Newsgroups & Discussions
Reiner Pope
some at address.com
Sun Apr 1 06:26:03 PDT 2007
Alexander Panek wrote:
> Please, step back a bit and think twice about what you're going to post,
> for the sake of productivity and helping each other. There's no point in
> killing each other, verbally, as we're all grown up and shouldn't
> actually let ourselves be led by our animal instincts that much.
It's not right to expect John Reimer and kris to be silent without some
kind of resolution. They have aired grievances with a backing, and it
is understandable that these should not simply be silenced for the
purpose of 'keeping the NG together.' They have said specifically what
about Andrei's online presence they found objectionable, so a general
counter to such arguments does not do them credit; the correct way to
respond to such points is address each of them individually and
specifically. Calling it a dead horse because we do not share these
objections is unfair to them, as I see it.
When I first read the stdio threads, I very much agreed with the JJR's
and kris's point of view: Andrei seemed very pushy, seeming to repeat
the idea, 'Phobos is now really good at IO, so why isn't Tango.' Andrei
raised objection after objection to what Tango did: first speed, then
readln discarding the newline; then speed again; then objecting to the
call-chained code sample; objecting to Cout(a)(b) in general instead of
Cout(a,b); then objecting to C/stdio incompatibility. Taken in context,
it is easy to see this as an attack on Tango:
1. Andrei talks about Phobos improvements.
2. Andrei starts 3 successive threads questioning Tango's
collections, and IO
3. kris and Sean asked Andrei to submit Tango tickets; Andrei didn't.
4. Comparisons between Phobos and Tango ensue; Tango appears to
perform better, so Andrei's objections could be taken as 'excuses' for
Phobos
Could.
This is certainly how I read it at first, so JJR and kris's posts about
Andrei seemed spot on. However, on re-reading the stdio threads, with
Andrei's posts in particular, his responses seem much more
straight-forward: the objections he raises *do* have merit, and he seems
to be trying to help avoid bad design decisions in Tango IO -- a tough
critique, but an altruistic one.
Yes, he does compare Tango with Phobos at times, like in
http://www.digitalmars.com/webnews/newsgroups.php?art_group=digitalmars.D&article_id=51365
but that can be interpreted as him pointing out how a 'bona fide
programmer' would see the choices for D. The comparison is to illustrate
the problem, not to flame Tango.
He even completely admitted a mistake to James Dennet: 'If you did,
fine. I take that part of my argument back.'
So, from the evidence of this alone, Andrei seems not to be following a
secret agenda. However, he has (IMHO) made a few mistakes which make his
D persona appear very aggressive:
1. He, apparently without any evidence, implied that Tango's IO is
probably not 'up to snuff' -- an allegation which seems to be completely
unjustified.
2. He didn't submit tickets for Tango.
I cannot understand why Andrei did #1 -- I will assume this is just a
mistake, and I hope that Andrei acknowledges it as such.
#2 is very important: submitting tickets turns a Tango bash into a
constructive design discussion. However, it is possible to understand
Andrei's hesitance: having already pointed out the problem, couldn't
kris or Sean simply submit the ticket?
In future, I hope that Andrei will do as kris and Sean ask and submit a
ticket, if just to show good will.
I hope that kris and JJR could re-read Andrei's posts from the last week
or so, supposing that Andrei didn't make the mistakes mentioned above.
Hopefully, you will agree that in that light, Andrei seems straight-forward.
Hope that helps,
Reiner
More information about the Digitalmars-d
mailing list