pkg/25000: mailing lists archives browsing is complicated on netbsd.org

[dotz%irc.pl@localhost]

>Number:         25000
>Category:       pkg
>Synopsis:       mailing lists archives browsing is complicated on netbsd.org
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    pkg-manager
>State:          open
>Class:          change-request
>Submitter-Id:   net
>Arrival-Date:   Wed Mar 31 20:33:00 UTC 2004
>Closed-Date:
>Last-Modified:
>Originator:     Michal Pasternak
>Release:        N/A
>Organization:
>Environment:
N/A
>Description:
It is really hard to do something with current list archives.

For example, this one post is very interesting for me:
http://mail-index.netbsd.org/tech-pkg/2004/01/19/0028.html

I'd like to view replies for it, but the only way to do this is to go
here:
http://mail-index.netbsd.org/tech-pkg/2004/01/
and look for all posts having similar subject.

Perhaps there were some more replies, not only in January, so I have
to go to February archives by changing the URL (it is impossible to click on 
"February" here - http://mail-index.netbsd.org/tech-pkg/2004/01/).

Okay, nothing in February? So perhaps I should go to March archive... and
so on.

Do I need to see all John Doe posts? Certainly. www.google.com,
query for [site:mail-index.netbsd.org "john doe"]. Okay, so I get
there Jon Doe's posts and Jon Doe in the mail index. If I click
wrong link (the index), I have to browse a long page and search
for all John Does.


>How-To-Repeat:
See above.
>Fix:
Please install some *useful* software, that allows to browse
mailing lists archives in an easy way.

Mainly, I'd like to see there *threaded* viewing of the discussions.
I hope current archives are complete enough (eg. posts are saved
in full form, with all headers) to allow such browsing.

Also, I wonder why searching of lists has to be done via Google. If
we'd include threaded browsing in the archives, doing search on our
own would allow us to maximize searching options.

In case currently available software for threaded browsing is too
slow or wants too much resources, I am ready to write some optimized
threaded viewing software, which could work in distributed environments,
to minimize netbsd.org CPU time.

Good, comfortable archive browsing is IMO a *critical* target in 
open operating system development, as it allows new developers to 
easily check, what has been done and what has been discussed. Current 
mailing list archives don't allow such option; I'd like to see them
changed and I'd like to help with such transition. I know, how to write
code for the web, we could distribute the database / browsing to 
2 - 3 hosts and finally get rid of the current list archiver, as ,
without threading, it is not very useful.

If you see such interest/need, please contact me directly. Also, a 
tarball with 1-month archive for a list would be useful, if we decide
to write software on our own.

BTW it could also include some nice statistics, with nice graphs
(perhaps in some chart-drawing program format, if we want to save
the bandwidth); perhaps some more options.

Comments?

>Release-Note:
>Audit-Trail:
>Unformatted: