Re: more Keep-Alive - possible problem

["Roy T. Fielding" <fi...@avron.ICS.UCI.EDU>]

> This isn't a big problem, I don't think. The patch should be modified so 
> anything under, say, 100 bytes, isn't Keep-Alived, and I think that's a 
> hack we all can live with. Not many documents are under 100 bytes anyhow, 
> but for those who are, I think it's a small enough comprimise to do this 
> so it doesn't get people who are using Netscape on small pages annoyed 
> with Apache.

Like Ben said, it needs to be a flag.  It would also be nice if it
checked the User-Agent as well, but that may be too resource-intensive.
Fortunately, even a tiny greek letter xbm file is at least 100 bytes.
The fix should only affect the hit counters, which is fine by me. :)

.......Roy

Re: more Keep-Alive - possible problem

[Brian Behlendorf <br...@organic.com>]

On Tue, 6 Feb 1996, Roy T. Fielding wrote:
> > This isn't a big problem, I don't think. The patch should be modified so 
> > anything under, say, 100 bytes, isn't Keep-Alived, and I think that's a 
> > hack we all can live with. Not many documents are under 100 bytes anyhow, 
> > but for those who are, I think it's a small enough comprimise to do this 
> > so it doesn't get people who are using Netscape on small pages annoyed 
> > with Apache.
> 
> Like Ben said, it needs to be a flag.  It would also be nice if it
> checked the User-Agent as well, but that may be too resource-intensive.
> Fortunately, even a tiny greek letter xbm file is at least 100 bytes.
> The fix should only affect the hit counters, which is fine by me. :)

Maybe there's a way to send an *extra* chunk of data to keep the 
connection alive?  efficient transfer of small data is the whole point of 
persistant connections!  

	Brian

--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
brian@organic.com  brian@hyperreal.com  http://www.[hyperreal,organic].com/