@Christopher2222 

The issues of patching and backporting have been considered e.g. in this thread.

The prefix % added to a function name makes it inert (see ?value for details). It was introduced in recent versions of Maple, and in particular, %diff is largely equivalent to Diff (in most older versions of Maple, capitalized function names frequently meant inert). Here an inert form like %diff is needed in the rule because the active diff would give an error:

der:=[ diff( f(x::name,y::name) , x::name ) = f1x(x,y) , diff( f(x::name,y::name) , y::name ) = f1y(x,y) ]:
Error, invalid input: diff received x::name, which is not valid for its 2nd argument

And the backquotes `` around diff/f make it a name.

The prefix % added to a function name makes it inert (see ?value for details). It was introduced in recent versions of Maple, and in particular, %diff is largely equivalent to Diff (in most older versions of Maple, capitalized function names frequently meant inert). Here an inert form like %diff is needed in the rule because the active diff would give an error:

der:=[ diff( f(x::name,y::name) , x::name ) = f1x(x,y) , diff( f(x::name,y::name) , y::name ) = f1y(x,y) ]:
Error, invalid input: diff received x::name, which is not valid for its 2nd argument

And the backquotes `` around diff/f make it a name.

Use Maple 15 Classic GUI. Its LaTeX export is the same as Maple 11 Classic in breaking lines. On the other hand, neither Maple 15 Standard, nor Maple 11 Standard break lines when exporting to LaTeX. 

@PatrickT 

Certainly, the comercial nature of Maple implies an disincentive for open source development. Despite that, some people went along the OpenMaple part of the issue and developed things like bindings to free languages as Python and Perl. This part seems the hardest one because of documentation problems, and possible incompleteness. The other part, the opensource libraries for development of multiplatform GUIs, is quite standard. E.g.: wxWidgets, Qt and FLTK. 

Certainly, Maplesoft shortage of in-house resources for GUI development are long-term. Simultaneously, there is a sizable need for a reliable, high performance GUI oriented towards programming. A possible solution to this conflict is offering a monetary incentive to a third party for developing such a GUI. It might involve funding a start-up company oriented towards OpenMaple based applications and providing support on the rough parts of this API.

@PatrickT Actually Maplesoft has Online Math ORACLES (beta), presumably devoted to students. It predates WolframAlpha, I think. It is based on Java, and its design is somewhat clumsy, in my opinion. Very seldom it has worked for me.

@PatrickT Actually Maplesoft has Online Math ORACLES (beta), presumably devoted to students. It predates WolframAlpha, I think. It is based on Java, and its design is somewhat clumsy, in my opinion. Very seldom it has worked for me.

WolframAlpha gets the exact solution.

WolframAlpha gets the exact solution.

Equivalent but even shorter input is:

eqs:={x^2+y^2+z^2-2*x-4*y-6*z+13 , 2*x^2+3*y^2+z^2-4*x-12*y+6*z-2}:
s := solve(eqs, {x, y}):
allvalues(s);

  {x = 1 + %2, y = 2 + %1}, {x = 1 - %2, y = 2 + %1},

        {x = 1 + %2, y = 2 - %1}, {x = 1 - %2, y = 2 - %1}

               2        1/2
  %1 := (32 + z  - 18 z)

                         2 1/2
  %2 := (-40 + 24 z - 2 z )

But, presumably, conditions on z for real solutions were required by the OP, and

SolveTools:-PolynomialSystem(eqs,{x, y},domain=real);
                2           2
  {x = RootOf(_Z  + 41 + 2 z  - 2 _Z - 24 z),

                     2    2
        y = RootOf(_Z  - z  - 4 _Z + 18 z - 28)}

does not provide them either.

Equivalent but even shorter input is:

eqs:={x^2+y^2+z^2-2*x-4*y-6*z+13 , 2*x^2+3*y^2+z^2-4*x-12*y+6*z-2}:
s := solve(eqs, {x, y}):
allvalues(s);

  {x = 1 + %2, y = 2 + %1}, {x = 1 - %2, y = 2 + %1},

        {x = 1 + %2, y = 2 - %1}, {x = 1 - %2, y = 2 - %1}

               2        1/2
  %1 := (32 + z  - 18 z)

                         2 1/2
  %2 := (-40 + 24 z - 2 z )

But, presumably, conditions on z for real solutions were required by the OP, and

SolveTools:-PolynomialSystem(eqs,{x, y},domain=real);
                2           2
  {x = RootOf(_Z  + 41 + 2 z  - 2 _Z - 24 z),

                     2    2
        y = RootOf(_Z  - z  - 4 _Z + 18 z - 28)}

does not provide them either.

@PatrickT interesting stuff for a psychoanalyst!

@mich 

You can check that most of Laurent's publications as listed here or here were clearly written using LaTeX. So, he knows it very well and the reason must lay somewhere else.

Patrick, I also keep missing primes 1. Overall, with all its problems, it was better than primes 2, in my opinion.

The first warning against wheel reinvention that I have found was this one by Jacques. I will add yours later on if I find it (for these purposes I prefer searching over my own archive).

PD: And keep this link for the thread about Classic.

@Christopher2222 Indeed, save as text is not a good option but the less bad option, in my opinion. I would not expect an option for saving the page as a web archive any time soon.