On 10/23/2012 9:00 AM, Scott Hirsch wrote: > Thanks, dpb, for jumping in. My apologies for coming across as if > justification for the previous workflow were needed - I was just > interested in understanding the workflow better. > > I wanted to take a minute to clarify the behavior change, just because I > realize there may be some confusion among other readers of this thread. > R2012b preserves the ability to manage editor files either as a group > within the editor container, or as individual system-level windows. This > allows the choice between using the operating system to manage the > editor document windows or the MATLAB desktop. Windows within the editor > container can either fill the entire container or be tiled as before. > The specific change is that windows within the desktop editor container > can no longer "float" such that individual documents are overlapping > each other, while still contained within the editor group. One > alternative to this behavior is to undock individual editor files, > managing their relative positioning as operating system-level windows. ...
OK, that makes _some_ since, but begs the question of "why remove a functionality" other than for the obvious of it releases TMW of the burden of maintaining that piece of code.
I know it isn't really intended as a justification request from TMW's viewpoint but I respond as a gut-level reaction of long-term user. Don't know if you monitor cs-sm very much or not, but there were a couple of other threads here just recently on the new desktop and a change in behavior of a couple of functions. Specifically, one of the function changes removed a fairly obvious generalization in usage that the suggested workaround is, simply put, just butt-ugly in comparison. A TMW followup requested for the same input in that regards and I was the first to point out that to a user it seems like having to beg to keep what we already had/have. It's a point-of-reference thing.
Since I don't have the feature to play with I can't really comment on the real difference in how the suggested replacement functionality works just the general drift of modifications seem to be being made on superficial things while more basic areas don't get much attention.