Re: [Jmol-users] Jmol prerelease 10.00.52

[Miguel <mi...@jm...>]

Frieda wrote:
> ZOOM
> The zoom with perspective depth turned off is
> nothing short of amazing. Great work=21

Good.

> With perspective depth ON, zoom still stops
> short of what might be desired in certain circumstances, but I think
> it is improved, and the complete lack of distortion at high zoom
> levels is lovely. Go, Jmol team=21

Perspective depth has always been a problem. Bob has spent a lot of time
looking at this code and has some good ideas on how to address some of th=
e
perspective depth issues.

> ******
> This is a bit of an aside, NOT specific to 10.00.52: With perspective
> depth on, centering an atom or a subset of atoms that are toward the
> =22front=22 of a structure (i.e., closer to the user) automatically mak=
es
> them appear smaller (they move farther away from the user). Zooming
> in all the way on that centered set may or may not go as close as one
> wants - but if you do zoom in as far as possible, and then rotate the
> structure 180 degrees in the y axis, wow, are the atoms on the other
> side ever close=21 I assume this is because the centered set is the
> center of rotation and so there has to be space for the opposite end
> of the structure to swing around into view.

Correct.

Or, you may decide to display some of the atoms on the near side at any
point ... so they are still there.

In some sense, zoom does the calculations based upon the atoms that are
closest to the user. All other atoms are smaller. This happens regardless=

of whether or not the atoms are visible.

> So in a way, this is what is limiting zoom on
> the centered set... centering it=21 It is ironic,
> since that set is what you want to examine most closely. It is
> tantalizing to look at the atoms on the 'other side' but without
> being able to rotate around them it does not provide a good view of
> interactions. The solution is to turn perspective depth off for close
> examinations of small regions in large structures

Yes.

I should have realized this and told people about it years ago.

Even before the distortion stuff was fixed, it would have given much
higher effective levels of zoom.

> HBONDS and RESTRICT - BUG?
> Issuing a restrict command causes hbonds to appear. For example
> 'restrict protein'
> turns on h-bonds, as does
> 'restrict not water'

That must be broken :-(

> The restrict command does not turn hbonds on if they have been
> explicitly turned off prior to the restrict command. So,
> 'hbonds off; restrict protein'
> does not turn on hbonds.
>
> =5BIn 10.00.52, when a protein loads, hbonds are not shown by default
> in the 2 pdb files I am using at present. I think this is as expected
> for this release,

Yes, that is the expected behavior. Although different poeple are
reporting different problems.

> although I am a little lost on whether the current
> discussion of having hbonds come on by default was meant for this
> release or the next.=5D

The change in hbond behavior is intended to be in this release because of=

the addition of the 'connect' command.

> OTHER
> Is there a sense of what is going to happen to the new spin-related
> commands that were removed? Are they still slated for implementation
> in Jmol (I hope)?

They are still slated for future integration into Jmol, after the officia=
l
10.1 release.

> Also, with the 10.1 release...
> -could 'set windowCentered on/off' be implemented?
> This command would be very useful to have in place

Probably not.

> -will the planned smoother translations be implemented?

No


Miguel