Yes, indets seems to be what I was looking for... Thanks!

Yes, indets seems to be what I was looking for... Thanks!

The helppage for $ says it all: Note: It is recommended (and often necessary) that both 'expr' and 'i' be enclosed in single quotes to prevent premature evaluation. (If i had been assigned a value, the i in the expression would be evaluated to its value, and so could not be used as a counting variable.) The most common use is 'expr' $ 'i' = m..n; So using 'modp'(jj,3) $ jj=0..8 instead show the correct values 0,1,2,0,1,2,0,1,2 Regards, Franky.

Yes indeed, it seems that using forget(evalf) eliminates this problem... thanks! Now, is there a way to avoid using remember tables alltogether? Because I am not sure which routines use them and which don't, so I *am* sure to forget forgetting some of them... ;-) Thanks for any reply, Franky.

The file Maple10LinuxUpgrade.bin is only for 32-bit, so you need to download the Unix Network Update instead (for Maple 10.03, its ftp://ftp.maplesoft.com/pub/maple/downloads/maple10-03/Maple1003UnixUpgrade.tar.gz). Regards, Franky.

The file Maple10LinuxUpgrade.bin is only for 32-bit, so you need to download the Unix Network Update instead (for Maple 10.03, its ftp://ftp.maplesoft.com/pub/maple/downloads/maple10-03/Maple1003UnixUpgrade.tar.gz). Regards, Franky.

Which version are you running? Assuming the command line version works, what does it show? Here, it shows Maple 10 (X86 64 LINUX)... just to be sure we're talking about the same version. FYI, here we also have SuSE 9.2 on a dual Opteron and everything works fine, so you should be able to get it to work. What's on your PATH? Here are my settings: ~/bin:/home/backelj/bin:/usr/local/bin:/usr/bin:/usr/X11R6/bin:/bin:/usr/games:/opt/gnome/bin:/opt/kde3/bin Maybe there's a path to a conflicting java installation or so... not sure though. Otherwise, try installing it again and let me know. Regards, Franky.

Which version are you running? Assuming the command line version works, what does it show? Here, it shows Maple 10 (X86 64 LINUX)... just to be sure we're talking about the same version. FYI, here we also have SuSE 9.2 on a dual Opteron and everything works fine, so you should be able to get it to work. What's on your PATH? Here are my settings: ~/bin:/home/backelj/bin:/usr/local/bin:/usr/bin:/usr/X11R6/bin:/bin:/usr/games:/opt/gnome/bin:/opt/kde3/bin Maybe there's a path to a conflicting java installation or so... not sure though. Otherwise, try installing it again and let me know. Regards, Franky.

We are experiencing the same problem on Linux. So I guess this does not have to do with Mac OS X but rather with the GUI itself... hope they can fix this, as this makes it unworkable.

That's the case - I have Mac OS X 10.4.3... hope they fix this soon.

I would like to have the name 'a' appear in the result, just like in the first example where the result is somefunc(a). But since 'a' is effectively replaced with its corresponding table (to which the assumption has been appplied), there's no way to be able to refer to the original argument that was begin passed (namely 'a'). So I would like to know whether it is possible to retrieve the original argument that is passed to the procedure...

I would like to have the name 'a' appear in the result, just like in the first example where the result is somefunc(a). But since 'a' is effectively replaced with its corresponding table (to which the assumption has been appplied), there's no way to be able to refer to the original argument that was begin passed (namely 'a'). So I would like to know whether it is possible to retrieve the original argument that is passed to the procedure...

Already one step in the right direction - thanks! But unfortunately it seems to work only for the first message in a thread. What I wanted was to be able to print the full thread... hopefully, this will also be possible?

Unfortunately, this does not help either - read my post above... but thanks anyway! > evalf(Re(evalc(Psi(2,2+2*I)))); 0.03902435405 + 0.1565674760 I

I found the problem - a rather silly bug. In the helppage for Re, we read : "If x includes a function f, then Re(x) and Im(x) attempt to execute the procedures `Re/f` and `Im/f` to determine the real and imaginary parts of the corresponding part of x." So, look at the definition of `Re/Psi`: > eval(`Re/Psi`); proc(x) local a; if nargs = 1 then a := Re(x) else a := Re(args[2]) end if; if a = x then 'Psi'(args) else FAIL end if end proc So here we go... in the case of Re(Psi(2,2+2*I)), nargs is 2 so a := Re(args[2]) which equals 2. Now x is the first argument, which is 2, so 2 = 2 and we return 'Psi'(args), which is in fact the same as the original, namely Psi(2,2+2*I). If we now evaluate this using evalf, we get a complex floating point number. Rather silly, don't you think...