On 23 October 2012 16:51, Robert L. Savoy <[log in to unmask]> wrote:
> When teaching people how to use the file-selection window in SPM, those people using Macs have extra, unnecessary headaches. I will ask about two, in the hope that there is both specific information about how to deal with them, and perhaps a general site/message that I missed about this kind of question in the past.
Not many of the SPM developers have access to Macs, so this sort of
thing tends to be a bit neglected.
>
> 1) The font size in many SPM8 windows, especially in buttons, but also in figures generated by SPM8, is often inappropriately large or small (respectively). Is it possible/easy to change this?
I'll leave this to someone else. I don't personally like the excess
of figure windows that SPM uses, because I have to open and close a
number of different windows in order to find the appropriate one on
which to click a button. I had hoped that we'd be doing everything
from a pull-down (similar to the old TASKS pull-down in the graphics
window) by now.
>
> 2) The following is much more specific, and clearly a "bug", as it doesn't serve any good purpose and is different on Macs from PCs. When in a File-Selection window, if one "right clicks" to do a "Select All" (files) or "Deselect All" (files), either by having a right-button, or using the more common "Control-leftbuttondown", the "Select All" or "Deselect All" pop-up button appears. On a PC, one has to release the (physical) button to get the (on-screen-software) button. And then one moves the cursor to the active portion of the button to SelectAll or DeselectAll. On a Mac, as soon as one pushes the (physical) button, the "Select All" (or Deselect All) on-screen-software button appears. Now comes the critical part: If one KEEPS THE PHYSICAL BUTTON PUSHED (i.e., "buttondown" state), one can move the cursor to the active part of the screen button to do the appropriate action. HOWEVER, if one releases the physical button ("buttonup" state), the on-screen "SelectALL" or "DeselectAll" button stays on the screen, and one can subsequently move the cursor into the active part and click again, BUT ONE FILE HAS ALREADY BEEN SELECTED...namely, the one where the cursor happened to be in the select or deselect portion of the file-selection window. If one then absent-mindedly doesn't notice this, and then clicks the button, the remainder of the files are selected (so it LOOKS like you've correctly selected all the files you were trying to select), but, in fact, they are in the wrong order.
I'll leave this one to someone else, as it's part of matlabbatch, so
we in the FIL don't have control over that part of theuimenu code. I
suspect that it may not be so straightforward to change the behaviour
anyway, as it seems to relate to differences between how MATLAB does
stuff with uicontextmenu and uimenu on the different platforms.
>
> I suspect that it may simply be a matter of changing the way the Select All and Deselect All portion of the code for the Mac is implemented, being careful to do some actions only after the buttonup state is reached. But it has been several decades since I programmed Macs at that level, so I don't really know how difficult this is to fix.
It's all MATLAB stuff anyway, hidden away in /matlabbatch/cfg_getfile.m .
> -ps: While I am here: Have the "bounding box" default values been changed to something like -70 for z, to include the bottom of the cerebellum? For years I have had to change the spm_defaults file to get that part of the brain included, i.e., within my spm_defaults.m I have the lines:
> %defaults.normalise.write.bb = [[-78 -112 -50];[78 76 85]];
> defaults.normalise.write.bb = [[-78 -112 -70];[78 76 85]];
OK. Now we get to why I answered this one. I plan to extend the
default bounding box to z=-70 for SPM12. I should have done this ages
ago, but kept forgetting. It's not the sort of thing that can just be
slipped in among a set of minor updates, and really has to be part of
a major release.
All the best,
-John
|