 will most likely not work
> * Reading a PickleTrajectory from an untrusted or unauthenticated
> source is a security problem
>
> Yesterday, I added a new trajectory writer/reader that doesn't
> have any of those problems - it doesn't use pickle :-) I propose
> that we make the new trajectory the default in ASE now so that we
> can move away from pickle files for long term storage.
>
As of today, the development version of ASE (also called trunk) will use
the newly added pickle-free trajectory as default trajectory. You can
still work with the old ones, but you will get a warning if you do. I
need as many brave ASE uses as possible to try out the development
version of ASE (to become version 3.9) and try to convert some of your
old trajectory files to new trajectory files. It's done like this:
$ python -m ase.io.trajectory a1.traj ...
Let me know if there are any problems.
Jens Jørgen
>
> What do you think?
>
> Jens Jørgen
> _______________________________________________
> ase-users mailing list
> ase-users at listserv.fysik.dtu.dk
> <mailto:ase-users at listserv.fysik.dtu.dk>
> https://listserv.fysik.dtu.dk/mailman/listinfo/ase-users
>
>
>
>
> --
> Andrew Peterson
> Assistant Professor
> Brown University School of Engineering
> Barus & Holley 247
> 184 Hope Street
> Providence, RI 02912
> (401) 863-2153
> http://brown.edu/go/catalyst
More information about the ase-users
mailing list