You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hc.apache.org by Jeffrey Dever <js...@sympatico.ca> on 2003/02/17 18:07:45 UTC
2.0 alpha3 - no rush
We were going to do a 2.0 alpha3 release, but this will wait untill the
threading work Mike is doing with Aurelien columnates into a conclusion
and a patch.
Jandalf.
Re: 2.0 alpha3 - no rush
Posted by Michael Becke <be...@u.washington.edu>.
The version that Aurelien was using is attached in bugzilla:
http://nagoya.apache.org/bugzilla/showattachment.cgi?attach_id=4878
Mike
On Monday, February 17, 2003, at 12:33 PM, Oleg Kalnichevski wrote:
> Mike
>
> Do you think it would be possible to post even a very rough,
> preliminary
> patch? Seeing how the patch is shaping up would aid us greatly in
> assessing various alternatives.
>
> Cheers
>
> Oleg
>
> On Mon, 2003-02-17 at 18:21, Michael Becke wrote:
>> It seems that the connection wrapper solution will work. I will go
>> ahead and and clean up the code some and submit a patch tonight or
>> tomorrow. If anyone has suggestions/ideas about this change please
>> let
>> me know. In particular I would like to hear some comments regarding
>> the
>> following:
>>
>> - what is a good name for this connection wrapping class, currently it
>> is HttpConnectionAdapter?
>> - should this class be available outside of the
>> MultiThreadedHttpConnectionManager?
>> - once the wrapped connection has been released what should be the
>> default behaviour for the connection methods? I was thinking any
>> methods that perform I/O should throw an exception. Other methods
>> would
>> just return default/empty values or do nothing.
>>
>> Mike
>>
>> Jeffrey Dever wrote:
>>> We were going to do a 2.0 alpha3 release, but this will wait untill
>>> the
>>> threading work Mike is doing with Aurelien columnates into a
>>> conclusion
>>> and a patch.
>>>
>>> Jandalf.
>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail:
>>> commons-httpclient-dev-unsubscribe@jakarta.apache.org
>>> For additional commands, e-mail:
>>> commons-httpclient-dev-help@jakarta.apache.org
>>>
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail:
>> commons-httpclient-dev-unsubscribe@jakarta.apache.org
>> For additional commands, e-mail:
>> commons-httpclient-dev-help@jakarta.apache.org
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> commons-httpclient-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail:
> commons-httpclient-dev-help@jakarta.apache.org
>
Re: 2.0 alpha3 - no rush
Posted by Oleg Kalnichevski <o....@dplanet.ch>.
Mike
Do you think it would be possible to post even a very rough, preliminary
patch? Seeing how the patch is shaping up would aid us greatly in
assessing various alternatives.
Cheers
Oleg
On Mon, 2003-02-17 at 18:21, Michael Becke wrote:
> It seems that the connection wrapper solution will work. I will go
> ahead and and clean up the code some and submit a patch tonight or
> tomorrow. If anyone has suggestions/ideas about this change please let
> me know. In particular I would like to hear some comments regarding the
> following:
>
> - what is a good name for this connection wrapping class, currently it
> is HttpConnectionAdapter?
> - should this class be available outside of the
> MultiThreadedHttpConnectionManager?
> - once the wrapped connection has been released what should be the
> default behaviour for the connection methods? I was thinking any
> methods that perform I/O should throw an exception. Other methods would
> just return default/empty values or do nothing.
>
> Mike
>
> Jeffrey Dever wrote:
> > We were going to do a 2.0 alpha3 release, but this will wait untill the
> > threading work Mike is doing with Aurelien columnates into a conclusion
> > and a patch.
> >
> > Jandalf.
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail:
> > commons-httpclient-dev-unsubscribe@jakarta.apache.org
> > For additional commands, e-mail:
> > commons-httpclient-dev-help@jakarta.apache.org
> >
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: commons-httpclient-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: commons-httpclient-dev-help@jakarta.apache.org
>
Re: 2.0 alpha3 - no rush
Posted by Jeffrey Dever <js...@sympatico.ca>.
Thanks Mike. You rock.
>
> - what is a good name for this connection wrapping class, currently it
> is HttpConnectionAdapter?
Sounds reasonable.
>
> - should this class be available outside of the
> MultiThreadedHttpConnectionManager?
Prefer to have as much isolation as is reasonable. Particularly keep
things out of the public interface if possible.
>
> - once the wrapped connection has been released what should be the
> default behaviour for the connection methods? I was thinking any
> methods that perform I/O should throw an exception. Other methods
> would just return default/empty values or do nothing.
This would seem to be an exceptional case that the programmer needs to
be warned about. I don't like excessive exceptions, but it seems
justified here.
Re: 2.0 alpha3 - no rush
Posted by Ortwin Glück <or...@nose.ch>.
Michael Becke wrote:
> It seems that the connection wrapper solution will work. I will go
> ahead and and clean up the code some and submit a patch tonight or
> tomorrow. If anyone has suggestions/ideas about this change please let
> me know. In particular I would like to hear some comments regarding the
> following:
>
> - what is a good name for this connection wrapping class, currently it
> is HttpConnectionAdapter?
PooledHttpConnection
> - should this class be available outside of the
> MultiThreadedHttpConnectionManager?
No I think package access should be enough.
> - once the wrapped connection has been released what should be the
> default behaviour for the connection methods? I was thinking any
> methods that perform I/O should throw an exception. Other methods would
> just return default/empty values or do nothing.
NullPointerExceptions are fine too, if they occur due to improper use.
So throwing away the intrinsic reference to the wrapped object will do.
Re: 2.0 alpha3 - no rush
Posted by Michael Becke <be...@u.washington.edu>.
It seems that the connection wrapper solution will work. I will go
ahead and and clean up the code some and submit a patch tonight or
tomorrow. If anyone has suggestions/ideas about this change please let
me know. In particular I would like to hear some comments regarding the
following:
- what is a good name for this connection wrapping class, currently it
is HttpConnectionAdapter?
- should this class be available outside of the
MultiThreadedHttpConnectionManager?
- once the wrapped connection has been released what should be the
default behaviour for the connection methods? I was thinking any
methods that perform I/O should throw an exception. Other methods would
just return default/empty values or do nothing.
Mike
Jeffrey Dever wrote:
> We were going to do a 2.0 alpha3 release, but this will wait untill the
> threading work Mike is doing with Aurelien columnates into a conclusion
> and a patch.
>
> Jandalf.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> commons-httpclient-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail:
> commons-httpclient-dev-help@jakarta.apache.org
>