[Ometa] OMeta2 for Squeak?

[Enrico Spinielli]

Hi Alessandro,
is the sar in [1] pointing to your newly bootstrapped OMeta2?

Bye
Enrico

[1] http://tinlizzie.org/ometa/ometa2.html

On Sat, Apr 3, 2010 at 09:58, Alessandro Warth <alexwarth at gmail.com> wrote:

> the OMeta2 for Squeak (http://tinlizzie.org/ometa/ometa2.html) is
>>> designated as a release candidate, but it has some rough edges, and
>>> AFAIK it never has reached the "release" stage.
>>> Is there any news on that? I'm currently using it for a project, and I
>>> like the new syntax much better (less tedious typing).
>>> However, the production rules implementing it show as Squeak source
>>> code,
>>
>>
>> I wrote the rules that you're talking about by hand in Smalltalk, using
>> the OMeta classes as a library. My intention was to eventually rewrite t=
hem
>> in OMeta -- as I did in the old implementation -- to complete complete t=
he
>> bootstrapping. This should be very straightforward, and you're welcome to
>> give it a try if you want. Otherwise, I'll try to get around to it soon.
>>
>
> FYI, I've already been able to make a lot of progress on completing the
> bootstrap, and expect to be able to finish it early next week (possibly
> Monday afternoon).
>
> Cheers,
> Alex
>
>
>>
>>
>>> and if you rewrite them in OMeta2, you see that they obviously
>>> have been hand-massaged (OM2 generated methods show suspicious "true
>>> ifTrue: []" expressions for #And clauses).
>>>
>>
>> "Suspicious"? That that back! :)
>>
>> This "true ifTrue: [...]" idiom is just a straightforward (and reasonably
>> efficient) way to make an expression out of a list of statements, which =
was
>> something that I needed to be able to do in the target language in order=
 to
>> compile OMeta rules. Another alternative that I considered was "[...] va=
lue"
>> but that was a bit slower. (The block is compiled away for the former, b=
ut
>> not for the latter -- try it and look at the bytecodes, you'll see what I
>> mean.)
>>
>> Since it has been almost a year since the RC, it would be nice to get
>>> this to move a little forward, so that the Squeak/Smalltalk users of
>>> OMeta can benefit from the new syntax and features, too.
>>> I'd be willing to spend some time on this, if there is interest.
>>>
>>
>> The real reason why there hasn't been a proper release of OMeta2 is that=
 I
>> was never able to find a reasonable way to make Monticello work w/ a pro=
ject
>> like OMeta. The problem is that things must be loaded into a Squeak imag=
e in
>> a certain order -- e.g., the code that implements the OMeta parser and c=
ode
>> generator in Squeak must be loaded before any code written in OMeta synt=
ax.
>>
>> This is something that I would love some help with. Are there any
>> Monticello experts on this list who would like to volunteer to package t=
his
>> new version of OMeta/Squeak?
>>
>> Thanks!
>>
>> Cheers,
>> Alex
>>
>>
>>
>>
>>>
>>> Cheers,
>>> Hans-Martin
>>>
>>> _______________________________________________
>>> OMeta mailing list
>>> OMeta at vpri.org
>>> http://vpri.org/mailman/listinfo/ometa
>>>
>>
>>
>
> _______________________________________________
> OMeta mailing list
> OMeta at vpri.org
> http://vpri.org/mailman/listinfo/ometa
>
>


-- =

Enrico Spinielli
"Do Androids dream of electric sheep?"=97 Philip K. Dick
"Hear and forget; see and remember;do and understand."=97Mitchel Resnick
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://vpri.org/pipermail/ometa/attachments/20100412/34da9a22/attachme=
nt.htm