* On 01 Dec 2012, Tony's unattended mail wrote: > > Regardless of which standards a mutt user endorses, a good quality > tool is lenient in what it accepts, handles it well, while being > strict in what it produces.
Yes, that's generally our principle. > Yet mutt is not good at handling common deviations from standards and > conventions. Mutt would improve if mutt developers received more > garbage that could be salvaged into something that's readable. No? I find it's good. I receive a lot of garbage, and I find it's pretty readable by means of salvage. Is there something specific you have in mind? I use mutt, and I have no issues with anyone's lines. If you do, and you're not Chris Bannister (who already did), explain the problem. I don't think I'm sheltered. I work in a Microsoft environment, and a majority of my coworkers have never used UNIX, an 80-column display, Usenet, etc. They write without line breaks, they sign their e-mail in green Comic Sans with institutional logos, and they top-post. I'm only sampling this thread, but I think most of it is academic tussling. I don't see anyone illustrating any actual problems in mutt, aside from Chris's noting that mutt says "(all)" when it hasn't shown all of the message. That's a genuine usability problem, albeit not a major one. > I think this is why mutt's encryption interoperability suffers. Mutt > is good with GPG, but it's lousy with S/MIME. While Outlook does > S/MIME okay, it does not even recognize a GPG message at all, and > Outlook plugins are lousy with GPG. If you can't submit a patch, talk about what needs to be changed, and submit descriptions of a proposed workflow, UI model, etc. I don't use S/MIME, have no use for it, so it's hard for me to extend my thinking to what's wrong with our implementation. Likewise I can't fix any problems in the Swiss German translation, because I don't use Swiss German and can't think like a natural user of Swiss German and I don't have time to learn to use Swiss German and find Swiss German speakers to communicate with just to address a software issue. One of mutt's original core design goals was good OpenPGP support. Our S/MIME support was added in stages by early adopters of S/MIME in a predominantly PGP world, and though I lack data, I suspect our PGP users still far outnumber our S/MIME users. So if our S/MIME support is lousy, that's completely undertandable. We need someone who uses it to deal with it. Abstractly, I would like for mutt to have good S/MIME support, but I can't worry about it enough to start S/MIME-based conversations on a significant scale for the sole purpose of learning for myself what might be wrong with our S/MIME support. At the risk of more gratuitous intellectualization of a simple issue, I'm not keen on most descriptions of free or open or whatever source. I think they miss much of the point. I think of mutt, and most other popular "open source" software, as *community-sourced*: it depends not only on the openness of its codebase, but on the communal origins of its ideas. So if there's a problem, let's commune. "Mutt would improve if mutt developers [were or did whatever]." Turn that around. If you are or do whatever, become a developer. You don't have to write code. -- David Champion • d...@bikeshed.us
