renggli at gmail.com
Mon May 14 06:20:12 UTC 2007
> I, as a lot of people, are using the excellent job you made "mapping"
> such a thing I wanted to ask what is your opinion about the starting point
> of such kind of work for other toolkits for smalltalkers. I mean, I have
> idea of what to expect of the development in the formula
> more deeply the "path" to them.
first I must say that the Scriptaculous package is not entirely my own
work, Avi started it all back in June 2005. In the beginning there was
just basic AJAX functionality and some effects. The evolution of
Scriptaculous was a highly iterative process. While I tried to add
more functionality, and eventually supported all the things available
Smalltalk object model several times to accomodate new needs.
language, even though most libraries (including prototypes (sic!) and
script.aculo.us) mostly use it like a class based system. So the first
thing to notice is that the Smalltalk object hierarchy does not
For example there were some conflicts in SUAjax and SUAutocompleter,
where some Smalltalk objects understand messages they shouldn't. I
guess this is not really a problem, and if necessary it could be
cleanly solved using some well known technique that should not be
mentioned in the Seaside mailing-list ;-)
method has a Smalltalk equivalent + some Smalltalk glue code. I added
a lot of unit tests to make sure the transformation works as I
expected. Also nice is to have a functional test suite that makes sure
serves as an example for users of the library.
> To be honest I'm interested in a couple of widgets of the dojo
> want to "map" a smalltalkish access to them into Squeak for instance, which
> should be my starting point? how the normal cycle of development should be
> that are good to smalltalkers, etc)?
I would just start with something simple that works and then refine it
on the go. I am sure that you can take some ideas from Scriptaculous
and maybe even build on top of it reusing its serialization scheme.
More information about the seaside