[9fans] 9fans-ego-wars?

[9fans] 9fans-ego-wars?

[Sam]

I'm as guilty as any about shooting spitwads
from the back of the room, but if the ego
battles must continue with such fervency
perhaps we could create another list so
people can insult each other and still feel
like they have an audience?

Or maybe all that wasted energy could be
redirected at writing a web browser?

Sam

Re: [9fans] 9fans-ego-wars?

[Dan Cross]

It exists.  Hangar18-general@open-forge.org.

	- Dan C.

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> Or maybe all that wasted energy could be
> redirected at writing a web browser?

sounds like a plan, but it's an 'army of programmers' because of
the complexity of the problem.  i don't like the exploder, but it
works a darn sight better than netscrape.

unless i'm missing something, this is a serious 'death march'.

mothra might be a start and the netscrape code is pub domain
(iirc and probably bloody awful).

duff would know what to do.

Re: SPAM: Re: [9fans] 9fans-ego-wars?

[Charles Forsyth]

>>duff would know what to do.

i think he did: he stopped (didn't he)

Re: SPAM: Re: [9fans] 9fans-ego-wars?

[David Presotto]

Is howard's browser being used in Inferno at all?  The auto
translated limbo->c version 'i' is pretty crufty but only
because the conversion never got finished.  It would be
good starting point.  Even if you disagree with the style,
it is relatively small and might have pieces that can be
salvaged.

Re: [9fans] 9fans-ego-wars?

[Charles Forsyth]

[-- Attachment #1: Type: text/plain, Size: 1574 bytes --]

yes, and chris has worked on it (and john for Ecmascript) so it does
well enough that amongst most (not all) other sites
i can use it to access my credit card, and complete
the Inland Revenue online Tax Return.  the Charon source is in the
free download.  amongst other things,
you might need to change the configuration so that
it lies about which browser it is.  it's a reasonable place to start.
it's probably easier to work on its interface to plan9
than it is to try translating it all into C or C++ or Erlang or something.
(the emu source is also now on our Plan 9 CD
and will be in the download shortly.)

for listeners in the United Kingdom:
for the Revenue site, it's important not to use a Digital Certificate
to access via the Government Gateway: use the password scheme.

if you try the digital certificate, first
you'll need to buy one after playing
a guessing game about your credit record with a credit
company that then gives you a certificate with a root CA
in California to identify me to my own non-Californian Government;
then you'll need to pay them to renew it each year, playing the
same game again; and finally,
it doesn't actually work even with Windows.
(``ah yes, we've had a lot of trouble with those digital certificates.'')

it should be obvious that the only thing the game does is show
that i have access to my credit record not who i am.
the really clever bit comes when a loan has been farmed out by
one company to another so you haven't got the slightest
idea what they're talking about.  but you can guess.

[-- Attachment #2: Type: message/rfc822, Size: 2213 bytes --]

From: David Presotto <presotto@closedmind.org>
To: 9fans@cse.psu.edu
Subject: Re: SPAM: Re: [9fans] 9fans-ego-wars?
Date: Thu, 25 Sep 2003 15:35:31 -0400
Message-ID: <5242a264d58cea1f282b7d81b87fa21f@plan9.bell-labs.com>

Is howard's browser being used in Inferno at all?  The auto
translated limbo->c version 'i' is pretty crufty but only
because the conversion never got finished.  It would be
good starting point.  Even if you disagree with the style,
it is relatively small and might have pieces that can be
salvaged.

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> you might need to change the configuration so that
> it lies about which browser it is.

that's a very good point.  these dickhead web 'designers'
who use browser specific junk should be hunted down.

but it's one big 'arms race'; they change stuff on a whim
and then it doesn't work anymore.  the number of broken
web pages/sites is truly astounding/horrifying.

my swedish bank went beserk when it saw i was using
mozilla.  they do, however, hand you a cryptokey so
you can auth yourself over the TLS/SSL connection.

so i went back to the exploder, on a non lunix box.

i know one bank that uses 6 char [a-z0-9] case insensitive
'pawwords' over TLS/SSL.  said bank asked me for my
'password' over the phone, in _direct violation_ of their
own stupid 'security' rules.  my response to that 'raised
some pulses'.

Re: SPAM: Re: [9fans] 9fans-ego-wars?

[C H Forsyth]

>>my swedish bank went beserk when it saw i was using
>>mozilla.  they do, however, hand you a cryptokey so
>>you can auth yourself over the TLS/SSL connection.

i could have added that the problem--sorry, one problem--with the Government
Gateway and certificates was that it used a Java applet to
fetch your certificate from your certificate safe (ha ha), and sign something with it,
and the implementation produces a bizarre diagnostic if you happen
to have ticked the box that chooses Sun's Java in Explorer
(diagnostic roughly equivalent in appearance to ed's `?'
but manages to be less informative yet somehow more worrying); if you
work out your mistake and untick the box to use Microsoft's
Java it still fails, with a different but at least legible
diagnostic (``the certificate is invalid'').

Re: [9fans] 9fans-ego-wars?

[ron minnich]

On Thu, 25 Sep 2003, boyd, rounin wrote:

> but it's one big 'arms race'; they change stuff on a whim
> and then it doesn't work anymore.  the number of broken
> web pages/sites is truly astounding/horrifying.

it's even worse. FrontPage generaete HTML that works with ... Internet
Explorer. Sometimes it works with other browsers, usually by happenstance.

ron

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> it's even worse. FrontPage generaete HTML that works with ... Internet
> Explorer. Sometimes it works with other browsers, usually by happenstance.

yup been there.  had to fix that.  non-trivial, given the way they had
set up their dmz, firewalls and web servers.

Re: [9fans] 9fans-ego-wars?

[Tristan Seligmann]

[-- Attachment #1: Type: text/plain, Size: 287 bytes --]

On Thu, Sep 25, 2003 at 16:07:04 -0600, ron minnich wrote:
> it's even worse. FrontPage generaete HTML that works with ... Internet 
> Explorer. Sometimes it works with other browsers, usually by happenstance.

Even with Internet Explorer, you still take your chances.

mithrandi

[-- Attachment #2.1: Type: text/plain, Size: 309 bytes --]

The following attachment had content that we can't
prove to be harmless.  To avoid possible automatic
execution, we changed the content headers.
The original header was:

	Content-Type: application/pgp-signature; name="signature.asc"
	Content-Description: Digital signature
	Content-Disposition: inline

[-- Attachment #2.2: signature.asc.suspect --]
[-- Type: application/octet-stream, Size: 196 bytes --]

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQE/c2jTgfSlm3IcLfgRAsbbAKC7HRYDlhQxulenDokHRhV7iUOgtQCgm8rL
m6QCkHmZgutaeqwJjG7JXFo=
=ZlEo
-----END PGP SIGNATURE-----

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

    From: "Tristan Seligmann" <mithrandi-9fans@mithrandi.za.net>

ditto

Re: [9fans] 9fans-ego-wars?

[Tristan Seligmann]

On Fri, Sep 26, 2003 at 00:35:59 +0200, boyd, rounin wrote:
>     From: "Tristan Seligmann" <mithrandi-9fans@mithrandi.za.net>
>
> ditto

Yeah ok, I normally use PGP/MIME...

mithrandi

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> if you try the digital certificate, first
> you'll need to buy one after playing
> a guessing game about your credit record with a credit
> company that then gives you a certificate with a root CA
> in California to identify me to my own non-Californian Government;
> then you'll need to pay them to renew it each year, playing the
> same game again; and finally,
> it doesn't actually work even with Windows.
> (``ah yes, we've had a lot of trouble with those digital certificates.'')

yes, the PKI is broken.

Re: SPAM: Re: [9fans] 9fans-ego-wars?

