// Systems research? Did you actually "research" how a normal user used their // computer? Did you even try to guess how a normal user used their system? // Did you do that and end up with a technical manual whose prime example for // backup strategy involves a "Jukebox?"
You clearly have a very particular, narrow idea of what a "user" is, and a very muddy idea of how research works. Obviously getting an optical jukebox isn't practical for Joe Public sitting in his flat, but it makes great sense for lots of users in larger settings. Perhaps more to the point, experience with fs(4) led pretty directly to the current construction of fs(3), fossil(4), and venti(6) - all of which are much more suitabe for Joe. Put another way: the topic under research wasn't "how do we provide the backup functionality people are asking for?", but "how would having daily dumps change the way you work? would that be useful?". It's a less product- oriented set of questions, but produces more fundamental results. // Plan 9 seems to be a "niche" OS, as I pointed out before. That may well be true, or at least that it isn't mainstream and mass-market. That's never been its objective, and I'm sorry if you wasted your time based on misunderstanding that. // UTF-8 in an English-only "user" paradigm is only extravagance. We've got enough folks around here who use something other than English as their primary language with their computer that this complaing falls down. You're right that there's more research to be done here, such as on right-to-left input methods and composing characters, but that's far from the same thing. If the UI model doesn't work for you, well, that's a shame, I guess. Based on the bash love from earlier posts, I'm going to hazard a guess that your complaints are largely based on the old keyboard vs. mouse argument. I doubt hauling out the old references would be convincing once you've already made up your mind. Anthony
