[Date Index] [Thread Index] [Author Index]
Re: v6: still no multiple undo?
David Reiss wrote: > The minimal approach to Undos is part of the reason why I created the > notebook backup and tracking functionality that I include in A > WorkLife FrameWork (shameless plug!). It is sort of a mini CVS in > Mathematica. > > I do think though that Undo is tricky in Mathematica. One reason is > that some of what one does and might want to undo can be **very** > large. Another is the question of what one means by Undo and redo. > Does this include evaluations? Does it include formatting changes? > And so on. I suspect that the question is rather more complex than > what one might have to deal with in a simple text editor or word > processor. But, on the other hand, I am sure that some cleverness > could be brought to bear on this... > > --David > > http://scientificarts.com/worklife > Now Mathematica 6 compatible > > > On Jun 7, 3:54 am, Selwyn Hollis <sh2.7... at earthlink.net> wrote: >> Ah, that's at the top of my gripes list too. I'm pretty sure that >> Undo is even less functional in 6 than in 5.2. >> >> Another shortcoming is the lack of AutoSave or AutoBackup capability. >> It's too easy to lose track of time, crash, and then discover that >> you haven't saved anything in 2 hours. (Or maybe its just dementia on >> my part.) I've created a little palette of my own with an AutoSave >> function that saves the InputNotebook every few minutes. That works >> pretty well, and if anyone would like to test it, let me know. I've >> made an attempt at an AutoBackup function, but I don't know how to >> make it sufficiently non-intrusive. >> >> -- Selwyn >> >> On Jun 6, 2007, at 6:43 AM, Will Robertson wrote: >> >>> Hello, >>> I am baffled that technical software is released in 2007 with only a >>> single undo level. Am I missing something here? Is it because I'm >>> using a student license? How hard can an undo stack really be in the >>> notebook interface? >>> I've been irritated a few times in v5.2 from clumsy keypresses losing >>> me work that I would expect to be two "undo"s away. Alas not. Still. >>> Will Robertson > > > A multi-level undo that let you go back to the last evaluation would be a step forward. Alternatively, a multi-level undo that ignored generated cells might be a possibility - after all, it is usually (but not always) the input that is really valuable, the output can be regenerated. David Bailey http://www.dbaileyconsultancy.co.uk