Re: about coders, docs and users
On Tue, Aug 29, 2006 at 10:38:06AM EDT, Rado S wrote:
> =- cga2000 wrote on Mon 28.Aug'06 at 21:59:06 -0400 -=
[..]
> > Even if they are new to mutt those that have mastered this
> > "culture" will instinctively make the best of mutt's
> > capabilities in a relatively short time and will be content with
> > the current "manual".
>
> But even those would be even more efficient with better docs.
> Why waste time on something everybody else has wasted his already?
That is also my opinion.
Another danger of the "experimental approach" -- AKA reinventing the
wheel .. is that well .. we're not really created equal and some will
eventually end up with a less than optimal implementation of mutt and
friends .. or even a right mess that turns out to be so inconvenient ..
unreliable .. generally unsatisfactory .. etc. that they will just give
up and start looking into ways to switch back to a GUI mailer.
> > One possible failing of the current manual where outsiders are
> > concerned is that it was written by UNIX guys for UNIX guys.
>
> My talking exactly!
>
> > > There are far too many variations to document all of them.
> >
> > Agreed. But it would be a lot more profitable where I am concerned
> > to adopt one of the ur-users' ways of doing things than spend
> > countless hours reinventing the wheel.
>
> Righto, see above. But which ur-user should do it?
?
After only a few months on this list it's pretty obvious that there are
a handful of contributors who seem to fulfill most of the requirements.
In my mind .. a strong *nix background .. in-depth knowledge of e-mail ..
I don't think you can document a mailer without a solid understanding of
the fundamentals .. long-time experience of the implementation .. you
need to know mutt very well indeed .. but also the ability to
communicate with the general population .. which means both the ability to
understand and to be understood by the common man .. the ability and the
desire to write, naturally .. and being a native speaker of English is a
big plus.
As always, availability and motivation are likely a problem ..
> Sure, there are some "basic" functions that are generally useful
> (which I tried to {get} record{ed} on the wiki), but some/ many apply
> only for special cases/ conditions, and on the other side don't cover
> what people actually need (first). So we need to find those "basic
> functions".
>
> > > See MuttGuide on wiki, add YOUR ideas to UserGuide. See also my
> > > agenda on RadoS: the MuttFaq shows what people miss and therefore
> > > needs better coverage in a guide.
> >
> > At this point my "ideas" are so vague that I don't see the point.
>
> *BZZZZZTTTTT* wrong! The 1st step is always the hardest! You don't
> have to come up with a perfectly outlined & complete plan! Just
> begin, and let things evolve: this is the wiki way! When things reach
> a critical mass, it will move on its own, others will jump in, because
> then they feel it's little enough that they can afford to add/ fix it.
> ;)
And it turns out the MuttGuide is pretty much the approach that I had in
mind. Although I switched to mutt only 3-4 months ago I just did not
find the MuttGuide/Wiki at that time. Possibly because I had looked into
switching in the past when this framework did not exist and therefore was
not looking for it when I did switch ..
I realize the project is in its early stages but all the same, at some
point in the future, the www.mutt.org page should have a reference to
this ambitious documentation project somewhere right near the top .. A
short paragraph under "latest news" perhaps would both ensure that
potential new users would not miss it and make a statement to the effect
that a new documentation project deserves the same recognition as new
releases .. major patches .. etc.
> > Incidentally, there's an "interesting" thread in "debian-user" at
> > the moment regarding mailers in general and mutt in particular. I'm
> > rather surprised to find that some of the more aggressive flamers
> > seem to have tried mutt .. used it for some time .. and turned
> > against it .. and yet are not even aware of some of its basic
> > capabilities.
>
> ... I didn't know about this thread, I learned 1st + 2nd hand how
> things are (wrong) and could be better.
... ran into a breed I never knew existed .. the mutt hater.. fancy
that ..
:-)
A good deal of criticism focused on mutt not providing MRA/MTA
functionality .. as a consequence mutt suffers from a fundemental design
flaw that makes it near impossible for Joe User to configure .. and
therefore to use .. !!
I am neither an email expert nor a *nix wizard and yet I personally did
not run into serious difficulties setting up a traditional *nix mail
environment on my debian laptop. OTOH, it is probably fair to say that
it is a bit more work than entering your username .. your ISP's "mail
server" .. and a couple of other things in the habitual GUI wizard.
So maybe the Mutt Guide needs to look beyond mutt at the typical
full-blown mail setup in an introductory chapter explaining how it's
done and more importantly why it is worth going to a bit more trouble
rather than emulating the Microsoft Outlook model.
Naturally, it is possible to do basic testing by sending mail to
yourself. But this is very limited and sooner rather than later you
need to switch to your live mail with the risk of losing important
messages or inconveniencing your habitual correspondents. Sandboxing
these activities effectively is not really an option especially for
users with little or no experience.
Another criticism centered on mutt being "obsolete technology" .. ie.
since it's not a GUI it must be considerably less powerful than
state of the art software such as TB .. Evolution .. Kmail .. etc.
Since I switched primarily because I realized I was unable to keep up
with the daily volume of my mail in these environments I totally
disagree.
Maybe something in the Mutt Guide .. or some other place on the web page
should demonstrate that due to its remarkable flexibility (scriptablity
is probably a misnomer but that's what it amounts to..) and its
well-designed keyboard-only interface, mutt is an amazing productivity
tool for anyone who receives more than a dozen of messages a day.
As long as you invest a few hours in learning the interface, it just
lets you get the job done a lot faster and with minimal overhead on
your own personal CPU than any GUI.
Heck .. having decided that gmane was rather inconvenient, I remember
that on several occasions I subscribed to high-volume mailing lists ..
posted .. and unsubscribed immediately afterwards because if for any
reason I had been unable to check my mail for a couple of days ..
I knew that cleaning up my "inbox" was going to be a nightmare.
Not so with my current setup with messages from mailing lists showing
up each in a dedicated folder .. created automatically as soon as I
subscribe to a new ML.
So maybe it is time for a little bit of advocacy .. showing mutt in
action .. with plenty of screenshots just to show new users that with
256 colors to choose from mutt can also look rather kewl ...
> However, some people might argue "they're too stupid for mutt" ...
> Well, if this is mutters' attitude... I applaud you for your
> confidence in your superiority and justification not to improve. :-(
I can only judge by the contributors to this list but there are also a
few knowledgeable users here who are always extremely helpful and ready
to share their knowledge with the neophyte.
No matter how trivial the issue.
Thanks
cga