[Python-ideas] Browser for mailing lists

[Cameron Simpson]

On 31Mar2014 08:37, Nick Coghlan <ncoghlan at gmail.com> wrote:
> On 31 Mar 2014 08:22, "Cameron Simpson" <cs at zip.com.au> wrote:
> > Please tell me Mailman3 still includes the pipermail stuff still.
> 
> Mailman3 includes a pluggable archiver model - that's how HyperKitty can be
> a separate project, specifically aimed at tackling the "seamless web
> gateway" problem.

If it can be run in parallel, fine. If it replaces pipermail, it
looks like a step backwards.

> > That gets me:
> >   - local search of local data
> >     access when offline, privacy, speed, whatever tools or indices I
> prefer
> >
> >   - search tools of my choosing (eg mairix or mu), including the
> >     search facilities of my mail reader
> >
> >   - using my mail reader search lets me read messages in my _preferred_
> form,
> >     as it would for anyone else using their preferrer mail reader
> >
> > Without an archive download, this is all far far less useful.
> 
> The pluggable archiver model in MM3 means HyperKitty itself doesn't need to
> handle that usage model. However, I'm not aware of any current efforts to
> create an MM3 archiver that is closer to a pure pipermail replacement (the
> workflow you describe strikes me as being incredibly unusual these days, as
> it requires either running your own mail server, or only accessing your
> email from one device)

My workflow is unusual only because most people don't think about
it and because the existing MUAs do not make it easy. There _should_
be a "join a mailing list" button with optional archive suck.

Most modern MUAs have a "search my mail" facility. Making the raw
list archive hard to obtain gets in the way of offering the user's
own search field for search.

_Nothing_ in my workflow requires running my own mail server, it
merely requires having local folders; which is entirely reasonable
with an IMAP service, and entirely reasonable with multiple devices.

Using one's own search tools is only unusual because unless you're
a nerd it is hard.

Likewise, sucking down the archives is difficult for end users for
various reasons:
  - even pipemail makes you do that in pieces
  - the mbox archive URLs (often gzipped) don't come down with a
    cool MIME type meaning "UNIX mailbox format"
  - the pipermail archives need some unobfuscation before use as
    pure mbox files
  - most MUAs do not install a platform hook to handle an mbox MIME
    type (if there even is one), thus making an mbox download that
    much harder on the user.

None of these is hard to address. I do it with a tiny shell script,
but that is not for most people.

I still content that any change that removes the simple ability to
fetch the list archive for use by the end user is a huge step
backwards, regardless of the prettiness of whatever replaces it.

Cheers,
-- 
Cameron Simpson <cs at zip.com.au>

Early to bed and early to rise makes a man healthy and wealthy with bags
under his eyes. - Dave Cochran, <cochran at dg-rtp.dg.com>