@Carl Love 

I wonder what exactly are keywords in 2D input. Certainly, ?keyword does not explain correctly this issue in the 2D language.

@Carl Love 

OK, for this equation, an asymptotic expansion could be given for large abs(x), as the roots approach the poles of tan(x). Something along these lines:

st:=series(tan(x),x=n*Pi/2,2) assuming n::odd;
                   (-1)                   /            3\
       /    1     \       1     1         |/    1     \ |
      -|x - - n Pi|     + - x - - n Pi + O||x - - n Pi| |
       \    2     /       3     2         \\    2     / /
s:=solve(convert(st,polynom)=x,{x});
            /                              (1/2)\   
            |    1        1 /    2  2     \     |   
           < x = - n Pi + - \9 Pi  n  - 96/      >, 
            |    8        8                     |   
            \                                   /   

              /                              (1/2)\ 
              |    1        1 /    2  2     \     | 
             < x = - n Pi - - \9 Pi  n  - 96/      >
              |    8        8                     | 
              \                                   / 
series(subs(s[1],x),n=infinity,2);
                     1         2      /1 \
                     - n Pi - ---- + O|--|
                     2        Pi n    | 3|
                                      \n /

@Carl Love 

OK, for this equation, an asymptotic expansion could be given for large abs(x), as the roots approach the poles of tan(x). Something along these lines:

st:=series(tan(x),x=n*Pi/2,2) assuming n::odd;
                   (-1)                   /            3\
       /    1     \       1     1         |/    1     \ |
      -|x - - n Pi|     + - x - - n Pi + O||x - - n Pi| |
       \    2     /       3     2         \\    2     / /
s:=solve(convert(st,polynom)=x,{x});
            /                              (1/2)\   
            |    1        1 /    2  2     \     |   
           < x = - n Pi + - \9 Pi  n  - 96/      >, 
            |    8        8                     |   
            \                                   /   

              /                              (1/2)\ 
              |    1        1 /    2  2     \     | 
             < x = - n Pi - - \9 Pi  n  - 96/      >
              |    8        8                     | 
              \                                   / 
series(subs(s[1],x),n=infinity,2);
                     1         2      /1 \
                     - n Pi - ---- + O|--|
                     2        Pi n    | 3|
                                      \n /

The equation tan(x)=x is trascendental. What useful formula do you expect?

The equation tan(x)=x is trascendental. What useful formula do you expect?

@mehdi jafari 

Actually, the Standard GUI has an auto save feature, set up  by default at every 3min, I think, see ?worksheet,managing,OptionsDialogGeneral#AutoSave . The problem, as I see it,  is that this backup and restore process may be not reliable enough, specially for large/complex worksheets.

@Carl Love 

First test in Maple 17.00 64-bits on Sabayon Linux 10 does not end after some 500s. It seems that there is something quite strange here...

@Carl Love 

First test in Maple 17.00 64-bits on Sabayon Linux 10 does not end after some 500s. It seems that there is something quite strange here...

@Axel Vogt 

If English were so much "The Language", which would be the need for Maplesoft of going into the complication of making "localized" versions of Maple? (Japanese, Spanish, and so on).

@roman_pearce 

Do you mean running Typesetting:-Parse:-ConvertTo1D and related machinery for every keystroke? How large would be the overhead?

@fereydoon_shekofte 

I can reproduce this copy & paste bug. The expression pasted at the 1D prompt that I get is (a__n+`a__n-1`)/(`a__n-2`+`a__n__-3`). Apparently, after this comment, a new implementation based on the undocumented command Typesetting:-mcomplete, is handling this operation. So, it might be a bug there.

@PatrickT 

Have you tried with the suboption fdiscont to discont? I understand that in such case it uses a (possibly heuristic) numerical algorithm instead of a symbolic one, which may be faster. See ?plot,discont .

Actually, I have never had time to test this numerical algorithm over a significant sample of 2D plots with discontinuities. In particular how much longer it would take wrt the plot without discont=[fdiscont]. The point is that if this overhead were small enough, it might be sensible to keep it as default.

@PatrickT 

Have you tried with the suboption fdiscont to discont? I understand that in such case it uses a (possibly heuristic) numerical algorithm instead of a symbolic one, which may be faster. See ?plot,discont .

Actually, I have never had time to test this numerical algorithm over a significant sample of 2D plots with discontinuities. In particular how much longer it would take wrt the plot without discont=[fdiscont]. The point is that if this overhead were small enough, it might be sensible to keep it as default.

This issue has been discussed several times here. A single-language (English here) forum helps to keep a homogeneous comunity but tends to exclude all those persons not knowing, or not fluent in, that language. Indeed, my talks as Maplesoft ambassador in the '90s, or some private mails from Mapleprimes members, have shown me that many Maple users are left outside because of the language barrier, what is not good, I think.

Personally, I like to learn other languages, and I deeply regret not having time to learn more. And certainly, I would welcome Axel refreshing my German :)

@Axel Vogt 

I observe this kind of copy & paste crashes described by David in Classic for recent versions of Maple on Linux 32-bit (lots of error messages are thrown to the console), while not on Windows. But I am afraid that sooner or later they will reach you...