I currently developing some M2M-transformations with the Atlas Transformation Language (ATL).
During studying the language constructs and properties I have often read, that ATL is a prototype for QVT, but was not used for QVT. This leads me to the question, if ATL is a deprecated language.
Personally, I claim that ATL is not deprecated because there are some indications for this:
-
If I look at the website of ATL the last Update was August, 26th 2011. The first specification of QVT was 2008. So I think the development of ATL seems to proceed.
-
There are also some papers since 2011 on google scholar:
http://scholar.google.de/scholar?q=atl+m2m&btnG=&hl=de&as_sdt=0&as_ylo=2011 -
It is still part of the M2M-frameworks of Eclipse
What would you say? Would it be better to user another M2M language or is ATL future proof? If not, do you know possibilities to transform ATL into other M2M-languages?
1
In order to be deprecated, someone (that is, whoever maintains it) has to deprecate it. That doesn’t seem to be the case here, as OBEO (listed on Wikipedia as a maintainer) still lists it in the Elicpse market as a current offering. It looks like it was updated last month.
So, no, it’s not deprecated. Whether it’s a good choice for whatever you want to do and how far into the future it might be maintained are separate questions and probably not on topic here.