- From: David W. Morris <dwm@xpasc.com>
- Date: Mon, 2 Jun 1997 17:00:20 -0700 (PDT)
- To: Josh Cohen <josh@netscape.com>
- Cc: Albert Lunde <albert-lunde@nwu.edu>, rlgray@raleigh.ibm.com, http-wg@cuckoo.hpl.hp.com
On Mon, 2 Jun 1997, Josh Cohen wrote: > > > Albert-Lunde@nwu.edu > > Non-caching proxies may be used to access thru firewalls, or as we > > are doing at our institution, for inward access to domain-limited > > services. In our case, we are using a netscape proxy server > > with caching off, but I think the firewall tool kit has > > a simpler non-caching proxy, and there is definely an > > application "niche" for this sort of thing. > > > This is true. I would definitely agree that the non-caching > proxy is common, and has different uses, as well as different rules > than a caching proxy. > > So, I think it would be worthwhile to list them differently. I agree that there a whole class of web facility augmentations which can be accomplished with a non-caching proxy. ZooWorks Research (from HitachiSoft) is an example of such an application. BUT I'm not convinced that is is necessary to list them differently, at least with a dedicated column. That is probably a judgement call as the Caching proxy column is created ... Perhaps add a code which says this applies to both kinds of proxies or just a caching proxy ... Working from memory, there aren't too many non-caching proxy considerations which don't apply to the client and server as well. Persistent connections, pipelining, and hop-hop headers so to keep the basic information more readible, perhaps a second short list of the 'simple proxy' considerations. Dave Morris
Received on Monday, 2 June 1997 17:06:00 UTC