The printing of routine content into strings and searching that could be done also on the  inert representation, like tools/get_functions_called does. Browsing its code suggests to me that making work something like that with module exports should not be hard.

Whether there are kernel and external library routines including calls to standard library routines is part of my other question. Indeed the names of some builtin routines like evalf/hypergeom/kernel or int/series suggest to me that they may be involved in an interplay with library routines. And there is also Java code. Some functionality is currently implemented in Java. 

One way to read in this data into the procedure is using global names for L and w as here:

 Shear := proc (n) local i, L1, f, add; global w,L,t; f := []; add :=
`#msub(mi("Q"),mn("1"))`; for i to n do add := add+eval(w[i]*L[i],
param); L1 := add; f :=[op(f), L1] end do end proc;

One way to read in this data into the procedure is using global names for L and w as here:

 Shear := proc (n) local i, L1, f, add; global w,L,t; f := []; add :=
`#msub(mi("Q"),mn("1"))`; for i to n do add := add+eval(w[i]*L[i],
param); L1 := add; f :=[op(f), L1] end do end proc;

I am not so familiar with those methods anyway. And it sounds to me like they might have limitations if routines in the kernel or external libraries are involved in the dependency chain.

The point is a (legal) diagnostic tool that estimates the risk on the system of patching a routine. I think that such a thing is important and should be available.

I think that you cannot do it without changing the expression like:

use `*`=`&*` in 2*<g.P,g.Q> end use;
    `&*`(2, Vector(2, {(1) = g.P, (2) = g.Q}))

But automatic simplification is not a particular issue of document mode.

Honestly, I prefer worksheet mode. Yes, the typesetting code above is horrible. That is why many requests for a TeXlike syntax were posted here.

I think that you cannot do it without changing the expression like:

use `*`=`&*` in 2*<g.P,g.Q> end use;
    `&*`(2, Vector(2, {(1) = g.P, (2) = g.Q}))

But automatic simplification is not a particular issue of document mode.

Honestly, I prefer worksheet mode. Yes, the typesetting code above is horrible. That is why many requests for a TeXlike syntax were posted here.

The code below typesets this simpler matrix:

Matrix(2, 2, {(1, 1) = a, (1, 2) = c, (2, 1) = b, (2, 2) = d})

Typesetting:-mfenced(`Typesetting:-mrow`(`Typesetting:-mtable`
(`Typesetting:-mtr`(`Typesetting:-mtd`(`Typesetting:-mi`("a", italic 
= "true", mathvariant = "italic"), rowalign = "", columnalign = "", 
groupalign = "", rowspan = "1", columnspan = "1"), `Typesetting:-mtd`
(`Typesetting:-mi`("c", italic = "true", mathvariant = "italic"), 
rowalign = "", columnalign = "", groupalign = "", rowspan = "1", 
columnspan = "1"), rowalign = "", columnalign = "", groupalign = ""),
 `Typesetting:-mtr`(`Typesetting:-mtd`(`Typesetting:-mi`("b", italic 
= "true", mathvariant = "italic"), rowalign = "", columnalign = "", 
groupalign = "", rowspan = "1", columnspan = "1"), `Typesetting:-mtd
`(`Typesetting:-mi`("d", italic = "true", mathvariant = "italic"), 
rowalign = "", columnalign = "", groupalign = "", rowspan = "1", 
columnspan = "1"), rowalign = "", columnalign = "", groupalign = ""), 
align = "axis", rowalign = "baseline", columnalign = "center", 
groupalign = "{left}", alignmentscope = "true", columnwidth = "auto", 
width = "auto", rowspacing = "1.0ex", columnspacing = "0.8em", 
rowlines = "none", columnlines = "none", frame = "none", 
framespacing = "0.4em 0.5ex", equalrows = "false", equalcolumns = 
"false", displaystyle = "false", side = "right", minlabelspacing = 
"0.8em")), mathvariant = "normal", `Typesetting:-msemantics` = 
"Matrix", open = "[", close = "]", `Typesetting:-msemantics` = 
"Matrix"); 
 

The code below typesets this simpler matrix:

Matrix(2, 2, {(1, 1) = a, (1, 2) = c, (2, 1) = b, (2, 2) = d})

Typesetting:-mfenced(`Typesetting:-mrow`(`Typesetting:-mtable`
(`Typesetting:-mtr`(`Typesetting:-mtd`(`Typesetting:-mi`("a", italic 
= "true", mathvariant = "italic"), rowalign = "", columnalign = "", 
groupalign = "", rowspan = "1", columnspan = "1"), `Typesetting:-mtd`
(`Typesetting:-mi`("c", italic = "true", mathvariant = "italic"), 
rowalign = "", columnalign = "", groupalign = "", rowspan = "1", 
columnspan = "1"), rowalign = "", columnalign = "", groupalign = ""),
 `Typesetting:-mtr`(`Typesetting:-mtd`(`Typesetting:-mi`("b", italic 
= "true", mathvariant = "italic"), rowalign = "", columnalign = "", 
groupalign = "", rowspan = "1", columnspan = "1"), `Typesetting:-mtd
`(`Typesetting:-mi`("d", italic = "true", mathvariant = "italic"), 
rowalign = "", columnalign = "", groupalign = "", rowspan = "1", 
columnspan = "1"), rowalign = "", columnalign = "", groupalign = ""), 
align = "axis", rowalign = "baseline", columnalign = "center", 
groupalign = "{left}", alignmentscope = "true", columnwidth = "auto", 
width = "auto", rowspacing = "1.0ex", columnspacing = "0.8em", 
rowlines = "none", columnlines = "none", frame = "none", 
framespacing = "0.4em 0.5ex", equalrows = "false", equalcolumns = 
"false", displaystyle = "false", side = "right", minlabelspacing = 
"0.8em")), mathvariant = "normal", `Typesetting:-msemantics` = 
"Matrix", open = "[", close = "]", `Typesetting:-msemantics` = 
"Matrix"); 
 

In this case, Word takes, instead, by default an image item from the clipboard. When items in several formats occur in the clipboard (plain text, rtf, image, etc) it might be better to have a menu for deciding which item should be uploaded.

Also, links to posts laying in the second (or later?) page are currently broken.

The issue of placeholders is a "problem" of Maple handling rtable objects, not something related to MaplePrimes. You will see the same in Classic GUI edit box when selecting a Matrix output.

I would also find useful a search by regular expressions,  by date (or date range), by author, plus any of them combined. What is available now, counting back days or weeks in "Recent posts" is something too primitive!

Also, I want to be able to save complete threads in text, print-friendly form.

You could easily find posts on these and many other improvement requests with a better search facility...

See the axiom-wiki for an example of good TeX based typesetting, e.g. SandBoxMaxima.

Bryon,

Are you happy with the current search facility of MaplePrimes, or you think that you could do better to provide easier access to those valuable older threads?

The equations should be explicitly real, so that Re(iQ) should be -Q_i, etc.

But the main issue is whether the Maple numeric pde solver could handle integrodifferential equations - I do not believe so. If not, you should go back and look for fully differential equations for Q_r and Q_i.