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

Re: FC2 ate my system

On Thu, 27 May 2004, James Wilkinson wrote:

Christopher Stone wrote:
How do you know he is using pine?
His message-IDs begin "Pine.". Dead give-away.

He should be using mutt if he is using a console,
Arguably, yes, but IMHO mutt took a lot more configuration before it
worked the way I wanted it to work. (I'm on it now, though...) I must

And there's no reason why he shouldn't be using text mode MUAs in an
xterm, either.

and anyway, if links can click on urls, so should pine.

That's very debateable. Links is a web browser, and pine (and mutt) aren't. Philosophically, an MUA has no business trying to interpret a non-MIME non-UUencode mail body, or a text/plain message part. Which is what it would take: either MUA can shell out to links, but the difficult bit is working out "what's a URL". (Links gets it easy, because HTML identifies a URL unambiguously).

Practically, identifying URLs gets messy. Something of the form
http://localhost/ might be easy, but then "why don't you support
www.example.com? Or http://www.example.com/test.php?";

(It might be interesting to see how many URL-identifying programs
include those question marks as part of the URL.) I maintain that there
is no Right Way to do it, especially when URLs line-wrap. Microsoft
programs can get hopelessly confused. It's a lot of extra complexity
incidental to the business of e-mail.

Also, links is designed around the concept that it's going to show
hypertext, and so the scrolling supports hyperlinks in the text (down
means down to the next hyperlink, which might be scroll down a page,
or down a line). Mutt and pine are designed around the concept that
they're going to be showing plain text, possibly with attachments.
Changing either MUA to show hypertext would be a major change to the way
either program worked, and inconvenience their main purpose of showing
plain text (down normally means scroll down, except occasionally?)

Philosophically, the Unix Way to do it would be to have a separate
find-url program that could be shared between programs (or easily
replaced). This would get you what you want without bulking up links or
the MUA: the MUA would pipe the body-text to find-url, which would
run links with the (user-selected) URL.

Not sure why pine bashing is hapenning - but it works pretty well for me. I've tried to switch to mutt - and stayed back with pine. (roles feature is great!)

The biggest drawback of mutt is - you need a functional
sendmail/postfix on the machine. And if you wish to use multiple smtp
servers (because of multiple mail accounts) - you are hosed. (It could
be argued that one doesn't need this feature but...)

And pine does process urls - I use 'epiphany' as my urlviewer - with
the following entry in ~/.mailcap

text/html; /usr/bin/epiphany %s ; copiousoutput

And I believe urls are properly defined by some RFC - and any url
processor should be able to grab the correct ones - and ignore the
wrong ones.

Mail-clients that wrap lines mess up things - but thats the
mail-client's (MUA) fault.


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