- From: Clemm, Geoff <gclemm@rational.com>
- Date: Thu, 24 Oct 2002 12:23:43 -0400
- To: "'WebDAV (E-mail)'" <w3c-dist-auth@w3.org>
- Message-ID: <E4F2D33B98DF7E4880884B9F0E6FDEE2B2901E@SUS-MA1IT01>
Just to clarify, the proposal I was supporting was to replace bytes with octets, not to leave the units unspecified. I don't think anyone wants to leave the units unspecified. There is a separate question of whether the units should to appear in the property name ... I'd probably leave it off, for uniformity with the rest of HTTP (it is the "Length" header, not the "Octet-Length" header). Cheers, Geoff -----Original Message----- From: Elias [mailto:elias@cse.ucsc.edu] Sent: Thursday, October 24, 2002 11:11 AM To: 'WebDAV (E-mail)' Subject: Re: FW: I-D ACTION:draft-dusseault-dav-quota-01.txt Hello, I feel pretty strongly that the units should be defined as bytes. This is a widely accepted standard unit of measure when talking about disk space used/available. If the unit is left undefined, the vast majority of servers would use some denomination of bytes as their unit of measure anyway, whether that be individual bytes, Kb, Mb, etc. Allowing this flexibility doesn't buy us much, just more unit conversions. Furthermore, writing a generic client to support multiple units (on different server implementations) would rapidly become an exercise in finding out what unit the server was using. This would, in turn, require either another round trip or, more likely, the introduction of a new header value, thus unnecessarily raising the complexity of the world... Elias Lisa Dusseault wrote: > [...] For the suggestion not to use 'bytes': I'm ok with 'octets' (is there a > difference important enough to make the change? ) but I reject the > suggestion that it should be an unmeasured unit. It's useless to know > that the quota is 33 frobnitzes and 18 frobnitzes have been used up if > you want to have some hint if you can upload a 1.5 Mb file.
Received on Thursday, 24 October 2002 12:24:15 UTC