Re: [ANN] Lua For Windows v5.1.3.12 released

["Andrew Wilson" <agrwagrw@...>]

Yes, Sébastien,

 I'll have to get Steve Donovan to comment specially about Scite and
Lua debugging, interactive sessions etc. but we'll definitely try to
fix these kinds of issues. And yes the difference between interactive
lua session versus running/debugging lua program is not clear, since
what I've experienced is that most debug systems wouldn't also have an
interactive session type mode. There does seem to be some bugs/issues
going between these two modes that will have to be explored a little
more.

Regards
Andrew

On Wed, Jul 9, 2008 at 2:57 PM, Sebastien Loisel <loisel@temple.edu> wrote:
> Dear Andrew,
>
> Thank you for your email.
>
>> Thanks for the feedback, we'll make improvements based on feedback.
>
> That sounds great. Just now, I decided to see what else I could find out, so
> I played around with SciTE and opened hello.lua in the editor. It seems like
> the "Run Program" and "Launch Interactive Lua" buttons interact poorly. When
> the program ends, it seems like the Lua session is killed. Then, I can click
> anywhere in the terminal pane, and the cursor goes there. I can edit the
> contents of the terminal pane, but that does not seem to make any sense.
> Then, I click on "Launch Interactive Lua" again, and I couldn't see a new
> prompt. I was typing in the terminal pane, but the cursor was somewhere
> random. After typing some more, I eventually got a prompt, and then I had
> the following interaction:
>
>>print "hello, world"
> variable 'rint' is not declared
> stdin:1: attempt to call global 'rint' (a nil value)
> stack traceback:
>     stdin:1: in main chunk
>     [C]: ?
>> print "hello, world"
> 'hello, world'
>
> So apparently, it lost the 'p' of my 'print' on the first try. Then, I tried
> the "Load Lua File into Interactive Session" button. When I clicked it, I
> got the following mysterious message:
>
>> Lua 5.1!
> true
>
> which looks like only a fraction of the hello.lua program.
>
> I don't know if it's my place to make suggestions, but I'll try one and you
> just tell me if you don't need my suggestions.
>
> I was thinking that it doesn't really need to work like this. As long as the
> console pane is open, it should be bound to a given Lua interpreter. You
> don't really need the button to start an interpreter -- although if you want
> multiple interpreters, you could have an "open new console pane" button, but
> each console pane should be permanently attached to an interpreter, in my
> opinion. So if you hardwire the single console pane into your editor, you
> don't need the button, I think.
>
> For the other two buttons, it seems like they attempt to do the same thing
> two different ways. Again, in my opinion, you should have a button that
> copy-pastes the entire buffer into the console window, and that's all you
> need. If you look at MATLAB, it also has a variety of good ways of doing
> this (are you familiar with MATLAB?) In MATLAB, what you can do is select a
> portion of your buffer, then hit F9, and it just pastes it into your console
> pane. In addition, MATLAB has something they call "cell mode". In Cell Mode,
> your script is divided into cells. Cells are separated by comments that
> begin with %%. When you hit CTRL-Enter, it automatically copies and pastes
> the entire contents of the cell where your cursor is, to the console pane.
>
> I'm not saying that's how you have to do it, I'm just saying that there's
> room for different ways of executing the code you're editing, but it seems
> to me like what your editor is doing now is... not following the principle
> of least surprise?
>
> Thank you very much for LfW, it is a very interesting program and I am
> looking forward to future versions and improvement!
>
> Sincerely,
>
> Sébastien Loisel
>
>
> --
> Sébastien Loisel