freebsd-chat-digest V5 #699
owner-freebs...@freebsd.org
freebsd-chat-digest Monday, February 10 2003 Volume 05 : Number 699
In this issue:
Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Re: languages
Re: languages
Re: languages
Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Re: languages
Re: languages
Re: languages
Re: languages
Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Re: Okay, I think I need some serious introduction ;-)
Communism
Re: languages
Re: Communism
Re: Okay, I think I need some serious introduction ;-)
Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Re: Communism
Re: Communism
Re: Communism
Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Re: matthew dillon
----------------------------------------------------------------------
Date: Sun, 9 Feb 2003 13:56:18 -0500
From: Rahul Siddharthan <rs...@online.fr>
Subject: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Colin Percival wrote:
> Continuing on with this process, what's the next step? Specifically,
> after finding a bug, fixing it, submitting a PR with included patches to
> -CURRENT and -STABLE, and watching it sit in GNATS for 8 weeks, is there
> anything to do other than keep on waiting?
GNATS does seem to be in a logjam, and will clearly get worse as the
userbase increases. Have the FreeBSD project thought about bugzilla?
Both GNOME and KDE have switched to it recently, finding their old bug
system (debbugs, I think) didn't scale. In fact it seems to have very
useful features, like handling of duplicates, dependency tracking of
bugs, etc, which would (I presume) reduce the logjam quite a bit. The
couple of times I submitted Mozilla bugs, I was impressed by the rapid
response (in one case it was a duplicate, and flagged as that within
minutes of my sending it).
R
------------------------------
Date: Sun, 9 Feb 2003 20:12:03 +0100
From: Erik Trulsson <ertr...@student.uu.se>
Subject: Re: languages
On Sun, Feb 09, 2003 at 01:46:58PM -0500, Rahul Siddharthan wrote:
> Mark Murray wrote:
> > Dag-Erling Smorgrav writes:
> > > All right, show me where in the XML 1.0 specification the
> > > interpretation of the following snippet of XML described:
>=20
> Presumably in some DTD somewhere? That's possible with XML (hence
> "extensible")
>=20
> > Where in a dictionary is the meaning of Jabberwocky explained?
> >
> > 'Twas brillig. and the slithy toves did gimble on the gyre....
> >
> > Language is a structure, not necessarily a meaning.
>=20
> Well, it's explained later in "Through the looking glass", and some
> of it even got into the dictionary later ("chortle", "galumph").
>=20
> The newspeak words in Burgess's "A clockwork orange" aren't in fact
> explained anywhere -- the reader understands them by context. But
> they still have a meaning. Ditto with some of Edward Lear's nonsense.
>=20
> I think language is a structure *and* a meaning, but the meaning
> doesn't necessarily come from an authoritative dictionary (though the
> Academie Fran=E7aise may disagree)
The definition of "language" in mathematics (which is also used in
computer science) is as follows:
An alphabet L is a finite non-empty set of symbols.
Let L* be the set of all strings of elements in L (including the empty stri=
ng.)
A _language_ over L is a subset of L*.
Note that this is a very broad definition and does not concern itself
with any meaning of a language.
For example does "All strings containing exactly 2 instances of the
letter 'a'" define a language over the alphabet {a,b,d,@,2,k}. Not a
very useful or interesting language, but a language anyway.
So, yes, HTML and XML are languages. They might not be programming
languages but they are certainly languages.
- --=20
<Insert your favourite quote here.>
Erik Trulsson
ertr...@student.uu.se
------------------------------
Date: Sun, 09 Feb 2003 14:12:51 -0500
From: Bill Moran <wmo...@potentialtech.com>
Subject: Re: languages
David Schultz wrote:
> Thus spake Gary W. Swearingen <sw...@attbi.com>:
>
>>Bill Moran <wmo...@potentialtech.com> writes:
>>
>>>OK, I'll give you XML, but despite the name, I don't really consider
>>>HTML a "language".
>>
>>XML and HTML are both languages in which you may tell the computer what
>>to do.
>
> Sure. Just don't categorize them as such on your resume. ;-)
Are you crazy? I've actually had interviewers ask me if I knew HTML and
when I told them "yes", they asked why it wasn't on the resume. "It's
under 'other' at the end," I'd say. "Why isn't it under programming
languages?" they asked.
