[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnu-arch-users] Re: Ruminations on Arch Desiderata

From: Paul Snively
Subject: Re: [Gnu-arch-users] Re: Ruminations on Arch Desiderata
Date: Wed, 17 Sep 2003 07:22:39 -0700

Hash: SHA1

Hi Doran!

On Tuesday, September 16, 2003, at 10:21  PM, Doran Moppert wrote:

How useful, really, is automatic registration of archive names beyond about two dozen anyway? I have 20-odd archives registered and already that's too much to reliably remember without some kind of meta-data (in my case, a text
file and ArchiveRegistry on wiki).

That's an interesting point. One would hope that the name would be significant enough, but I can certainly see a motivation for additional comment-style metadata on the archive name. OTOH, it might be sufficient, albeit an amount of work that would be nice to avoid, just to go through the archives and learn what categories they held, etc. Given arch's fairly uniform naming conventions, it might even be possible to, e.g. determine when the same category existed across multiple archives and present that view of the world. This obviously requires more thought.

Really it sounds to me as though you're barking up the wrong tree: it would make more sense to me to have an HTTPD somewhere (which can register itself through DHCP or various DNS hacks - ask Verisign (
about this) with a well-known URL featuring a table similar to
ArchiveRegistry. On the same box I'd chuck ViewARCH and get it to update its
cache every 15 mintues or so.

Yeah, my example wasn't as well-chosen as it could have been: I was trying to describe an environment without such fixed infrastructure that someone has administrative control over.

If I'm walking into a conference and I want to auto-publish, I don't think this is arch's job but rather that of the network services on my box. As
(Miles?) pointed out, arch doesn't care whether anyone else can see my
archive: why should it?

Yes, that's correct. It seems likely that the "let the local link know that my archive lives at location X" process would be distinct from any of the arch tools.

 When DHCP gives me an address the server can query a
well-known URL on my machine or send me instructions on manually publishing. Heck it could probably be done by creative use of a VCard field. Which seems
closer to what would be desirable at a tech conf in my world ...

Perhaps "pragmatic," but "desirable" seems quite debatable to me. :-)

(disclaimer: I know nothing about rendezvous, zeroconf, DNS-SD or any of these
things. I am however speaking from experience making a place for ad-hoc
networks when I can say "just let the DHCP server set you up, then go to
'www' and join the forum.")

Sure, that's largely how things are done today for lack of a good automated alternative. I'm proposing that we explore how we might take advantage of a good automated alternative.


Best regards,

Version: GnuPG v1.2.3 (Darwin)


reply via email to

[Prev in Thread] Current Thread [Next in Thread]