On 10/21/2014 06:55 PM, go wrote: > On 2014-10-21 18:39, Rolf Huehne wrote: >> It is not that simple. >> The current user might not have the right to write in the install >> directory. >> What about multiple users on the same computer? > > Hi Rolf, > > You are right. Then the choice of the directory requires more > consideration, but I am sure it is not an unsolvable problem. > >> One set of problems here will be caused by multiple instances of Jmol >> run simultaneously by the same user on the same computer. >> If you write one line at a time you will get a mixture of different >> sessions that will not be very helpful. You must also make sure that >> the >> instances are writing one after another. > > Good point again. > Writing a single history file by multiple instances should never ever > happen. > >> And what about command-line usage? >> We use Jmol.jar for example for producing automatically about 300000 >> images for all PDB structure by a complex script with about hundred >> lines of code. I wouldn't want that to be slowed down by writing into a >> history file for each line (and coordinating to do this from 30 >> processes in parallel). > > Your massive use case and a usual simple experimenting differ a lot. > I still think that a history file across multiple sessions > is of great value while one is experimenting with Jmol. > In "expert mode" is should be just switched off completely. > I don't doubt the usefulness of a history file across multiple sessions. My point here is that it looks much more simple to add such an automatic feature than it actually will be to develop a consistent one. (Even the developers of the Unix/Linux shells I have worked with havn't solved many of the problems regarding to write a consistent command history across sessions from multiple instances. But of course I don't know how hard they tried.) My suggestion would be to leave out the autmatic part and let the user control it manually. Then the only thing that would be needed would be a 'load history' command. Everything else would already be there.
As far as I know you can already define a script file that is executed at startup. There you could load your history file (or choose between multiple ones). I don't know if it is available yet, but helpful would be to have the same possibility at the end. (A finish script, where you could also do other useful stuff like writing the current state). I know that by Murphy's law the commands that are most important and most difficult to reproduce will get lost. But you might do an extra manual saving for really difficult or important commands to minimize this. Regards, Rolf -- Rolf Huehne Postdoc Leibniz Institute for Age Research - Fritz Lipmann Institute (FLI) Beutenbergstrasse 11 07745 Jena, Germany Phone: +49 3641 65 6205 Fax: +49 3641 65 6210 E-Mail: [email protected] Website: http://www.fli-leibniz.de Scientific Director: Prof. Dr. K. Lenhard Rudolph Head of Administration: Dr. Daniele Barthel Chairman of Board of Trustees: Dennys Klein VAT No: DE 153 925 464 Register of Associations: No. 230296, Amtsgericht Jena Tax Number: 162/141/08228 ------------------------------------------------------------------------------ Comprehensive Server Monitoring with Site24x7. Monitor 10 servers for $9/Month. Get alerted through email, SMS, voice calls or mobile push notifications. Take corrective actions from your mobile device. http://p.sf.net/sfu/Zoho _______________________________________________ Jmol-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/jmol-users