I don't really consider HTML to be nearly as impressive as ASP, PHP, C,
C++, perl or any of the other languages that fall into that category,
but I'd still put it in the "computer languages" section of my resume.
- --
Bill Moran
Potential Technologies
http://www.potentialtech.com
------------------------------
Date: Sun, 9 Feb 2003 14:23:09 -0500
From: Rahul Siddharthan <rs...@online.fr>
Subject: Re: languages
Erik Trulsson wrote:
> The definition of "language" in mathematics (which is also used in
> computer science) is as follows:
>
> An alphabet L is a finite non-empty set of symbols.
> Let L* be the set of all strings of elements in L (including the empty string.)
> A _language_ over L is a subset of L*.
>
> Note that this is a very broad definition and does not concern itself
> with any meaning of a language.
Well, it arguably defines a vocabulary.
What about syntax, grammar? Aren't those part of a language?
R
------------------------------
Date: Sun, 9 Feb 2003 14:32:50 -0500
From: Rahul Siddharthan <rs...@online.fr>
Subject: Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
The Hermit Hacker said on Feb 9, 2003 at 15:23:52:
> For example, take a look at the oldest Critical ticket ... its from '98
> *and* deals with v3.x ... chances are the user has long since moved on to
> newer hardware (deals with the Cyrix CPU) or moved away from 3.x to the
> newer versions ...
Well, if the bug still exists, it's possible there are related later
bug reports, and if it doesn't exist, it was possible it was fixed as
followup to a different bug report. Either way, dependency tracking
would help: if a later bug was classified as dependent on an earlier
one, then if one was closed by a developer, the other would be also
(without explicit action from the developer).
Even if the developer forgot to close the bug, the whole "tree" of
dependencies still exists, so when someone wakes up and starts to
close these reports, the entire tree can be killed off in one stroke.
Plus, duplicate bug handling will reduce the number of open bugs.
Having a list of "most frequently reported bugs" and "most recently
duplicated bugs" on the webpage (as bugzilla.mozilla.org does) would
also potentially reduce the number of bug reports quite a bit.
As I understand, all this is essentially impossible to do in GNATS,
and things are going to get worse.
R
------------------------------
Date: Sun, 9 Feb 2003 15:39:01 -0400 (AST)
From: "Marc G. Fournier" <scr...@hub.org>
Subject: Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
On Sun, 9 Feb 2003, Rahul Siddharthan wrote:
> As I understand, all this is essentially impossible to do in GNATS, and
> things are going to get worse.
'K, one thing that I don't recall when we investigated Bugzilla way back
when ... GNaTs has a 'send-pr' functionality that is installed by default,
so that submitting bug reports is integrated as part of each release?
And, I think, the biggest nightmare ... how do you switch over from GNaTs
- -> Bugzilla without losing a *very* large portion of your user base?
You'd need some sort of 'GNaTs Report -> Bugzilla Entry' gateway so that
users of older versions would still have acess ...
------------------------------
Date: Sun, 9 Feb 2003 20:56:51 +0100
From: Erik Trulsson <ertr...@student.uu.se>
Subject: Re: languages
On Sun, Feb 09, 2003 at 02:23:09PM -0500, Rahul Siddharthan wrote:
> Erik Trulsson wrote:
> > The definition of "language" in mathematics (which is also used in
> > computer science) is as follows:
> >
> > An alphabet L is a finite non-empty set of symbols.
> > Let L* be the set of all strings of elements in L (including the empty string.)
> > A _language_ over L is a subset of L*.
> >
> > Note that this is a very broad definition and does not concern itself
> > with any meaning of a language.
>
> Well, it arguably defines a vocabulary.
> What about syntax, grammar? Aren't those part of a language?
That is included in the above defintion.
Note that "symbols" can include things like space and newline.
This means that for example the following
int main(void)
{
return 0;
}
is a string over the set of ASCII symbols. It happens to be a valid C
program while the following
aad sdfsd &734 11
s
s
dfsdf
43534
which is also a string over the set of ASCII symbols is not a valid C
program. Both are obviously elements of ASCII* but only the first is an
element of the subset of ASCII* which is the set of all valid C
programs.
To define a particular language over some alphabet you need to give
some kind of description on which strings are part of the language and
which are not. I.e. you need to define which subset of L* you are
talking about, since each subset defines a separate language.
A grammar is a very useful way of doing that for a large class of
languages. (Actually I believe that all languages in the above sense
can be described by a grammar, but for some languages (e.g. English) it
can be quite complicated (and in many cases completely infeasible) to
write a correct grammar that fully describes the language.)
- --
<Insert your favourite quote here.>
Erik Trulsson
ertr...@student.uu.se
------------------------------
Date: Sun, 09 Feb 2003 20:58:43 +0100
From: Dag-Erling Smorgrav <d...@ofug.org>
Subject: Re: languages
Rahul Siddharthan <rs...@online.fr> writes:
> Dag-Erling Smorgrav writes:
> > All right, show me where in the XML 1.0 specification the
> > interpretation of the following snippet of XML described:
> Presumably in some DTD somewhere? That's possible with XML (hence
> "extensible")
DTDs specify a grammar, not an interpretation.
The ISO C standard specifies the syntax, grammar and semantics of C.
The XML specification however only specifies a syntax, and a mechanism
for describing a grammar using that syntax. It says nothing about
semantics.
> The newspeak words in Burgess's "A clockwork orange" aren't in fact
> explained anywhere
Don't you mean George Orwell's _1984_?
DES
- --
Dag-Erling Smorgrav - d...@ofug.org
------------------------------
Date: Sun, 9 Feb 2003 15:04:58 -0500
From: Rahul Siddharthan <rs...@online.fr>
Subject: Re: languages
Dag-Erling Smorgrav said on Feb 9, 2003 at 20:58:43:
> > The newspeak words in Burgess's "A clockwork orange" aren't in fact
> > explained anywhere
>
> Don't you mean George Orwell's _1984_?
No, I meant Burgess, but I didn't mean newspeak - sorry. I forget
what Burgess called it. It was a sort of street lingo.
R
------------------------------
Date: Sun, 09 Feb 2003 20:48:47 +0000
From: Mark Murray <ma...@grondar.org>
Subject: Re: languages
Dag-Erling Smorgrav writes:
> > The newspeak words in Burgess's "A clockwork orange" aren't in fact
> > explained anywhere
>
> Don't you mean George Orwell's _1984_?
1984 == Newspeak.
A Clockwork Orange == Nadsat (and there is a dictionary at the back
of the book).
M
- --
Mark Murray
iumop ap!sdn w,I idlaH
------------------------------
Date: Sun, 9 Feb 2003 22:06:05 +0100
From: Brad Knowles <brad.k...@skynet.be>
Subject: Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
At 3:39 PM -0400 2003/02/09, Marc G. Fournier wrote:
> 'K, one thing that I don't recall when we investigated Bugzilla way back
> when ... GNaTs has a 'send-pr' functionality that is installed by default,
> so that submitting bug reports is integrated as part of each release?
If this feature isn't there today (and I figure it could be),
then I'm sure it wouldn't be too hard to add it. The program would
still be called send-pr, but obviously it probably wouldn't function
much like the send-pr we all know and love to hate.
> And, I think, the biggest nightmare ... how do you switch over from GNaTs
> -> Bugzilla without losing a *very* large portion of your user base?
> You'd need some sort of 'GNaTs Report -> Bugzilla Entry' gateway so that
> users of older versions would still have acess ...
Damn. I wish you had asked these questions a couple of days ago.
I had some of the key bugzilla developers sitting with me at the
pre-FOSDEM drink on the Grand Place here in Brussels on Friday night,
and there was a apparently a pretty big bugzilla representation at
the conference. We could have made this a pretty short task....
Oh, and wait for a huge, huge announcement to come soon from
them. I can't tell you who's switching over, but it's a really big
announcement.
- --
Brad Knowles, <brad.k...@skynet.be>
"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
-Benjamin Franklin, Historical Review of Pennsylvania.
GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E-(---) W+++(--) N+
!w--- O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++)
tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)
------------------------------
Date: Sun, 9 Feb 2003 16:40:09 -0500
From: Jeremy Faulkner <gldi...@gldis.ca>
Subject: Re: Okay, I think I need some serious introduction ;-)
On Sun, Feb 09, 2003 at 06:45:26PM +0000, Mark Murray wrote:
> Colin Percival writes:
> > At 18:16 09/02/2003 +0000, Mark Murray wrote:
> > >Use the OS. When something bothers you, fix it :-). Submit your fix
> > >back through the usual channels to learn the project norms and customs.
> >
> > Continuing on with this process, what's the next step? Specifically,
> > after finding a bug, fixing it, submitting a PR with included patches to
> > -CURRENT and -STABLE, and watching it sit in GNATS for 8 weeks, is there
> > anything to do other than keep on waiting?
>
> Yeah. Do enough of the above to get a good portfolio of patches "on
> register". Participate in mailing lists such that your name and
> credibility are recognized. Sooner or later someone will get annoyed by
> this unprovoked display of compatibility and competence and punish you
> with a commit bit.
>
> After that, you commit your own damn patches. ;-)
>
> M
> --
> Mark Murray
> iumop ap!sdn w,I idlaH
And finally, you too can be imitated by the project's pet troll.
- --
Jeremy Faulkner http://www.gldis.ca
------------------------------
Date: Mon, 10 Feb 2003 00:38:41 +0000 (utc)
From: "David A. Hoffman" <d...@m-net.arbornet.org>
Subject: Communism
Certainly, Greg Russo has commited a grave injustice in making Kevin FW of
the polytarp conference. I demand excision of this.
------------------------------
Date: Sun, 09 Feb 2003 18:37:06 -0800
From: Terry Lambert <tlam...@mindspring.com>
Subject: Re: languages
Dag-Erling Smorgrav wrote:
> "Zoidberg" <dr.zo...@telia.com> writes:
> > says who?
> >
> > XML = Extensible Markup Language
>
> All right, show me where in the XML 1.0 specification the
> interpretation of the following snippet of XML described:
I think DES is saying that it has to have a grammar, not just
a syntax, in order to be a language. ;^).
FWIW, what XML *really* is turns out to be "data conforming to the
SGML DTD for XML".
And "an invention by IBM engineers to work around not being able
to move data through a hole in a firewall, without six months of
paperwork to get the hole approved".
In practical terms, with the sole exception of the financial
community, who, for any acceptable-to-them data marked up with
XML that DES could post, would be able to point to a standards
body ratified standard for a data dictionary that could be used
to translate the XML into other record formats (for the financial
community, XML is minimally a record format, and occasionally a
file format).
Can we now go on to argue about whether X Servers are "servers"
or not, and whether or not "Internal Revenue Service" is actually
a "service", and whether or not "technical support" is neither
"technical" nor supportive", yet again?
Thanks,
- -- Terry
------------------------------
Date: Sun, 9 Feb 2003 21:35:13 -0500 (EST)
From: soupman <so...@m-net.arbornet.org>
Subject: Re: Communism
I fully support the above measure.
On Mon, 10 Feb 2003, David A. Hoffman wrote:
> Certainly, Greg Russo has commited a grave injustice in making Kevin FW of
> the polytarp conference. I demand excision of this.
>
------------------------------
Date: Sun, 09 Feb 2003 18:57:56 -0800
From: Terry Lambert <tlam...@mindspring.com>
Subject: Re: Okay, I think I need some serious introduction ;-)
Colin Percival wrote:
> At 18:16 09/02/2003 +0000, Mark Murray wrote:
> >Use the OS. When something bothers you, fix it :-). Submit your fix
> >back through the usual channels to learn the project norms and customs.
>
> Continuing on with this process, what's the next step? Specifically,
> after finding a bug, fixing it, submitting a PR with included patches to
> -CURRENT and -STABLE, and watching it sit in GNATS for 8 weeks, is there
> anything to do other than keep on waiting?
You can apply the patch locally, so you can continue finding more
bugs and submitting more PRs.
You can write regression tests that can prove whether or not your
bug has been fixed, so that when the next release comes you, you
can regress your list of bugs, and update the "Effects" field of
the existing PRs, each time a -RELEASE happens.
You can submit PRs containing your regression code, so that others
can tell when the bug comes back, if it ever does, even after it has
been fixed and marked closed.
You can buy hardware that doesn't run under FreeBSD, and obtain the
information from the manufacturer necessary to write a FreeBSD driver
(this is often easier than obtaining information from FreeBSD on how
to write a driver, with some kernel code being undocumented ;^)).
You can buy hardware that doesn't run under FreeBSD, and for which
the manufacturer refuses to provide information, and, if you live in
certain countries, buy a copy of "Sourcer" from V Communications, Inc.,
and disassemble and reverse engineer the manufacturer's Windows driver,
for compatability purposes, as allowed by law in those countries (e.g.
Germany).
If you live in one of those countries, you can probably get a lot of
free hardware for the cost of shipping, from people who aren't in
those countries, and found out that the manufacturer would not give
you documentation, no matter how hard you pled (usually, a hell of a
lot longer than just the 8 months you were complaining about 8^p).
You can keep doing this until you get a "mentored commit bit", at
which point you can vote for core team members, and project management.
You can keep doing this with your "mentored commit bit" until you
get a "real commit bit", at which point you will be permitted to
perform small amounts of architectural work on your own, and you are
still able to vote for core team members and project management. You
can also request code by backed out, and the project by-laws require
that it be backed out, unless it's a lot of code, in which case it
stays, for fear of the backout and recommit process causing "repo bloat"
(i.e. if the expectation is that it will be recommitted later, in some
form, then your request will be denied as spurious, which it will be).
You can get elected to core, and pretty much get away with significant
architectural work, which no one else can get away with, as long as you
do not step on the toes of other core team members.
In other words, it's very much like going to work for an established
company, and being hired in at the bottom rung as "customer support
specialist", and having to work your way up.
Does that answer your question?
- -- Terry
------------------------------
Date: Sun, 09 Feb 2003 19:22:41 -0800
From: Terry Lambert <tlam...@mindspring.com>
Subject: Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Rahul Siddharthan wrote:
> GNATS does seem to be in a logjam, and will clearly get worse as the
> userbase increases. Have the FreeBSD project thought about bugzilla?
Every time someone complains about GNATS (including myself), and
then again, every time someone posts a Bugzilla root compromise
to BugTraq. 8-).
- -- Terry
------------------------------
Date: Mon, 10 Feb 2003 13:56:43 +1030
From: Greg 'groggy' Lehey <gr...@FreeBSD.org>
Subject: Re: Communism
- --tuYRN1zEaS85jg/Y
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
[trimming obviously disinterested parties]
On Monday, 10 February 2003 at 0:38:41 +0000, David A. Hoffman wrote:
> Certainly, Greg Russo has commited a grave injustice in making Kevin FW of
> the polytarp conference. I demand excision of this.
What on earth is this about?
Greg
- --
See complete headers for address and phone numbers
- --tuYRN1zEaS85jg/Y
Content-Type: application/pgp-signature
Content-Disposition: inline
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.0 (FreeBSD)
iD8DBQE+RxvzIubykFB6QiMRAqBVAJ9BvjKbWWe5+68p6MitPfgpSOscKACgmVss
pfqhUpuOSkWjxFUDe80anck=
=XAmH
- -----END PGP SIGNATURE-----
- --tuYRN1zEaS85jg/Y--
------------------------------
Date: Sun, 9 Feb 2003 22:37:11 -0500 (EST)
From: <jl...@cyberspace.org>
Subject: Re: Communism
Check out m-net.arbornet.org they are ran by communists, we think!On Mon,
10 Feb 2003, Greg 'groggy' Lehey wrote:
> [trimming obviously disinterested parties]
>
> On Monday, 10 February 2003 at 0:38:41 +0000, David A. Hoffman wrote:
> > Certainly, Greg Russo has commited a grave injustice in making Kevin FW of
> > the polytarp conference. I demand excision of this.
>
> What on earth is this about?
>
> Greg
> --
> See complete headers for address and phone numbers
>
Jeremy Allen Lamb
303 N. Mill
Camargo, IL 61919
217-832-9712
------------------------------
Date: Mon, 10 Feb 2003 03:30:30 +0000 (utc)
From: "David A. Hoffman" <d...@m-net.arbornet.org>
Subject: Re: Communism
Did you not make twinkie FW of the polytarp conference, despite: a) His
obvious intent to ruin it; and b) more people wanting styles to beFW?
On Mon, 10 Feb 2003, Greg 'groggy' Lehey wrote:
> [trimming obviously disinterested parties]
>
> On Monday, 10 February 2003 at 0:38:41 +0000, David A. Hoffman wrote:
> > Certainly, Greg Russo has commited a grave injustice in making Kevin FW of
> > the polytarp conference. I demand excision of this.
>
> What on earth is this about?
>
> Greg
> --
> See complete headers for address and phone numbers
>
------------------------------
Date: Mon, 10 Feb 2003 00:30:38 -0800
From: David Schultz <dsch...@uclink.Berkeley.EDU>
Subject: Re: Bugzilla? (was Re: Okay, I think I need some serious introduction ;-)
Thus spake Marc G. Fournier <scr...@hub.org>:
> On Sun, 9 Feb 2003, Rahul Siddharthan wrote:
>
> > As I understand, all this is essentially impossible to do in GNATS, and
> > things are going to get worse.
>
> 'K, one thing that I don't recall when we investigated Bugzilla way back
> when ... GNaTs has a 'send-pr' functionality that is installed by default,
> so that submitting bug reports is integrated as part of each release?
>
> And, I think, the biggest nightmare ... how do you switch over from GNaTs
> -> Bugzilla without losing a *very* large portion of your user base?
> You'd need some sort of 'GNaTs Report -> Bugzilla Entry' gateway so that
> users of older versions would still have acess ...
Given that they've figured out how to convert entire GNATS
databases to Bugzilla databases automatically, I'm sure backwards
compatibility for new bug reports could be implemented. Besides,
everyone and his dog uses the web these days, so despite how much
I hate the web, I have to say that the worst case is that people
with old versions are annoyed when they are forced to use lynx or
(heaven forbid) a mouse.
------------------------------
Date: Mon, 10 Feb 2003 17:58:18 -0800
From: Terry Lambert <tlam...@mindspring.com>
Subject: Re: matthew dillon
Dave Hayes wrote:
> Terry Lambert <tlam...@mindspring.com> writes:
> > Do you unsubscribe from mailing lists you merely monitor for
> > interesting content, rather than subscribing to them, when some
> > jerk fills up your POP3 maildrop because they have an axe to
> > grind, and, as a result, mail which you consider "important",
> > compared to the list traffic, bounces?
>
> I don't use POP3, precisely because of that reason. Do you?
What you do or not do is irrelevent to the fact that some people
can not obtain service that doesn't involve their email piling
up somewhere it has to be downloaded from.
In addition, not everyone can run a mail server, for lack of
IPv4 address space, and due to service provider restrictions on
the ability to run servers on their network connections due to
active firewalling to create an artificial tiering of pricing,
while avoiding the oversight of the PUC by not seperating it
into a new tarrif group.
> > People who advocate "receiver filtering" (either of the active
> > variety, or of the "just ignore" variety) is the answer to all
> > SPAM-like problems apparently do not understand the realities
> > of many people using pull-based rather than push-based email
> > transports.
>
> We do understand those realities, which is why we contend that
> pull-based systems aren't the correct technology to use for receiving
> randomly ubiquitous content such as humans are likely to generate.
Until the technologies are no longer being deployed against
new users, live in the world as it is, not as you wish it
were.
> I recognize that some people are unable to leave their POP client
> connected 24/7 with "leave mail on server" unchecked and with
> a scan rate of "once every 2 minutes". Perhaps a digested form
> of the mailing list or a web browsable archive should exist for
> those people's needs?
The problem is that a denial of service attack can be successful,
even in that case, by using a sufficiently large message size, or
a sufficiently high message frequency, or a combination of the
two (e.g. the recent troll repetitive mailings that cause this
thread to be started were once-a-second, from my reading of the
email headers).
How is it that you suggest people defend against people with
bigger pipes for shoving messages out than people have for
messages coming in? In the limit, the same argument will apply
to push-based systems, eventually, since you can not RED-queue
persistent TCP connections, only incoming connection requests.
> The technology is supposed to serve you, not dictate how you
> are supposed to communicate.
Feeel free to correct it, and every exisitng instance of it on
the Internet, and then, after you have done that, get back to
me, and I may indeed be willing to agree with your arguments.
NB: If you are going to deal with this, then please, at the
same time, fix the FIN-WAIT-2 problem, which is caused by a
protocol design error in TCP, which requires two responses to
a single request, with no way for the requester to re-request
the first of the two responses.
> > Please understand the technology involved before telling people
> > how they should use it.
>
> Please understand the people involved before attempting to force
> people to behave based on a particular choice of technology. =)
The technology used dictates the permissable behaviours of
the people using it; whether you like that fact or not, it is
nonetheless true.
- -- Terry
------------------------------
End of freebsd-chat-digest V5 #699
**********************************
To Unsubscribe: send mail to majo...@FreeBSD.org
with unsubscribe freebsd-chat-digest in the body of the message