abdera-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From James M Snell <jasn...@gmail.com>
Subject Re: svn commit: r428793 [1/2] - in /incubator/abdera/java/trunk/client: ./ src/ src/main/ src/main/java/ src/main/java/org/ src/main/java/org/apache/ src/main/java/org/apache/abdera/ src/main/java/org/apache/abdera/protocol/ src/main/java/org/apache/
Date Fri, 04 Aug 2006 17:59:36 GMT
Yeah, those are coming. Man you're picky ;-).  I rewrote the client
yesterday following my initial post and thus still have to go through
and rewrite my tests.  For now I've been testing the client against Mark
Nottingham's set of XMLHttpRequest caching tests[1]. The ant build will
be coming shortly.

- James

p.s. don't expect much of the caching impl yet.  I'm pretty well certain
it's busted as is and will need lots of work to get right.  I really
hate the fact that there doesn't appear to be any Java language client
side cache implementations out there.

[1] http://www.mnot.net/javascript/xmlhttprequest/cache.html

Garrett Rooney wrote:
> On 8/4/06, jmsnell@apache.org <jmsnell@apache.org> wrote:
>> Author: jmsnell
>> Date: Fri Aug  4 10:36:34 2006
>> New Revision: 428793
>>
>> URL: http://svn.apache.org/viewvc?rev=428793&view=rev
>> Log:
>> This still needs LOTS of work to complete.  The client itself is
>> rather straightforward and complete, conditional gets,
>> cache-control, gzip/compress/zip compression is supported and enabled
>> by default, extension HTTP methods (e.g. COPY)
>> can be used with nominal effort..
>>
>> The part that really needs the most work is the client side HTTP
>> cache.  It's generally functional but still needs lots of
>> refinements, testing, tweaking... and a
>> generally-better-all-the-way-around implementation.  The goal is to
>> provide a functionally
>> complete client-side private cache that supports most of the nuances
>> of HTTP caching (e.g. Vary headers, max-stale, min-fresh,
>> no-cache and private headers, and so on).  Use of the cache can be
>> controlled via the RequestOptions interface
>> (e.g. options.setNoCache(true) to disable the use of the cache via the
>> Cache-Control header).
>>
>> I had imagined ultimately shipping three cache implementations with
>> this.  An in-memory LRU cache, A write-to-disk cache, and a
>> JCS-backed cache implementation.  Other impls are possible (e.g.
>> derby-based, etc).
>>
>> The client code currently depends 100% on the Apache Commons HTTP
>> Client. Changing the http stack would require fairly significant
>> changes to this code.
> 
> Any chance of getting this hooked into the ant build?
> 
> Also, tests and/or example code would be useful, although I can
> certainly understand how it would be hard to test the client without a
> working server...
> 
> -garrett
> 

Mime
View raw message