It looks like that Maple tolerates Windows better than Linux. However, it is still not perfect.

@C_R 
Thank you
No, I have no Java installed. I'll test Maple on the same laptop on Windows (my laptop has dual boot).

PS
The Maple GUI has never been one of Maple's strong points. However, a decade ago, when I had less powerful hardware, the Maple GUI operated smoothly )

@Alejandro Jakubi 

Laurent is not GUI developer.
I suppose, he rather deals with large projects and thus works within Maple IDE from Digiarea where this problem doesn't appear. Here he says that really uses IDE.

BTW
he noticed (in 2011) that they "were looking into providing a desktop shortcut that defaults to 1D, in order to make this even easier." However, this shortcup is still absent.

@Carl Love 
Thank you.

It seems that Maple GUI develpers do not use Maple themselves.

Thank you for reply.
My question is specifically about cutting and pasting code from Maple's Standard GUI. Sorry for my misleading original question.

The problem appears when I copy the code presented in math mode and then paste it in any text editor. The same problem rises at converting the code from math mode (2D-Math) into text mode (Maple input or 1D).

I've attached a zip file comprising a maple worksheet and a text file. Worksheet contains two pieces of code: the first on is presented in math mode (2D) and the second one is presented in maple input mode. The text file contains the result of copying/pasting of the contents of Maple worksheet.

strange.zip

@Rouben Rostamian  
My OS is Windows, text editor is Notepad++ or standard Windows Notepad.

@Carl Love 

As I understand it is a relatively new interface variable. Maple documentation hides information about it very well.

@Christopher2222 

Thank you for the answer. But it is not what I am looking for. If you previously called, for instance, 

currentdir(kernelopts(homedir))

then the path to a dir where current document is located gets lost. Of course we can do something like this

st:=currentdir()
...
currentdir(kernelopts(homedir))
...
currentdir(st)

but what to do if we don't call st:=currentdir() at the beginning? I'll be wondered if correspondent variable doesn't exist.

@ecterrab 
Could you give me an example of an expression that is a product of commutative and noncommutative variables where the first operand is NOT commutative? I could not construct such expression.

The simplest example:

restart; 
with(Physics): 
Physics[Setup](op = {A, B}):
z := (A*k*(B/r))/g; 
op(z);

returns

k, 1/r, 1/g, A*B

@ecterrab

Now difference between Physics:-`*` and Physics:-`.` is clear.

Regarding my question
z1:=Physics:-`*`(A,B,C)
is equal to
z1:=Physics:-`*`(A,Physics:-`*`(B,C))
so formally z1 matches the type 'Physics:-`*`'(anything,anything).

Such was my thinking. However, as I now understand, equal here is not equivalent.

A desired output for the case when an argument has the form z:=add(k[i]*W[i],i=1..N) where N is some integer number, k[i] is a commutative part, W[i] is a quantum operator such that expand(W[i])=W[i] (exactly my case) can be achieved with the function f presented below:

restart; 
with(Physics);

Physics[Setup](op = {A, B}, quiet); 
`type/ProdComOp` := `&*`(commutative, specfunc(anything, Physics:-`*`)); 
z0 := u*A.B-A+k*B+B.A.B+g; 
z1 := z0+v*A.B.A; 
z2 := z0-v*A.B.A; 

f := proc (z) 
   local A, z2; 
   Physics[Setup](op = A, quiet); 
   z2 := map(abs, z+A); 
   indets(z2, ProdComOp) 
end proc; 

S1 := f(z1); 
S2 := f(z2)

Result:

S1:={u*A.B, v*A.B.A}
S2:={u*A.B, v*A.B.A}

@Alejandro Jakubi
Really? In such case I'll avoid a picture-form posts at all.

@Alejandro Jakubi 
I was sure that my problem could be easily solved by experienced user without running Maple. The reason I presented a picture-form is that it gives easier perception of the code then a typed text. As I expect to get help from community with my problems I'll present my problems in as comfortable way for readers as possible. So, if you say that form of my post doesn't encourage you to help me then it is not a problem to correct this post and to provide a code in visible text form here and in future.

@acer 
oh... of course `Or`. Thank you.

@Carl Love 
Thank you for explanation. It was not clear from the help page.