I wrote: > But what are you saying to those users who don't like it that GNU programs > abort silently when they discover bugs in themselves? Aren't you saying > "tough" in a somewhat more polite way?
On Wed, Mar 16, 2005 at 11:23:55AM -0500, Richard Stallman wrote: > No, because nobody has complained about it. The idea that Emacs > should not use plain abort to crash has only been raised here, not by > Emacs users. The vast majority of Emacs users will never send you mail, and they certainly will not send you mail about problems with debugging Emacs. They run stripped versions of Emacs, included as part of their GNU/Linux or BSD distribution or built for them by their Unix system administrator. When they see abort: core dumped, they just curse Emacs for losing their work and switch to vi. It's very hard to get users to file useful bug reports. If a program crashes, we should make it as easy as possible for the user (even if the user is a non-programmer) to produce a useful bug report, ideally by simply cutting and pasting. > The real complaint that I really got was about cross-jumping. Let me assure you, people have complained about naked aborts. It's true that they don't complain about it on the Emacs developer list, where you participate, because end-user complaints usually go to the GNU/Linux distributions first. If you look at what the complaint you got was, it was that the developer could not isolate which abort() was being reached. There are two ways to solve that: one is to avoid cross-jumping abort() calls; another would be for Emacs to produce more information when it aborts. Either way would solve the problem. Now, as I said, I won't stand in the way if people want to avoid cross-jumping to abort(). But I still think that just producing abort() when an inconsistency is detected, without providing any other information to the user, is rude and inconsiderate to the user. It's not like we have to squeeze our code into 64k byte segments any more.