[C H Forsyth]

[-- Attachment #1: Type: text/plain, Size: 299 bytes --]

yes and no.  yes, x.509 chain-to-essentially-ignorant-root-CA is broken,
but i gathered that at least some of the PKI inventors
did not really intend that, but its liberal, decentralised opposite.
fortunately there is a counter-culture.
(i suppose one could add: unfortunately, there are three.)

[-- Attachment #2: Type: message/rfc822, Size: 2853 bytes --]

From: "boyd, rounin" <boyd@insultant.net>
To: <9fans@cse.psu.edu>
Subject: SPAM: Re: [9fans] 9fans-ego-wars?
Date: Thu, 25 Sep 2003 22:30:05 +0200
Message-ID: <01ef01c383a3$ce075200$b9844051@insultant.net>

> if you try the digital certificate, first
> you'll need to buy one after playing
> a guessing game about your credit record with a credit
> company that then gives you a certificate with a root CA
> in California to identify me to my own non-Californian Government;
> then you'll need to pay them to renew it each year, playing the
> same game again; and finally,
> it doesn't actually work even with Windows.
> (``ah yes, we've had a lot of trouble with those digital certificates.'')

yes, the PKI is broken.

Re: SPAM: Re: [9fans] 9fans-ego-wars?

[Fco.J.Ballesteros]

> Is howard's browser being used in Inferno at all?  The auto

If you refer to charon, it's what I use all the time.

Re: [9fans] 9fans-ego-wars?

[Andrew Simmons]

>sounds like a plan, but it's an 'army of programmers' because of
>the complexity of the problem.  i don't like the exploder, but it
>works a darn sight better than netscrape.
>
>unless i'm missing something, this is a serious 'death march'.
>
>mothra might be a start and the netscrape code is pub domain
>(iirc and probably bloody awful).
>

Don't shout at me, but I've been using Mozilla Firebird for the past few
months, and really like it. It feels much snappier and less bloated than
Mozilla, especially when you start it up, and the screen layout is, to my
eye, very clean and uncluttered. I particularly like the layout of the
options - all the ones I want to change are easy to get at, not buried ten
levels deep - and the fact that you can leave Javascript on but disable
popup windows. No idea what the source code is like though.

Re: [9fans] 9fans-ego-wars?

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 500 bytes --]

Firebird has turned into the browser of choice in our group also.  It
is less bloated, and can navigate most of our internal, built expressly
for Exploder, web pages.  I haven't looked at the code but ehg has
gotten it to compile for µsoft I believe.  I use it by vncing
to a linux box.

To tell you the truth, in the absence of anything, I'ld be more
than willing to accept whatever is offered.  I'll start by
getting a new emu (mu old one just crashes and burns)
and running charon again.

[-- Attachment #2: Type: message/rfc822, Size: 3011 bytes --]

From: Andrew Simmons <andrew.simmons@monitorbm.co.nz>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] 9fans-ego-wars?
Date: Fri, 26 Sep 2003 10:42:04 +1200
Message-ID: <5.1.1.6.0.20030926103508.00b193e8@pop.monitorbm.co.nz>


>sounds like a plan, but it's an 'army of programmers' because of
>the complexity of the problem.  i don't like the exploder, but it
>works a darn sight better than netscrape.
>
>unless i'm missing something, this is a serious 'death march'.
>
>mothra might be a start and the netscrape code is pub domain
>(iirc and probably bloody awful).
>

Don't shout at me, but I've been using Mozilla Firebird for the past few
months, and really like it. It feels much snappier and less bloated than
Mozilla, especially when you start it up, and the screen layout is, to my
eye, very clean and uncluttered. I particularly like the layout of the
options - all the ones I want to change are easy to get at, not buried ten
levels deep - and the fact that you can leave Javascript on but disable
popup windows. No idea what the source code is like though.

Re: [9fans] 9fans-ego-wars?

