Mailing List Archive

What JDK level should lucene code be compatible with?
When adding/fixing code in the main source tree?

thanks
eric


--
To unsubscribe, e-mail: <mailto:lucene-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:lucene-dev-help@jakarta.apache.org>
Re: What JDK level should lucene code be compatible with? [ In reply to ]
Eric Fixler wrote:

> When adding/fixing code in the main source tree?
>
> thanks
> eric
>
>
> --
> To unsubscribe, e-mail:
> <mailto:lucene-dev-unsubscribe@jakarta.apache.org>
> For additional commands, e-mail:
> <mailto:lucene-dev-help@jakarta.apache.org>
>
>
We did hold a poll of users on this some months ago and everyone who
responded said 1.2. Shall we hold a committer vote? If we up our JDK
we can use more efficient collections!

-Andy



--
To unsubscribe, e-mail: <mailto:lucene-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:lucene-dev-help@jakarta.apache.org>
RE: What JDK level should lucene code be compatible with? [ In reply to ]
> From: Andrew C. Oliver
>
> Eric Fixler wrote:
>
> > When adding/fixing code in the main source tree?
>
> We did hold a poll of users on this some months ago and everyone who
> responded said 1.2.

Yes, but since then I've seen folks mention that they're using Lucene in JDK
1.1, e.g.:

http://nagoya.apache.org/eyebrowse/ReadMsg?listName=lucene-user@jakarta.apac
he.org&msgId=117018

> Shall we hold a committer vote? If we
> up our JDK we can use more efficient collections!

I wouldn't expect big performance improvements. Someone could experiment by
switching things to use, e.g. HashMap instead of Hashtable, and ArrayList
instead of Vector, but all that would really be saved is some unneeded
synchronization, and synchronization overhead is pretty low in JVMs these
days.

If someone can demonstrate that there is appreciable performance
improvement, that would be a strong reason to remove JDK 1.1 compatibility.

Doug

--
To unsubscribe, e-mail: <mailto:lucene-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:lucene-dev-help@jakarta.apache.org>
Re: What JDK level should lucene code be compatible with? [ In reply to ]
Lets get 1.2 out first :).

----- Original Message -----
From: <cutting@lucene.com>
To: <lucene-dev@jakarta.apache.org>
Sent: Thursday, May 16, 2002 1:51 PM
Subject: RE: What JDK level should lucene code be compatible with?


> > From: Andrew C. Oliver
> >
> > Eric Fixler wrote:
> >
> > > When adding/fixing code in the main source tree?
> >
> > We did hold a poll of users on this some months ago and everyone who
> > responded said 1.2.
>
> Yes, but since then I've seen folks mention that they're using Lucene in
JDK
> 1.1, e.g.:
>
>
http://nagoya.apache.org/eyebrowse/ReadMsg?listName=lucene-user@jakarta.apac
> he.org&msgId=117018
>
> > Shall we hold a committer vote? If we
> > up our JDK we can use more efficient collections!
>
> I wouldn't expect big performance improvements. Someone could experiment
by
> switching things to use, e.g. HashMap instead of Hashtable, and ArrayList
> instead of Vector, but all that would really be saved is some unneeded
> synchronization, and synchronization overhead is pretty low in JVMs these
> days.
>
> If someone can demonstrate that there is appreciable performance
> improvement, that would be a strong reason to remove JDK 1.1
compatibility.
>
> Doug
>
> --
> To unsubscribe, e-mail:
<mailto:lucene-dev-unsubscribe@jakarta.apache.org>
> For additional commands, e-mail:
<mailto:lucene-dev-help@jakarta.apache.org>
>
>


--
To unsubscribe, e-mail: <mailto:lucene-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:lucene-dev-help@jakarta.apache.org>