[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Line numbers notation.
[Strictly speaking, this is not neccesarily an issue we need to address in the spec, but for the interest of being pedantic, I'm going to continue this thread] Peter- I was using the # as used in HTTP URLs -- I think this is somewhat unsettled, but generally the #'s interpreted outside the of the URL resolution itself - in other words its purely for use by the client. Thus, it depends on the media type involved. #53 was a made-up way of saying "line 53". But it seems to me to be consistent with the use of # in other URI types. If you use [:0.5#53], that actually seems (according to the current proposal, modified to use ) to be incorrect syntax. ':' means that everything following is a number dot-path. I think of the # to be outside any syntax we've yet defined. (Think of it as "lowest precedence operator"). So [:0.5] means version 0.5, and [:0.5]#53 means "part 53 of version 0.5". I think this is one of those cases where obviousness is in the eye of the beholder ;-) I think my proposal is more regular w/r/t to the entire XRI (and even URI) syntax - maybe a little less intuitive when used separate from a XRI/URI though. -Gabe > -----Original Message----- > From: Peter C Davis [mailto:email@example.com] > Sent: Monday, April 21, 2003 12:59 PM > To: OASIS XRI > Subject: Re: [xri] Version enclosures > > > Wachob, Gabe wrote: > > > The only question I have is about the status of the > '['/']' characters vs. '('/')' characters - are there > situations where  must be escaped? (I don't think so). > > > > > No more so than "(" and ")" AFAICT > > >P.S. > > > > I really dig your approach wrt referring to sections of text: > > > >[:version#linenum] > > > > But perhaps we use (in the approach of "eat your own dogfood"): > > > >xri://(urn:oasis:xri)/documents/syntax-and-resolution[0.5]#53 > > > The question for me in this would be: Does [0.5]#53 refer to > line 53 of > document instance version 0.5 (obvious with the [0.5#53] notation)? > Less obvious, but intended with the [0.5]#53 notation. (can > you tell i > spend a good deal of time reading meaning into URI's :-) > > -- peterd > >