[racket-dev] [plt] Push #23587: master branch updated

From: Robby Findler (robby at eecs.northwestern.edu)
Date: Thu Sep 22 12:34:36 EDT 2011

Oh, if what you're saying is that we never had a transparent REPL all
along, then fine. We still don't. (I don't know about "blessing".
That's something that type systems and priests do and I'm neither of
those.)

BUT, in my mind, in a practical sense, what we have a better than the
old Emacs style REPL because you see the sequence of expressions in
the REPL. I don't know how things are in recent days, but in the past,
sending expressions to the REPL in Emacs wouldn't actually print the
expressions in the REPL transcript; you'd just get a series of
prompts. Now, at least, you (or your instructor/TA) can scroll back up
and say "you're doing something dumb here... click Run".

Robby

On Thu, Sep 22, 2011 at 11:28 AM, Matthias Felleisen
<matthias at ccs.neu.edu> wrote:
>
>
> How does this preserve transparency? Yes, you can break transparency with a handful of keystrokes of your own, but when you do so, you know you're doing it. Now you're supporting it, blessing it.
>
>
>
>
> On Sep 22, 2011, at 12:25 PM, Robby Findler wrote:
>
>> On Thu, Sep 22, 2011 at 11:17 AM, Matthias Felleisen
>> <matthias at ccs.neu.edu> wrote:
>>>
>>> On Sep 22, 2011, at 11:52 AM, robby at racket-lang.org wrote:
>>>
>>>> | Add the following keybindings in a (hopefully) transparent REPL-friendly way:
>>>
>>> HOW?
>>
>> They keybindings automates what you can already do with some three or
>> four or ten or seventeen keystrokes currently.
>>
>> For example, send-selection-to-repl is equivalent to this (on a mac
>> and assuming that the defs and interactions are both initially visible
>> and you have something selected (the keyboard focus also has to be in
>> the definitions, but that's a pre-req for send-selection-to-repl too))
>>
>>  cmd-c
>>  cmd-d
>>  esc;control->
>>  cmd-v
>>  return
>>  cmd-d
>>
>> except, of course, that send-selection-to-repl doesn't actually
>> clobber your copy/paste buffer like the sequence above does.
>>
>> There are, of course, alternative sequences that are also equivalent;
>> that's just the one I probably would have used if I wanted to do such
>> a thing.
>>
>> Robby
>
>



Posted on the dev mailing list.