mailing-list for TeXmacs Users

[Pierre-Henri Jondot <address@hidden>]

Hi Joris,

> 
> Yes, I introduced the fusing behavior after noting that a lot of users
> tended to create meaningless empty formulas before or after other formulas.
> I agree that this is counterintuitive from a structural point of view,
> but it avoids this very common type of error.

Understood. It makes sense, although maybe could it be possible to disallow 
empty formula, that is, automatically remove every empty formula ? I don't 
see any valid reason to keep any of them... 
> 
>> Last thing : it looks more like a (tiny) bug than a design decision : when 
>> I want to enter this kind of integral :
> 
> Note that the integral does not show up in all email clients.

Even such a small png ? Anyway the only point was that the upper bound was a 
fraction, as you probably guessed.
> 
>> then the Ctrl-F shortcut for the upper bound will be ignored once, which 
>> is not the case for the lower bound. The same is true for the square root 
>> shortcut Ctrl-S and the same behaviour applies to the discrete sum.
> 
> I don't understand.  If I do "S-F5 ^ C-f" (assuming the MacOS look and 
> feel),
> then I indeed start a fraction in the superscript.  Same if I do "\ i n t 
> return ^ C-f".

Well, I didn't understand either (but I think I might now have an idea...). 
The behaviour I describe is completely reproducible on my computer (I do use 
indeed macos keyboard layout for texmacs) : a macbook air running macos 
11.3.1, and of course TeXmacs 2.1.. I quit karabiner to see if it could make 
any change, but it didn't.

If I enter an integral sign, be it with the keyboard shortcut S-F5 S-I or 
\int return, then access the upper bound with ^, then C-f to put a fraction 
as the upper bound, the C-f has to be repeated. The first one is ignored (for 
the upper bound again, but not the lower bound)

Same thing if the upper bound is a square root : I have to type : S-F5 S-I ^ 
C-S C-S for the square root to appear.

I do have another computer, a desktop, running catalina at home, but I will 
only be there tomorrow night. I'll test this behaviour on the desktop when I 
come home.

Not being home, it so happens I do have a computer running windows 10 in 
front of me, and I just installed TeXmacs on it to give it a try. Actually, 
it felt a bit weird, and quite different, but I think I understand why :

On a pc keyboard with a french (azerty) layout, the ^ symbol appears twice. 
One is a deadkey, for accents, and the other is for the ^ symbol itself 
(altgr+9)

The deadkey ^ in a formula seems not to work, yet it does, somewhat :
supposing in a formula I type x then ^ : the cursor does not move to show 
that the focus is in the exponent of x. But the next char shows indeed as an 
exponent to x... Not that surprising actually.

For an integral symbol, it is even weirder : 
if I type : S-F5 S-I (the integral symbol appears) then the deadkey ^: no 
visual cue, then C-F and instead of having a fraction appear at the upper 
bound, I get a hat accent...

Of course, using altgr+9 for the ^ symbol and not the deadkey ^, the behavior 
is the one one should expect and wish for...

To be honest, altgr+9 to get ^ is a bit of a pain on a pc keyboard, but it 
works.

Now on a french keyboard mac layout, this is quite something else, as there 
is not a dedicated key for the symbol ^, but only for the accent. Contrary to 
windows, it does not seem to be completely a deadkey, as hitting the ^ key, 
the symbol immediately appears, yet the focus does not move.

All this to say that to observe what I did observe, I suppose that not only 
do you need tu run TeXmacs under macos, but you need to use a french mac 
keyboard layout...

Best regards,

Pierre-Henri