[Seaside] OSCON report

Wilkes Joiner wilkesjoiner at gmail.com
Sun Jul 30 01:48:44 UTC 2006


Thanks for the update Avi.  Good to hear.  I've been convinced for a
long time that the more attention Ruby/Python get the better off
Smalltalk is.  It will help to break the static branches of
programming and will bring a larger mind share into the late-binding
camping.

I've got a screencast in the queue to address the issue of learning
Smalltalk.  This is motived by Ralph Johnson's post
(http://lists.squeakfoundation.org/pipermail/squeak-dev/2006-May/104071.htm).
 Smalltalk is so radically different that the conventional ways of
learning a new programming language are not applicable.  Tutorials and
api documentation are the main ways of learning most envirionments.
In Smalltalk, the trick is learning how to browse the image and
experiment with running code using the workspace and the inspectors.
I think this is more effective once you figure out how to do it.

The thing that is killing me with Rails is not being able to read the
code I'm relying on.  I can open a text on editor on the source once I
find it, but this brutally primitive compared to what comes for free
with Smalltalk.  Porting ActiveRecord to Smalltalk with its own
browser would really leap frog what is available in Ruby.  Someone
will write a good IDE for Ruby and Rails.  It's inevitable.  However,
given the heavy reliance on runtime code generation in both Ruby and
Rails,  it seems to me to be a really challenging prospect.  Not only
do you have to deal with dynamic typing.  The definition of a lot
classes only exist at runtime.  In Smalltalk, we can do that, but it
is done far less frequently than in Ruby. I would rather just port
ActiveRecord.

I'll get a screencast together tomorrow and post a link on this list
and squeak-dev for feedback.

- Wilkes

BTW, I prefer Sails to StOR.

On 7/29/06, Avi Bryant <avi.bryant at gmail.com> wrote:
> Since this StOR thread seems to be about reaching an audience we
> haven't had before, it's probably worth mentioning what I've been up
> to for the last week.
>
> I've just returned from the O'Reilly's open source OSCON conference.
> I gave a talk on Seaside very intentionally angled not towards
> convincing people to jump ship from Python/Perl/Rails etc, but rather
> towards getting them to consider some of the basic technologies -
> like the the Canvas, Component, Callback and Continuation model (hm,
> never noticed all those C's before) - for use in their environment of
> choice.  However, interest was high enough that:
>
> - I gave a 45 minute talk to a packed room (some good photos at
> http://flickr.com/photos/jacobian/)
> - followed by about 30 minutes of Q&A with most people still in the room
> - followed by 3 hours of conversation with a large group at dinner
> about Seaside and Smalltalk
> - followed by a subset of that group tracking me down first thing the
> next morning and making me spend another couple of hours doing a
> Squeak demo
>
> What I found is that a) a huge number of these people have tried
> Squeak before, and that b) they were all very excited by what they
> saw in my (very simple) demo, but none of them had been able to make
> the journey themselves from the one to the other - that is, it was
> impossible for them to just download a Squeak image and, with no
> prior Smalltalk experience, find their way through a Hello World
> Seaside app.  So they decided Squeak wasn't for them.  There's also a
> huge amount of confusion and disinformation about the state of things
> like version control: essentially everyone seems to believe that the
> image is the only way to distribute Smalltalk code, and is
> understandably leery of this.
>
> I sensed enough goodwill, and got enough concrete offers to publish
> articles/books/etc, to think that the opportunity is there to get a
> mainstream audience for Seaside *if* (and this is a fairly large if)
> we want it and have the resources to put into it.  What it would
> take, I think, is a custom Squeak distribution and tutorial/
> screencast that was aimed at a non-Smalltalk audience.  It would need
> to cover:
>
> - The browser
> - The workspace
> - Saving, loading, and merging in Monticello
> - The basics of Seaside (a revamped "Walk on the Seaside" tutorial
> would be fine)
>
> An ActiveRecord-like simple O/R framework would certainly help too,
> but I think the above is a higher priority.
>
> Incidentally, I was asked by a few people at the conference how big
> the Seaside community was, and I was very pleased to be able to say
> "well, I almost never post to the list anymore, because other people
> are doing all the question answering, discussing, and committing".
> That's pretty cool, and an important threshold to cross, I think.
>
> Avi
> _______________________________________________
> Seaside mailing list
> Seaside at lists.squeakfoundation.org
> http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
>


More information about the Seaside mailing list