- From: Gregory J. Woodhouse <gjw@wnetc.com>
- Date: Thu, 27 Feb 1997 11:39:36 -0800 (PST)
- To: -=jack=- <jack@twaxx.twaxx.com>
- cc: Dan Connolly <connolly@w3.org>, Jim Whitehead <ejw@ics.uci.edu>, w3c-dist-auth@w3.org
On Thu, 27 Feb 1997, -=jack=- wrote: > > I've been lurking also, and will now register my two cents on this. > I do believe that some manner of disconnected operation is desirable, > but I've never really seen an email interface to such functionality > that wasn't a blight upon the world of software in general. To get > more to the point, disconnected operation is good, (although I'll > agree it may take significant effort to get that functionality right), > but doing it via email is definitely bad. Anyway, that's my feeling, > > -=jack=- > > (This text composed by voice) > I've also been very quiet since the Xerox PARC meeting and have been reluctant to jump into the fray (so to speak). Basically, I also believe disconnected operation is a necessity, though I believe that it was declared out of scope because of difficulty of the issues involved, and the necessity of meeting specific deadlines. If disconnected operation is out of scope now, I'd like to see it listed as "future work". That being said, I see no theoretical reason why WEBDAV couldn't operate over SMTP (because HTTP is, after all, a request/reply protocol), but agree that most WEBDAV operations are sufficiently complex (involving multiple requests) that practical operation over e-mail with existing protocols seems unlikely. For example, what would be the implications of checking out (and locking) a document for hours, or even days at at time? It seems to me that the main problem here is finding a suitable way to synchronize multiple copies of a webspace when there is no global clock. I also suspect that a high latency environment like this would only be feasible with finely grained locking, and I still hve my doubts about rangle locking (and don't we really want to lock media specific units like HTML elements?) Perhaps if all HTTP proxies involved are able to cache byte ranges AND provide suitable validators in such a way tat an entity is valid whenever it is the union of valid byter ranges, then I think it could work. But I don't see how this is possible. --- gjw@wnetc.com / http://www.wnetc.com/home.html If you're going to reinvent the wheel, at least try to come up with a better one.
Received on Thursday, 27 February 1997 14:43:53 UTC