>>>>> "jiun" == jiun bookworm <thebookworm101 at gmail.com> writes:
jiun> Ray,
jiun> On Sun, Jul 21, 2013 at 10:53 PM, Raymond Toy <toy.raymond at gmail.com> wrote:
Ray> Last thing first: ?What kind of user "astonishment" do you
Ray> see? ?I think we would like to know what that is.
Ray> Now for the first thing. ?If you want to keep things as
Ray> "maxima-ish" as possible, then I think you're doing a
Ray> disservice by having a daemon rename the plot output files.
jiun> maybe your right, the point i was trying to get across (and
jiun> i seem to have failed to do that successfully) is that i did
jiun> not want users to have to learn new custom commands, ?
I don't see that that's a big deal. They have to learn maxima's
commands, so what's one more? :-)
Ray> ?I think Jaime's solution is the correct "maxima-ish"
Ray> solution. ?Or if you're going to hack on a multiuser setup,
Ray> have each user's maxima-init file load a system init file
Ray> where you can implement whatever functions you think all
Ray> users would need. (Kind of like how shell init files
Ray> typically work.) ?At least in this way, the changes are
Ray> discoverable. ?Your daemon approach will be impossible to
Ray> discover how or why it's done. ?Something thatjust magically
Ray> changes the names of files out from underneath users is very
Ray> odd.
jiun> yes this is true for a desktop application, but if its on a
jiun> remote machine and its accessed over a medium like a browser
jiun> (my intention), then i can send the image to the browser as
jiun> soon as its written to file and preserve the norm of viewing
jiun> an image right after running the command, this is my main
jiun> concern-- and the source of my seemingly misinformed choice
jiun> of words :).
If you're using a browser to access maxima, then you can do whatever
you want on the served end, so I don't see how renaming the plot file
is a problem.
Ray