Eric Martz wrote: >Yes, Bob, I think a subset may do the job. > >Again, the goal is to be able to align the two molecules >(approximately) with sync off, and to zoom as needed to get the >relevant portions to be approximately the same size. For example, if >comparing a single domain model to a model with that domain plus >another domain, the goal would be to zoom the models independently so >as to make the common domain about the same size. In this case, one >might also like to center the common domain, and perhaps hide the >"other" domain. > >Then after turning sync on, the follower should rotate and zoom and >translate according to the leader, but relative to the follower's >initial position. The followers center should not be affected. > now THERE's a tall order. Yow.
>More >specifically: > >ROTATION >What we want is for the follower to rotate relative to its initial >position when sync was turned on, rotating the same relative angles >as the leader. When the molecules are not aligned, we don't want the >follower to rotate to the absoute angles (relative to the coordinate >system) adopted by the leader. > >CENTER >Movements of the leader should not change the rotation center of the follower. > >ZOOM >Similarly, we want the follower to zoom relative to its initial zoom. >So, suppose that when sync was turned on, the follower was at 150% >zoom and the leader was at 100% zoom. If the leader is then zoomed to >120%, I think we'd want the follower to zoom to (1.2)*(150%). > >Incidentally, I think that in the current implementation, only >rotation triggers sending sync moveto commands. Specifically, in >Windows, where you can zoom with the mouse wheel without any >rotation, often the followers do not zoom until the leader is >rotated. It would be good if pure zooming would also trigger sending >sync moveto's. > > that's fixed, I think in the latest -- there was a mouse wheel zoom bug there. >TRANSLATION >Translation should be relative, like rotation. The follower should >translate relative to its initial position, following translation of >the leader relative to the leader's initial position. > > > I have to be convinced that is really worth it. Holy cow. >Since I want sync to do rotation, zoom, and translation, but relative >to the starting orientation, I suggest that the new set command might be named > >set syncRelative on > > > It will have to be considerably more than that, and for this sort of very specific set, I have to be convinced that what we would get would be a natural expectation. You're somehow going to have to define starting points, save orientations, take differences in rotations/translations/zoom, etc; apply them independently of the user's application of mouse movements on the target frame.... Well, I guess I'm just saying it isn't going to happen soon. >I would deem the current sync behavior to be "absolute". > >Thanks very much, -Eric > > > ------------------------------------------------------------------------- SF.Net email is sponsored by: Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php _______________________________________________ Jmol-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/jmol-users
