[r6rs-discuss] Implementors' intentions concerning R6RS
cph at csail.mit.edu
Tue Oct 30 23:52:43 EDT 2007
Ryan Newton wrote:
>> In the past, consensus was required, not just a large
> I doubt that previously consensus meant consensus of everyone on the
> internet. It will be a long time waiting for that!
This was consensus of those who showed up at the author's meetings. A
smallish group, but even so, it was hard to achieve.
> Gerry Sussman couldn't care less! ;) You might notice that he didn't
> vote (nor did Olin for that matter). The point is, that he's not
> going to complain about someone else making a new standard, even if
> he hated it. I'm sure he feels that he has much better things to
> do. He'll go on using what he's using (MIT Scheme) and be fine with it.
Wrong, and wrong. Gerry did care, and registered to vote, but flaked at
the last minute and failed to submit. Had he voted, it would have been
"no". He doesn't read this list, so you'll have to ask him directly if
you want his opinion.
> Older guys like Matthias may find this kind of thing "amusing". But
> for me it's just a bit depressing. For whatever reason, Scheme has
> played a big part in my life (15/27 years of it), and so it's worked
> its way into my self image; I'm a Schemer. So I just hate to think
> that this community I'm part if is not only small, but is festooned
> with bickering.
Get used to it. The Scheme community has been fractured and
argumentative since before you learned how to program. It's unlikely to
change now. Nor is this a bad thing; arguing is a sign of a healthy
Granted, this standard has generated more than its share of bad feeling.
I think that's due to some fumbling in the way the committees were set
up and run, which led to a lot of people feeling excluded or
marginalized. There was (and is, apparently) serious lack of open
discussion about overall goals and how to achieve them. The process was
poorly managed and consequently became dominated by strong personalities
rather than a spirit of compromise. And far too much of the process was
handled behind the scenes, although towards the end there was some
significant opening up. The late transparency didn't help very much,
since by then there were already serious problems, and it was much too
late to fix them.
Most of these failures can be attributed to inexperience; I don't think
there was any significant ill intent. But it's very important that the
next iteration be handled better.
As for the MIT/GNU Scheme implementation, neither Taylor nor I will be
implementing R6RS. Perhaps, someday, another contributor will do so,
but it will take some convincing before I'll allow it.
More information about the r6rs-discuss