There was an older comment I added to HTTPCLIENT-745 mentioning that a bunch of classes share variables directly instead of using setters/getters. I'll promote that to a full issue on its own. Other than that, there's only one more thing i can think of...
While adding keep-alive support, I noticed that code-changes that need to provide the ClientConnectionManager with more information aren't possible without API changes. In the rest of HttpClient's code, there's usually an HttpContext being passed around that can hold key/value pairs so things can easily be plugged in without requiring full-blown API changes. I can't see any way to do this with ClientConnectionManager currently. Maybe there should be some Context that that uses? Or share the HttpContext itself? This will likely become an issue after an API freeze if there's more data we find ourselves suddenly wanting. Sam On Thu, Jul 10, 2008 at 5:17 PM, Oleg Kalnichevski <[EMAIL PROTECTED]> wrote: > On Thu, 2008-07-10 at 17:02 -0400, Sam Berlin wrote: >> I noticed that the recent fix to multipart is dependent on a snapshot >> of mime4j. If we go BETA1 for HttpClient, does that also freeze the >> API of the mime component? Is that OK if it's dependent on a snapshot >> release of mime4j? >> > > I think it is not. However, chances are good mime4j version 0.4 will be > released soon. In the worst case we could ship HttpClient 4.0-beta1 > without HttpMime > > What's your feeling about the API of HttpClient proper? Is there > anything else worth delaying the release for? > > Oleg > >> Sam >> >> On Thu, Jul 10, 2008 at 4:59 PM, Oleg Kalnichevski <[EMAIL PROTECTED]> wrote: >> > Folks >> > >> > There are no feature requests / open bugs targeted for BETA1 left. How >> > do you feel about the current state of HttpClient's API? It is good >> > enough to move to the BETA phase? Does anybody want to spend more time >> > refining the API? >> > >> > What should be the plan now? >> > >> > Oleg >> > >> > >> > --------------------------------------------------------------------- >> > To unsubscribe, e-mail: [EMAIL PROTECTED] >> > For additional commands, e-mail: [EMAIL PROTECTED] >> > >> > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > >
