@Joe Riel 

Yes, once you realize what is going on, without any example/warning in the help page. Not very helpful for most users... About in vs =, yes, debugger rendering is an issue, but only for a tiny minority like us.

@Kitonum 

The problem with that help page is not showing this single argument syntax as new. Similar problem as in this thread.

@Joe Riel 

Would you stick because of syntax (e.g. clarity) and/or performance? In terms of conceptual clarity, I would prefer add( x*ln(x), x in X ). On the other hand, I find useful your remark as the new syntax add(U), here U a list, seems poorly identified as new in ?add or elsewhere. In particular, no example for this syntax seems available.

@Axel Vogt 

They must be using something else, as all these tools identify correctly my location and the advertising I get is clearly local, never German,  French or something else. The only site I know having problems with my location is maplesoft.com.

@Thomas Richard 

Not strange, but frequent. Actually this issue has been discussed here several times, see e.g. here. In my experience, all the sudden English default is replaced by German (fewer times by French). Fortunatelly, I can read any of these three languages :)

@acer 

Take the example of ?cat. It states in Maple 18:

Calling Sequence

cat( a, b, c, ... )

Parameters

a, b, c, ...-any expressions

And then the example in ?updates,Maple2015,Language:

Enable concatenation of a string with a constant:

cat("x=", 2.5);

Now, objects of type string (like "x=") or type constant (like 2.5) are expressions, OK? Then, not working in Maple 18 and earlier means that the Parameter section has been wrong. So, the first problem is that the complete detail of allowed types for the arguments is not given. If ?cat had informed earlier that the combination string with a constant was not allowed, then you could have added now the new allowed combination in ?cat and provide optional highlighting for that entry as Mathematica does.

This problem of missing type information in the help pages is endemic, and I have complained about it for years, but nobody at Maplesoft seems to care. Fixing this issue is necessary for diverse additional improvements, as the case here.

@Bendesarts 

For a module-based pakage, the output of with shows only the exported procedures (that were not hidden from view), in this case show. Code of the non-exported procedures, if they exist, can be seen by executing first the command kernelopts(opaquemodules=false), or by using the :: syntax, see ?showstat .

@Axel Vogt 

Electrical engineers have been using j as standard notation for the imaginary unit for very long. The problem in their context is rather that i or I are used for electrical current.

@MartinUser 

The code of asympt (more precisely the code of the procedures in the module asympt) does not call IntegrationTools:-Parts, while it makes calls to limit and family. In this sense, the failures of asympt seem more related to limitations in series than Parts. Note also that the code of asympt is copyrighted 1991 (and a bit updated in Maple 15) while Parts is available since Maple 10, year 2005. Another thing is that a better "asympt" could need a more powerful Parts.

@acer 

Yes, the derivative divides sector of the code has been updated.

@Markiyan Hirnyk 

Do you know whether these problems are specific of this version of Mathematica? Namely, whether it is a regression or a long standing issue.

@Carl Love 

Debugging shows that evalhf is tried first, but an exception is raised (Error, evalhf cannot use _passed or args) and the soft route is followed in the procedure `plot3d/surface`, where it gets stuck:

> stopat(`plot3d/surface`,5):
> p1 := plot3d(J, Ox-w .. Ox+w, Oy-w .. Oy+w, grid = [R, R], shading = ZHUE, style = PATCHNOGRID, orientation = [-90, 0], axes = BOX, labels = [x, y, iter]);
""
`plot3d/surface`:
   5*  if _plotDigits <= evalhf(Digits) then
         ...
       else
         ...
       end if;
DBG> into 2
""
`plot3d/surface`:
   6     try
         catch "cannot assign a complex value to a hardware float %1":
           ...
         catch :
           ...
         end try
""
`plot3d/surface`:
   7       result := evalhf(`plot3d/sgridhf`(ff,var(xa),a,b,c,d,m,n))
DBG> 
""
`plot3d/surface`:
  10       result := `plot3d/sgridf`(ff,xa,a,b,c,d,m,n)
...

Your workaround helps in avoiding to get stuck in `plot3d/surface`, which finally returns an error code. In this sense, may be that it works by chance.

@Mac Dude 

No, what Carl has found is a workaround. Moreover, Carl's workaround seems to confirm what I have said, as the usage of the evalhf version bypasses the route of numerical computation of logarithms described above. On the other hand, here Maple 12 on Linux 32-bit executes it as fast as Maple 13.

@Mac Dude 

There is not so simple correlation between what you feel as large and complex and what is large and complex for the Maple JVM. Again, much smaller worksheets of O(10^5)B storage size and no plot may suffer of the no open/save syndrome. If you were really interested, I may give you details by private mail.

Besides, a key issue is about the default maximum heap size being set to 512MB. If raising it to 1 or 2 GB were the cure for many such operational/performance problems, why it is not set higher? Or at least, why instructions are not provided to raise it as needed or when possible? For instance, I know cases of small worksheets producing some 10 "ordinary" 3D plots that do not finish saving with the default heap size setting. And increasing the heap size does make the process of save finish, but the resulting worksheet does not show any plot when opened. 

Conclusion, raising the maximum heap size is not documented and it may produce unexpected results. In this situation I take it as a potentially risky workaround and I think that it should be used only experimentally. 

@Mac Dude 

Certainly, Standard worksheets with plots (many/complex) are frequent candidates of the no-open syndrome. But I know of worksheets with storage size below 1MB and with no plot exhibiting the same syndrome. And I know of worksheets of this same class that open but do not save.