geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Blevins <david.blev...@gmail.com>
Subject Re: Opening up xbean svn access (Fwd: Jar listing performance)
Date Thu, 09 Jun 2011 06:21:39 GMT

On Jun 8, 2011, at 9:36 PM, David Blevins wrote:

> I and pretty much everyone with dual OpenEJB/Geronimo commit is likely way too busy to
commit this patch (XBEAN-176).

Ended up having to make CDI related change to xbean and cleared out this patch.  We should
still consider opening things up though.

> 
> I wonder in general if we shouldn't just open up the xbean section of the repo since
it is more or less a commons area.  Would be nice to open it up to at least OpenEJB.  I.e.
we just tweak the svn auth file like so:
> 
>  [/geronimo/xbean/]
>  @geronimo = rw
>  @openejb = rw
> 
> Could potentially open it to OpenWebBeans at some point later if it starts using xbean-finder.
 So far talk on the OWB side has always been "lets move everything commons" which I suppose
we could do, but that seems like a lot of pointless refactoring and repackaging and moving
when we could just add a couple lines to the svn auth file.
> 
> Thoughts?
> 
> 
> -David
> 
> Begin forwarded message:
> 
>> From: Romain Manni-Bucau <rmannibucau@gmail.com>
>> Date: June 7, 2011 10:22:22 PM PDT
>> To: dev@openejb.apache.org
>> Subject: Re: Jar listing performance
>> Reply-To: dev@openejb.apache.org
>> 
>> In fact there are 2 kind of API for Jar/Zip: the ZipFile and the
>> ZipInputStream (same for jars). If you use the inpout stream one ... you use
>> an input stream so you read everything. If you use the file one you directly
>> call native methods so under linux in particular it works better (i didn't
>> test under windows but i hope it is the same).
>> 
>> David said me we are using JarArchive:
>> https://issues.apache.org/jira/browse/XBEAN-176
>> 
>> - Romain
>> 
>> 2011/6/8 David Blevins <david.blevins@gmail.com>
>> 
>>> FYI, to Romain, regarding http://www.friendpaste.com/JU7WuKslKle9UaAL3r21M
>>> 
>>> There might be caching going on in the first call that makes the second
>>> call go faster.  We should check that angle.  Maybe switch the order or loop
>>> the main code a couple times.
>>> 
>>> -David
>>> 
>>> 
>>> 
> 


Mime
View raw message