On 4/6/23 16:49, Brian E Carpenter wrote:
The reason for cleaving to email is not, IMHO, traditionalism. It's
functional. Email has more flexibility and functionality than
forum-style interaction, including github issue tracking. This isn't a
new discovery; it's been like that since the buzzword was
"computer-supported collaborative work" at least 30 years ago. Email
is simply the survivor; I agree that it's beginning to look as if
github issues will also be a survivor. One reason is that the two are
well integrated - when a github issue that I'm interested in wakes up,
I get an email, and I can even respond by email.
I think the question is not getting away from email, it's how to
integrate email and forum-style interaction even better.
Mostly agree, though I would say that email is better in nearly every
respect than all of the alternatives in widespread use today, and it's
better by design and not merely by accident. And at least to some
degree, email evolved to facilitate "computer-supported collaborative work".
There's still a lot of room for improvement of email, though. Email grew
up before the web, and we really haven't discovered how to integrate
them well. HTML in email is a mess, for example, because HTML is really
not a good representation for real-world text (which isn't strictly
organized in a hierarchy) and an even poorer representation for
annotated text. Back in the 1980s I used to say that email only
worked well for people who had computers on their desks. These days
everyone has a computer that they carry around with them, but email
discussions work a lot better with larger screens and keyboards than it
works with something small enough to fit in one hand. (IMAP's ability
to flag messages helps a lot, if people use it in a consistent way.)
Always-on internet access (which wasn't assumed for email but is almost
universal now) makes interactive mail quite feasible, but
email/HTML/javascript aren't a good fit for this for several reasons.
In many ways I think IETF abandoned email in the late 1990s, and that's
really unfortunate. IETF is in a good position to make email an even
more effective collaboration tool, but recent management hostility to
"bespoke" tools may have discouraged this. Some experimentation has
been done with email user agents (especially webmail), but mostly in a
way that differentiates one company's product from another rather than
in a way that encourages collaboration across different companies' products.
Email enhancements would make good hackathon projects, provided there's
not hostility to actually trying and using those enhancements in IETF.
Keith
