Hi Nick, On Tue, Feb 7, 2012 at 03:31, Nick Dokos <nicholas.do...@hp.com> wrote: > suvayu ali <fatkasuvayu+li...@gmail.com> wrote: > >> On Sat, Feb 4, 2012 at 18:46, Nick Dokos <nicholas.do...@hp.com> wrote: >> > >> > Confirmed. If I add >> > >> > (require 'org-clock) >> > >> > to my minimal .emacs, it goes away. So it seems to be a missing dependenc= >> y. >> > >> >> It seems an odd dependency to have. > > True - org-in-clocktable-p does not have anything to do with clocks, but > it ended up in org-clock.el. All of these "where am I" functions should > probably be in org.el. In fact, they all are except for > org-in-clocktable-p and org-at-item-p. The latter is in org-list.el and > it is not giving an error a) because org-in-clocktable-p errors out > first and b) because org.el contains a (require 'org-list). So having a > (require 'org-clock) in org.el does have a precedent, but of course this > way you end up defeating autoloads: the moment org.el is loaded, all of > these things are too. There might be a better arrangement but it's like > Pascal's letter[fn:1] : it would require time to find it :-) >
True, actually this is one of my only gripes (if you could call it that) about org. Loading my org customisation is a fair part of my emacs start time. If I could improve this, it would be perfect. :) >> Anyway, I noticed something; >> pressing F10 does bring up the menu as expected without any backtrace. >> After that I can select anything on the menu with a mouse without >> problems. > > That's true: org-context is only called from mouse code, so if you drop > the menu without the mouse, there is no problem: selecting a menu item > with the mouse, once the menu has dropped, is all widget code and does > not invoke org code at all. > Thanks for clearing that up. I was always hazy about the emacs interface libraries. I think I'll look into them someday when I have some time. >> The error occurs only when bringing up the menu for the >> first time with a mouse. >> > > That I find surprising though: I would expect that if you try to > click on the menu again, you'd get the error again: nothing is done > to define the function, so why should anything be different the > second time? > Ah, sorry my mistake. I did not word it correctly. I didn't mean bringing up the menu with a mouse a second time. I meant once the menu was up (lets say the "File" menu), going to a different one by hovering over the menu (say "Edit" or "Options") with a mouse doesn't trigger the backtrace. Clearly it is obvious now why it would be so from your explanation above. :) > [fn:1] Approximately: "Forgive the length of this letter: I did not have > time to make it shorter." This is a very interesting and true remark. I was not familiar with it before. :) -- Suvayu Open source is the future. It sets us free.