Henrik, I find the new archive highly inconvenient for reviewing a thread or a chronological sequence on a given list. This can be done at glance with the old archives with one line per message. With the new one, it's matter of figuring what filter to set up. I have found it *much* harder to find a particular thread than with the old archive. Please consider providing a subject-threaded and a topic-threaded view of a single list into the new archive. We definitely need a view equivalent to this one for every list: http://www.ietf.org/mail-archive/web/apps-discuss/current/maillist.html A generalised search tool does not provide the same convenience. Also it would be handy to have a way to download a raw archive. Regards Brian On 09/03/2014 03:44, Henrik Levkowetz wrote: > Hi Björn, Dale, > > On 2014-03-07 23:02 Bjoern Hoehrmann said the following: >> * Dale R. Worley wrote: >>> I've noticed that recently messages to the Ietf mailing list have been >>> sprouting "Archived-At" headers. This seems to be a great >>> convenience. But when I attempted to use one of these links, >>> http://mailarchive.ietf.org/arch/msg/ietf/rvA1yLD7Ds2BAxWYv3iN1ONAxJs >>> it turned out not to be what I expected -- it contained the message, >>> but none of the threading that I've learned to love in IETF mail >>> archives. > > Right. However, updating the single-message view is on the to-do list. > >>> Which is all rather odd, since there *is* a URL that goes to the >>> "real" archive, and does give access to the threading: >>> http://www.ietf.org/mail-archive/web/ietf/current/msg86465.html >>> Why is that URL not put into the Archived-At header? > > 2 reasons: > > * That URL goes to the old archives, not to the new archives > at mailarchive.ietf.org. We're aiming at phasing out the old > archives once it's been established that the new service is > solid. > > * That url will break if for some reason the archive pages have to > be re-generated due to later inclusion of list messages which didn't > originally make it into the archive for one reason or another. It's > happened a few times, and the way MHonArc generates the web pages > guarantees breakage if a dropped message has to be added. This is not > the case with the current Archived-At headers, which use a hash generated > by hashing Message-ID and list acronym, and is independent of any particular > sequence or ordering of messages in the list archive. > >>> Though to put the shoe on the other foot, I notice that the *former* >>> archive will show me the full headers of the message if I want, but >>> the latter archive won't. > > Yes. The latter (the old MHonArc archive) only includes some selected > headers in what's made available in the web page; one of many deficiencies > the new archive software aim to fix. > >> I assume there are some technical reasons, Henrik Levkowetz (copied) >> might be able to comment. > > I hope the information above helps. > > > Best regards, > > Henrik > >