ATS and Groovy
50 views
Skip to first unread message
Brandon Barker
Dec 2, 2014, 6:49:16 PM12/2/14
to ats-lang-users
This is just a thought, not a serious interest at this point. Since Groovy targets the JVM and is supposedly a superset of Java, to some extent, it adds additional flexibility to the programmer-centric methodology by allowing easy use of optional types, followed by normal static types in Java, and finally using types in ATS for the most rigor, all while using the JVM. But -- aside from those wanting to target the JVM -- is this really worth it, or does it make better sense to just do co-programming in ATS and a completely dynamically typed language? I hope to get some actual experience with ATS co-programming soon, probably with Perl, so maybe I'll have a better sense then.
I'm aware that currently there is no code generator for Java. I don't know if a code-generator for Groovy would be the right way to go for efficiency reasons, but if it isn't much different, it seems like that might be preferable.
The caveat is, I don't know to what extent Groovy is really a superset of Java, so I don't know if by supporting Groovy code generation you could also claim to have Java support.
gmhwxi
Dec 4, 2014, 3:31:30 AM12/4/14
to ats-lan...@googlegroups.com
I had a brief encounter with Groovy last.
I think it makes a lot of sense to target Groovy (instead of Java).
I prefer targeting untyped languages in general as it means no need
to fight typechecking in the targeted language.
I am willing to implement atscc2groovy if someone can promise me
to use it to do a (small) project.
PS: By the way, targeting clojure makes a lot of sense, too :)
I think it makes a lot of sense to target Groovy (instead of Java).
I prefer targeting untyped languages in general as it means no need
to fight typechecking in the targeted language.
I am willing to implement atscc2groovy if someone can promise me
to use it to do a (small) project.
PS: By the way, targeting clojure makes a lot of sense, too :)
gmhwxi
Dec 4, 2014, 3:32:55 AM12/4/14
to ats-lan...@googlegroups.com
>>I had a brief encounter with Groovy last.
Last year.
Brandon Barker
Dec 4, 2014, 12:23:23 PM12/4/14
to ats-lang-users
On Thu, Dec 4, 2014 at 3:31 AM, gmhwxi <gmh...@gmail.com> wrote:
I had a brief encounter with Groovy last.
I think it makes a lot of sense to target Groovy (instead of Java).
I prefer targeting untyped languages in general as it means no need
to fight typechecking in the targeted language.
I am willing to implement atscc2groovy if someone can promise me
to use it to do a (small) project.
My current goal is to practice doing GUI programming in Java or Groovy (don't care which, the main issue is getting used to GUIs, and Groovy may be simpler for this due to being more concise).
As an example app using co-programming, I was thinking of keeping the GUI in Groovy (at least for now) to avoid having to model the API (SWT) in ATS, especially since some people may like Swing more (I might be one of them... packaging swt libs for different platforms is a bit annoying using java build systems). The underlying logic could be a simple calculator in ATS, ported from the existing calculator demo. I do not know if that is of sufficient interest, but in principle it could be a good template for many GUI apps.
Otherwise, I plan to just use JNI, which is fine for me as well for this purpose. Sorry I don't have anything more interesting at the moment. I have an existing (closed-source) app involving cattle simulation in Java, but that is not my baby, so it is too early to say if I could use it there.
PS: By the way, targeting clojure makes a lot of sense, too :)
Yes, I can imagine that would make porting to other lisp-like platforms rather easy as well. I haven't looked at it much though, and am less sure if it makes as much sense for co-programming, but I don't have any specific reservations; I just know it is quite easy to mix in Java-like code in Groovy. It doesn't seem to fare well here: http://benchmarksgame.alioth.debian.org/u32/clojure.php Unfortunately Groovy is not an option.
On Tuesday, December 2, 2014 6:49:16 PM UTC-5, Brandon Barker wrote:This is just a thought, not a serious interest at this point. Since Groovy targets the JVM and is supposedly a superset of Java, to some extent, it adds additional flexibility to the programmer-centric methodology by allowing easy use of optional types, followed by normal static types in Java, and finally using types in ATS for the most rigor, all while using the JVM. But -- aside from those wanting to target the JVM -- is this really worth it, or does it make better sense to just do co-programming in ATS and a completely dynamically typed language? I hope to get some actual experience with ATS co-programming soon, probably with Perl, so maybe I'll have a better sense then.I'm aware that currently there is no code generator for Java. I don't know if a code-generator for Groovy would be the right way to go for efficiency reasons, but if it isn't much different, it seems like that might be preferable.The caveat is, I don't know to what extent Groovy is really a superset of Java, so I don't know if by supporting Groovy code generation you could also claim to have Java support.
--
You received this message because you are subscribed to the Google Groups "ats-lang-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ats-lang-user...@googlegroups.com.
To post to this group, send email to ats-lan...@googlegroups.com.
Visit this group at http://groups.google.com/group/ats-lang-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/ats-lang-users/d02b0878-e974-46e2-90b4-1508e97ce690%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages