On Wed, Dec 31, 2003 at 01:32:40PM +0100, Enrico Zini wrote:
> On Tue, Dec 30, 2003 at 04:02:07PM +1000, Anthony Towns wrote:
> > Do you find the "Essential: yes" flag unreasonable too?
> No, because it talks about a quality. Something as "the way of testing
> if that profile is valid" is not about a quality, but more like an
> imperative command.
So rename the option, eg to "requires-cable" which is a quality of
the interface. Personally, I don't think that makes any difference to
whether a "yes" makes sense or not -- you can say a package is essential
without having to add a "yes" to it, and there's still no point saying
"requires-cable no". But hey, if it makes you happy, feel free.
> Do you think that in C having the construct "break;" illegal and
> requiring to use "break yes;" or "break please;" instead would be
> reasonable?
It's not relevant -- the syntax for C doesn't require name/argument pairs
there. I would certainly consider "Package: foo\nEssential\nVersion:
1.00-1" to be unaesthetic. Hrm. You might also note that C does require
dummy tokens in some places, eg in:
static int foo(int x) {
return 30;
}
static int bar(void) {
int (*f)(int) = foo;
f(20);
}
ANSI C requires the "x", the 20, and the 30, even though none of them
are actually used.
> For the records (as I have seen people saying that), the main reason I
> kept reopening the bug was not to force you on anything, but because I
> was doubting you were understanding at all why I was asking for that.
Well, you said you were reopening the bug "as I still wish ifupdown
changes attitude towards handling of /e/n/interfaces." As opposed to
saying something like "I have obviously reopened the bug, as I don't
think you've understood why I'm asking for this".
> Instead of getting any clues that you was getting what I was asking, I
> was getting the bug continuously closed in my face. With threatens.
Meanwhile, I'm getting the bug continually reopened in my face, and having
my suggested alternatives called stupid, crazy, nonsensical, and useless.
Why, you'd almost think close/reopen wars were completely
counter-productive.
> I find it reasonable to ask a maintainer to at least understand what a
> wish is about before closing it, and that's why I acted as I did.
Is that why you didn't ask once if I understood your request, nor did
you explain it in a different way, and instead just added demand after
demand ("keep the bug open", "don't complain about duplicate options",
"tag it wontfix and keep it open", "it's your job to convince me, not
vice-versa" - paraphrased in each case)?
> From your alternative proposal I was instead pretty sure we weren't
> talking about the same thing.
Huh? You want guessnot to be configured by a line in an iface stanza in
/etc/network/interfaces that ifupdown ignores. "test-missing-cable yes"
satisfies those requirements. You don't think that's grammatical/pretty. I
disagree, and don't really care that you're not completely satisfied.
People can disagree over something without misunderstanding the issues
involved, you know.
> I tried to explain you, and I'm sorry you
> didn't like the style (and I welcome you to be a non-native English
> speaker and cope with people treating you as a stupid on busy days), but
> the contents are there.
The only thing I can see you explaining in that bug log is that guessnet
wants to use /e/n/interfaces. Which is fine, but, again, as the newcomer
it gets to abide by the existing rules, not make up new ones, especially
when the existing rules are already flexible enough to do what's needed
in a simple fashion.
But the stuff you were explaining works equally well with the alternative
I suggested, and you just said that the alternative made you think I
didn't understand what you were talking about, which seems to indicate
you're either confused now, or you didn't try to explain whatever it was
you thought I was confused about. Is it really unclear why repeatedly
reopening the bug didn't improve matters?
Oh, also:
features apache firewall php proxy home
features ftp-server nfs-exports
is already do-able by saying:
features apache firewall php proxy home \
ftp-server nfs-exports
I suppose I should also note in response to:
> In which I showed how I didn't think the change you proposed
> was right, and I added how instead I planned to act in the direction you
> indicated me. (which plan was again wrong because of the other ifupdown
> enforcement I later reported, and which was pointed to me by other people
> and not by you)
that there were six hours between the message where you said you were
going to try a different way that wouldn't work, and the message where
you discovered that it indeed didn't work. Those six hours seem to have
been between 7pm and 2am local time for me. I'm afraid I'm not going
to apologise for not being around to point out that problem to you,
and I don't really understand why you seem to be offended that I wasn't.
Cheers,
aj
--
Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.
Linux.conf.au 2004 -- Because we can.
http://conf.linux.org.au/ -- Jan 12-17, 2004
Attachment:
pgpXrKCIL_Vss.pgp
Description: PGP signature