Subject: Re: New vote on SUAVE or Objective VHDL
From: John Michael Williams (jwill@pacbell.net)
Date: Sat Jun 24 2000 - 11:36:18 PDT
Hi Dale.
Dale E. Martin wrote:
>
> John Michael Williams <jwill@pacbell.net> writes:
>
> > No. I'm not off target: I'm complaining that there is no target, just
> > a lot of code circulating in the butts.
>
> > Nice that you and others MIGHT have provided documents. I guess it is
> > unimportant to you whether they are randomly chosen and conflicting
> > (as are the two or more competing prototypes). No need to decide on
> > anything, just have fun writing and debugging . . . .
>
> Really, you're insulting the entire group here. A lot of research went
> into this effort - user surveys, _extensive_ background research, many
> papers written, meetings held etc before there was ever a line of code.
> This group has been around much longer than the last year.
This always is the assumed overhead in standards development.
I'm pointing out a deficiency in the way "the past four years"
or more has been used. The longer this group has been around with
no standard to show, the less praiseworthy have been its
efforts--in the direction of a standard, which is what this
group is supposed to be doing. I don't have any objection to,
or criticism of, a reflector for exchanging ideas seriously
on OO progrmming, design, or language
implementation. But I don't think it should be done
under the aegis of "standards development".
There was no intention to insult you; however, I thought my original
suggestion was quite clear, so, when you said it was off target
and flying off the handle to question the proposed vote, I tried using a
little hyperbole to make the point more obvious. Presumably, you
would understand. But, sigh, I guess not.
I happen to love programming; so, I assumed that
if we can't have a standard, why not have some fun, was a reasonable
description of why no progress.
>
> > Why is not the group deciding on the documents, and not being
> > sidetracked into beta-testing programs which, you seem to be saying
> > below, are NOT up to commercial standards of functionality.
>
> What is the story - are you trying to start some sort of flamewar here?
>
> I didn't say anything about the quality of anything. I said "to my
> knowledge there is no commercial backing of these systems." I'm more
> confident of that statement about the SUAVE group than the Objective
> VHDL group, which is why it's "to my knowledge".
>
> > I'm an expert at C++ OO coding, and I know the OO-VHDL "iterature"
> > well enough to recognize a slump in progress caused by endless
> > prototyping.
>
> You're flat out wrong. I was there for the creation of SUAVE, and
> you're wrong. No code happened before the language was designed.
Here is the problem: You are confusing language development with
standards development.
>
> > I don't want to read any more random documents or peer documents; I
> > want to see some standards activity.
>
> So, you feel that the group isn't generating enough paper to be a
> standards activity, but you're sick of reading the paper it has
> generated?
"Paper"--not really. Electronic random articles or peer-reviewed
articles, too. They don't encourage agreement, although peers tend
to agree they should be circulated.
>
> > It fulfills the purpose of this group.
>
> You propose to design a language without going through any use case
> analysis, user feedback, peer review?
I don't think peer review, in the academic sense, should be lumped
with standards activity. Peered articles represent individual new
ideas, ruled acceptable in the group of peers. They do not
represent the uniformity of opinion necessary in a standard.
The problem seems to have been IGNORING user feedback in favor of the
preconceived opinions of language developers, who are very concerned,
and rightly so, about whether they are able easily to implement the
standard. Implementation problems should be ignored,
in my opinion. Preexisting code, like a prototype,
should be thrown away BEFORE starting standardization. But, you
yourself keep citing how important has been the work you did
on the software; I again respect this and understand
it, but it doesn't bear on the standards issue.
>
> > If a person can't understand what they are doing without running code
> > examples, then maybe the material is too complex--or maybe the person
> > should be doing examples, not standards!
>
> That's completely flamebait - stop insulting the group.
Flamebait? Exactly who are you seeing as being insulted here?
Which group? Am I the only one worried by incessant prototyping?
I think if this were so, then, as I suggested, the group should cease
calling itself a "standards" related group, but rather a
software development group. Nothing wrong with this.
No insult intended. Different keystrokes for different keyfolks.
>
> > Think of it: For the past year, this group has, in effect, been
> > granted an IEEE charter to come up with essentially an architectural
> > document. Instead, you tell me that when the house is done, you will
> > figure out the floorplan. It's easier on the brain that way.
>
> I didn't say that at all. I said "the documentation that you would like
> to see exists if you ask for it". You seem to think that the entire
> effort of the group should go into writing a new standard in a void?
The problem is that, after all this time, we STILL have two--
or actually probably about six--disjoint documents which "exist" and
apparently will continue to exist if I ask for them. I would like them
to be discarded so as to work up a standard unblemished by the
praiseworthy but nonstandard works of the past. Use the experience;
don't
keep trying to reuse the code.
>
> > That's what you are implying. Are you trying to say OO methods are
> > just for dumb people? Maybe you have a point there, but I think
> > overcoming that particular handicap is possible.
>
> ?! I'm not implying any such thing. If you're trying to start a fight
> with someone, a nice way to start is to insult the whole group, and then
> insult the rest of the world.
>
> > Just drag some well-intentioned people away from the debugger, or at
> > least limit theor hours watching debug, and get them thinking about
> > how OO-VHDL *SHOULD* work, not how a sw engineer has so far been able
> > to find time to make it run.
>
> Honestly, you don't know what you're talking about. That's not the
> process that has happened here in the least.
>
> > Yes: You are missing my motion to provide for a way to stop coding and
> > start thinking. I'd appreciate a second.
>
> You propose to go back to 1996(?) when this effort started? I don't
> second that.
If I knew how to go back to 1996, I would try it, just to see what
happens. However, don't you think it might be possible to profit
from experience without repeating it?
>
> > Again, I move that the proposed vote be withdrawn and replaced by one
> > which provides at least for a choice "neither". If "neither" wins, we
> > drop ALL reflector discussion of either program until we have a draft
> > LRM document (or, maybe, chapter of a VHDL document).
>
> > Off what handle? Do you have a handle on the standards document?
>
> You've come from out of nowhere and made accusations about the
> activities and motives of the group - that's what I'm categorizing as
> "flying off the handle". If you've been very active, at the last
> meetings and so forth (I have not been able to make it to those), then
> strike the "from out of nowhere".
Most of the meeting have been held nowhere near where I could travel,
and I have attended only two in the past few years. I share your
feeling of being left out by those who have the money and time to
spend on personal meetings. Minutes have circulated on the reflector.
>
> > My alternative IS ABOVE: NEITHER PROGREAMMING PROJECT should be
> > further considered until at least a working group has completed work.
> > But, instead of getting it adopted in the back room, I'd like it to be
> > voted.
>
> > It sounds trite, but you are insisting that now that we have the
> > cart before the horse, we push harder until it works.
>
> I didn't insist on anything. I asked you what you felt is more
> appropriate at this point. You seem to be saying, "I missed what was
> done 4 years ago, and now I want you all to start over".
No, I have been participating, but I felt that my input was
being ignored, so I ceased trying the bery imperfect OO VHDL prototype
software. At least I can write a computer program in C++.
>
> I have no objection to you making the proposals you have, but I don't
> second them.
Fair enough. I am sorry if you were insulted. I was trying to
use somewhat vivid language to make a point I was sure would
be understood. I know it is hard sometimes to see a mistake, and
harder when one feels wronged by someone pointing it out.
--
John
jwill@pacbell.net
John Michael Williams
This archive was generated by hypermail 2b28 : Sat Jun 24 2000 - 11:37:25 PDT