Hej,
What matters more to me at this point is the long-term memory footprint.
The byte code compression methods you mention are applicable generally,
and thanks for mentioning them, but they have nothing to do with JiBX in
particular.
Let me take to opportunity to make some publicity myself for a static
analysis and code optimization framework called SOOT:
http://www.sable.mcgill.ca/soot/
:-)
Thanks,
Philippe Laporte
Software
Gatespace Telematics
Första Långgatan 18
41328 Göteborg
Sweden
Phone: +46 702 04 35 11
Fax: +46 31 24 16 50
Email: [EMAIL PROTECTED]
Mocky Habeeb wrote:
I'm not sure what the minimum size is but if you a have a complete set
of tests for some code that uses marshalling and unmarshalling then you
can run your tests using java -verbose:class to see a list off all the
classes that get loaded. As long as the tests are thorough and include
failure test cases (so exception classes get loaded) then you can get a
good idea of the size.
You could also use a static analyzer on the class files after you use
the binding compiler to determine all the classes that can be loaded
using marshalling and unmarshalling as entry points. Classes that are
loaded using the reflection API may be an issue. If you use something
like Dash-O it can also repackage all the required classes into a single
jar file. It can also compact the classes by renaming the symbols to
single letter names. I don’t know what the licensing implications are
for either repackaging or renaming, but if you are on a limited memory
budget this might be something you could look into.
Mocky
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Philippe
Laporte
Sent: Wednesday, August 17, 2005 6:16 AM
To: [email protected]
Subject: [jibx-users] [Fwd: size of JiBX]
Hi,
Just making sure I'm being totally clear:
What is the "minimum" size of the run-time support for JiBX excluding
external components, but of course including marshalling and
unmarshalling engines, and required classes.
Making a footprint budget...
Thanks,
Philippe
-------- Original Message --------
Subject: size of JiBX
Date: Wed, 17 Aug 2005 12:08:38 +0200
From: Philippe Laporte <[EMAIL PROTECTED]>
To: [email protected]
Hi,
I'm wondering about the smallest possible size of a JiBX.
That is:
- excluding the class generation code (code to be generated off-line)
- excluding the parser
- excluding BCEL
- excluding any non-JiBX specific components
IE, only the marshalling and unmarshalling engines.
Anyone has an appreciation of J2ME compatibility?
Thanks,
-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
jibx-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jibx-users
