Re: [Jmol-users] anim command with multiple files

[Brandon Hespenheide]

Miguel, > I would like to display a movie of the conformational motion, but I do > not want to have to create an additional PDB file that is simply a > cat'ed version of all the other output files, as this will essentially > double the amount of space each run requires.   Please explain your concern about doubling the amount of space. For the size data that you are working with, disk space is effectively fr= ee. Client performance and network bandwidth are more important.   Our software generates unique, valid, conformations of proteins. The number of conformations produced depends on how long of a simulation the user runs, and how often they output files. Typically, I output ~30 conformations for a directed run (finds a pathway between two conformations) and 50-100 conformations for undirected (freely exploring space consistent with the constraints we have put into the program). We have a “file cabinet” functionality so that users can keep their data online for further analysis, rerunning, sharing with others, etc. If we plan for ~300 users, at 50 runs, each with 20 conformations of a 50KB file, that’s 15TB. Space is an issue for us, and user quotas are going to be implemented (I’ve only got 2.1TB to work with, for now..)   As you stated, bandwidth is also an issue. I’m trying to decide between generating the cat'ed files on the fly when the user loads the page, or using the load/delay method you describe below. I’ll probably try both, and see which is faster.   Thanks Brandon   >             Does anyone know if it is possible with scripting to load > each of my conformers sequentially, and then display them as an > animation?   Someone else asked the same question.   I recently changed the code so that loading a new molecule does not clear=   the old molecule. Therefore, a script that simply alternates load and delay script commands should work   load =22x000=22; delay 0.5; load =22x001=22; delay 0.5; load =22x002=22; delay 0.5;   You must interleave calls to either 'delay' or 'refresh', otherwise the script will keep running and the display will not get updated.   > A results similar to the anim command would be desirable.   > Otherwise, I think we'll need to hack something with php and rpc in > order to dynamically create and destroy a concatenated file for > viewing.   I suggest:  - create the catted file statically  - Don't worry about the space  - Compress everything with gzip.