Where xTalk Excels (Was Re: OT: Adobe kills Director)

[Dan Shafer]

There have been some wonderfully eloquent descriptions in this thread 
of why some of us prefer xTalk to more conventional programming 
languages. I liked the poetry vs. novel, sprint vs. marathon analogy a 
lot.

For me, writing software and writing prose and poetry are very similar 
processes. I spend virtually all of my time engaged in one or the other 
of those creative pursuits. And, for me, xTalks are the only languages 
that get out of the way and let me create. I don't have to spend a huge 
amount of time thinking about syntax, structure, and other elements of 
the language. i just use it. This comes from a couple of features of 
xTalks that are not generally shared with other languages: their 
human-like syntax and their encouragement of verbosity.

Many years ago, Adele Goldbert, a Smalltalk pioneer, wrote a wonderful 
paper for the ACM called "Programmer as Reader" in which she argued 
persuasively that since most programs spend far more of their life 
cycles in maintenance mode than in development mode, it was important 
that programs be inherently readable. Smalltalk took a very 
sentence-structure approach to creating readability and it is that 
feature which drew me to it.

Transcript code tends to be more readable than code written in other 
languages, in part, I think, because the syntax itself is verbose. 
Writing "put 14 into field 'age'" puts the coder in a different mind 
set than "age=14". At least it seems to in my head.

Perhaps in part because of this human-readability and verbosity (I 
suspect the two go hand-in-hand), it is far more often the case that a 
Transcript statement does what I expect it to do even when I'm unsure 
of the syntax than is true of most other languages (notable exception 
for me being Python). I'm amazed as I work on projects how often I go 
through this process: (1) I wonder how to say that in Revolution; (2) 
I'll try it this way and see what happens; (3) what do you know, it 
worked. This is in part, of course, because the very high-level nature 
of the language means that step 2 never (at least in my experience so 
far) causes any catastrophes. (That possibility is what ultimately 
caused my uneasiness with Smalltalk, where it was at least 
theoretically possible that I could munge up the entire Smalltalk 
image, not just my app, since I was never isolated from it.)

People who don't program for a living and who are generally not trained 
in computer art (I'm not sure it's really a science, are you?) 
generally find xTalk languages more comfortable, inviting and 
productive. And it's good to hear from some real professional coders 
who are finding the same experience.


On Apr 22, 2005, at 8:11 AM, Jesse Sng wrote:

>> You know, Jesse, I COMPLETELY get it now.
>>
>> xTalk is the main reason why I decided to check out (and then buy a
>> copy of) RR a couple of months ago.  I'm completely hooked on the
>> language, which is stupid and crazy, since I've worked in
>> 30+languages, most of them BEFORE xTalk, and I had more-or-less given
>> up on xTalk for the last 10 years.
>
> I've been in the same situation. I could do the other stuff better 
> than most people but there was something about XTalk that was 
> unmatched.
>
> I compared it to writing novels vs writing poetry. XTalk development 
> was like poetry, you do it in small chunks, in small, elegant bits and 
> instead marching down the timeline of a project, you could sprint in 
> short spurts and then get to enjoy the scenery at the same time.
>
>> It isn't rational to like xTalk more than BASIC.  OK, no, that's so
>> not true.  I'm shutting up now.  I actually was pondering writing an
>> xTalk interpreter/cross-compiler earlier this year until I discovered
>> RR.  With the innovations that SC brought after I gave up on it that
>> were then incorporated into RR I'm glad I didn't.  Of course, RR has
>> quirks in it that are extremely annoying, but with some suggestions
>> from Richard, I'm hopeful that I can ultimately overcome them.
>
> Folks used poo-poo the stuff I did in SC, saying that they could 
> better with all their advanced tools and superior brains, but nothing 
> ever came out of those projects, while I moved it from black and white 
> Hypercard to Hypercard with 24 bit colour using my own externals, to 
> SC.
>
> While lots of things took very little code in SC, you can build very 
> sophisticated things with much code. Getting to enjoy Rev a bit more 
> and it helps when you try it out, do some stuff, back off, think a bit 
> before going back to it some more.
>
> Getting into Rev isn't quite the same as SC or HC, but it isn't 
> difficult either.
>
> Ultimately, I think it's because the development model better suits 
> the playfulness in my mind when I'm initiating a new project. It 
> invites the playfulness and the experimentation, very much like 
> messing around with your kids' Lego bricks.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dan Shafer, Co-Chair
RevConWest '05
June 17-18, 2005, Monterey, California
http://www.altuit.com/webs/altuit/RevConWest