jakarta-bsf-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Rony G. Flatscher" <Rony.Flatsc...@wu-wien.ac.at>
Subject Re: A little remark ad "compatibility" of BSF3 ... (Re: svn commit: r757581 - in /jakarta/bsf/trunk/bsf3: ./ testing/jruby-1.1.2/ testing/jruby-1.1.2/src/test/java/org/apache/bsf/testing/javascript/ testing/jruby-1.2.0/ testing/jruby-1.2.0/src/test/j
Date Tue, 24 Mar 2009 16:04:43 GMT
sebb wrote:
> On 24/03/2009, Rony G. Flatscher (Apache) <rony@apache.org> wrote:
>>  sebb wrote:
>>  > On 24/03/2009, ant elder <ant.elder@gmail.com> wrote:
>>>>  That sentence is a direct copy from an email when this was discussed a
>>  >>  while back [1], though I agree it could sound better. How about
>>  >>  changing "is not a compatible implementation" to "can not claim to be
>>  >>  a compatible implementation", so:
>>  >>
>>  >>  "Apache BSF 3 is an implementation of JSR-223, the Scripting for the
>>  >>  Java Platform APIs. Note that this software hasn't been tested with
>>  >>  the JSR 223 TCK, and therefore can not claim to be a compatible
>>  >>  implementation of JSR-223."
>>  >>
>>  >
>>  > Seems good to me, though I think it could go further.
>>  >
>> +1
> In particular, it would be useful to mention that this implementation
> runs on Java 1.4+ whereas AIUI Java 1.6 is the first version which
> includes the API.
Yes, I think this is a very important piece of information for all those
Java coders who have to address Java 1.4+ environments in the business

Also, AFAIK, Harmony uses BSF3 for "javax.script", which might be an
interesting tidbit.

> Should probably also mention that this means that the API does not use
> generics (not sure how many classes this affects).
Hmm, this should not be a problem due to signature erasing in the byte
code (which therefore does not possess any information about the defined
generics in the source code).

"generication" of the "javax.script" JSR-223 was introduced quite late
in the process, if I remember correctly.


  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message