(2nd version of this message, the previous one did not make its way to the list
yet)
I am currently working on a viewer that embeds a sequence display with
mouseover capability : when the user hovers a residue in the sequence, a
command is sent to Jmol so that the corresponding residue is highlighted. This
is (very) cool and works well in Safari, at least on my computer.
When I tried it with firefox I ran in exactly the same problem as described by
Philip Bays recently. I tried the workaround given by Angel taken from the wiki
there :
3.2 Activating UseCommandThread
It worked... that is to say the files could be loaded in the applet. But the
sequence widget became sluggish : the highlighting in Jmol took a while to
appear, after the residue had been hovered in the sequence...
I wrote a small test to see if the setcallback was in cause in the drop of
performance :
a javascript for loop calls 20 times in a row the jmolscript method. The time
for the end of each script is reported to the console with the print @{now()}
command (apparently not documented, but found on the new.htm page).
When useCommandThread is not set, there is a lapse of 4ms between each script
with safari (100ms with firefox)
When it is set, the lapse is of 200 - 250 ms (firefox being a bit slower than
safari on my laptop) which explains why my page became suddenly slow.
I put the page online if somebody wants to check :
http://librairiedemolecules.education.fr/outils/test/test.htm
Is there, by any chance, a way to disable that setting once the model is
loaded, and to re-enable it later, when a new model is going to be loaded ?
-Paul------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
