Vote to release 2.0.3

["Daniel F. Savarese" <df...@savarese.org>]

Hi all,

I would like to finish changing all of the version tags in the javadocs
to pick up the release version rather than the CVS version and then
package up and release version 2.0.3 and start the code freeze I mentioned
earlier.  The principal changes are Mark's case modification patch for
Perl5Substitution and a bug fix in Perl5Debug that I discovered while
investigating Dan Shriver's bug report.

In an attempt to get more in line with http://jakarta.apache.org/site/proposal.
html, I'm calling for a vote
to approve the release of jakarta-oro 2.0.3.  Technically, only
committer votes count, but heck, any valid objection will be considered.
I should allow more time to vote, but I'd like to get this release out
by Wednesday, before I go away for a few days, so please post your vote
by Tuesday 0000 GMT and state your reasons if you are -1.   As part of the
development plan, I'm proposing required prerequisites for a release centered
around regression and performance testing as well as some documentation
requirements beyond the guidelines at
http://jakarta.apache.org/site/proposal.html.

At any rate, my vote is:
+1

daniel

Re: Vote to release 2.0.3

[Takashi Okamoto <to...@kun.ne.jp>]

> At any rate, my vote is:
> +1

I want to give +1 though I don't know I can do it or not...


BTW, I have a interesting bench mark data about Hotspot. 

Jakarta ORO 2.0.1            747 281 
Jakarta Regexp 1.2          1005 721 

left:  processing time before optimized by Hotspot 
right: processing time after optimized by Hotspot

JVM: Blackdown JDK1.3fcs
OS : Linux 2.2.17
CPU,MEM: PentiumII 266,128M

I have Japanese page that introduce it, 
http://www.mycgiserver.com/~tora/java/regexp.html

Sorry, I can't explain this contents at this moment in English.
(Maybe I need more time. I'm not good at English...)

But it's clealy, if the implementation of CharStringPointer close
Regexp's CharacterIterator, the performance which is optimized by
Hotspot will be much down.

We may want CharStringPointer then good performance because the
recent computer is enough fast.

regards.
----------------------------
 ^^  	Takashi Okamoto <tora at debian.org>
= ..=	Debian Project
 -o-	  Key fingerprint:
            8B37 1FE6 76B2 7BA6 D59A  9BF7 E7F4 46C8 5293 6E17