Re^2: [SUMMARY] issues under contention in debian metadata
Am 11.07.98 schrieb apharris # burrito.onshore.com ...
APH> Ah, now, see, this is excellent. It sounds like maybe you read some
APH> of the URN standards!
Yes, I#ve read some of the documents from the given URL at w3.org.
APH> the retrieval of the file, which the local URN2URL resolver (probably
APH> CGI) can use to redirect a user to where the document is.
As I#ve read the description URN could be something like
(a) a DNS for the WWW
(b) the rewrite module of apache
You have one server where you can add a unique key and the URL. And than
you can use the key as a reference like the ISBN number. Right?
I think that this is no solution for our problems, because we must have a
httpd and/or a running database to resolve the keys.
Again, I suggest to have a unique ID and a URL/local filename.
APH> You're right that this is a true URN scheme, and that it doesn't have
APH> *any* coupling with the locations of files/paths.
Of course it has a coupling in the urn2url resolver.
APH> I think we should offer both a 'file:' scheme and a 'urn:debiandoc:'
APH> scheme. Maintainers who care can register URNs and maintain this
APH> registration. Now the registration scheme and the resolver scheme
APH> would have to be worked out. I think we'd need a local resovler and a
APH> central resolver. Furthermore, I don't think URN2URL information
APH> should necessarily be transmitted to the local resolver by way of the
APH> docreg file. However, you're right, we could do it that way. It would
APH> take some more thought...
I don#t think that we should have several solutions for one problem. The
ID in my proposal is for local use only. As file: name you could use URL,
local filenames or WWW URNs (at the moment there#s no standard?).
<package name> <free>
Identifier: doc-linux-html DistributionHOWTO
Identifier: DebianHomepage BugReport
Identifier: DebianHomepage FTP
It think that this is a very clean and good solution.
APH> Indeed. But I still think we do need an implicit, easy to use
APH> relative URL scheme. Both you and I have workable schemes for this.
Well, I think that#s very easy to understand the scheme I#ve descriped in
this mail. The URL field behaves like the URL in the HTML standard. You
can have relative and absolute URLs.
Uni: Budde@tu-harburg.de Fido: 2:240/5202.15
Mailbox: firstname.lastname@example.org http://www.tu-harburg.de/~semb2204/
To UNSUBSCRIBE, email to email@example.com
with a subject of "unsubscribe". Trouble? Contact firstname.lastname@example.org