[ron minnich]

On Thu, 25 Sep 2003, David Presotto wrote:

> To tell you the truth, in the absence of anything, I'ld be more
> than willing to accept whatever is offered.  I'll start by
> getting a new emu (mu old one just crashes and burns)
> and running charon again.

This brings up a more interesting question. Now that the Plan 9 and
Inferno code bases are open source, is some sort of convergance in the
future? Seems like it'd be nice to have them all in one code tree,
somehow. Limbo and all that things that come with it seem quite neat.

Possible? Ridiculous? Uninformed, inquiring minds want to know!

ron

Re: [9fans] 9fans-ego-wars?

[Charles Forsyth]

>>Possible? Ridiculous? Uninformed, inquiring minds want to know!

there are points where the systems take significantly different directions,
but there are indeed significant areas where they don't.

>>Limbo and all that things that come with it seem quite neat.

i'd prefer to do most of the programming i do in Limbo (sounds quite detached, doesn't it),
and i've therefore got some incentive to make that possible without (completely)
screwing up either system.

Re: [9fans] 9fans-ego-wars?

[Charles Forsyth]

>>limbo is the business.  no more having to hack C in user mode

i've sometimes wondered whether i might like to have a variant that can
be used for lower-level implementation (cf. ESPOL).

Re: [9fans] 9fans-ego-wars?

[vdharani]

>>>limbo is the business.  no more having to hack C in user mode
>
> i've sometimes wondered whether i might like to have a variant that can
> be used for lower-level implementation (cf. ESPOL).

First, I am not an expert in languages and implementation but I just want
to share my opinion.

I used to think a new C derivative with Limbo features into it would be
very nice (and this should have been the actual sequel to C instead of
C++). There are nice features in Limbo (like modules, adts, slicing,
automatic declaration(?), etc) that could as well be in C while retaining
C's core style (without the need for GC) but increasing the reliability and
robustness.

Then, again, one has to worry about another C variations and cross-platform
issues.

In 1996, I wrote a small protocol stack in restricted C (that resembled
Limbo code). That time all I had was just Kernighan's Limbo tutorial paper.
Eventually I was easily able to convert it to Limbo by using scripts. After
that I followed the same style (restricted C, resembling Limbo) for
implementing many protocols and drivers in C. I like that approach and
would like to stick to it till I see a better alternative.

As as aside, it would be nice to have a Limbo-to-C transator so that one
can code in Limbo (atleast in the initial stages of the development
process) and then convert it to C. Of course, things like arrays (as they
use GC in Limbo) may not be trivial but then I think it would be a good
approach.

Thanks
dharani

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> I used to think a new C derivative with Limbo features into it would be
> very nice (and this should have been the actual sequel to C instead of
> C++). There are nice features in Limbo (like modules, adts, slicing,
> automatic declaration(?), etc) that could as well be in C while retaining
> C's core style (without the need for GC) but increasing the reliability and
> robustness.

what is this?  a language super-collider?  i don't want to see two perfectly
good languages smashed together, producing yet another bastardised
piece of junk.  have you looked at C# [reaches for ...]?

Re: [9fans] 9fans-ego-wars?

[Charles Forsyth]

i must remember `language super-collider'.
that reminded me of several remarks by Peter Brown:

If the programming language has all the grace
of Georgian architecture it is a sin to build
a concrete and glass extension.

Taken as a whole the OS macro-assembler is utilitarian
rather than conceptually beautiful; not the girl you
would dream about, but the sort you might marry.

{Macro Processors and Techniques for Portable Software,
P.J. Brown, Wiley & Sons, London, 1974.}

Re: [9fans] 9fans-ego-wars?

[vdharani]

>> I used to think a new C derivative with Limbo features into it would
>> be very nice (and this should have been the actual sequel to C instead
>> of C++). There are nice features in Limbo (like modules, adts,
>> slicing, automatic declaration(?), etc) that could as well be in C
>> while retaining C's core style (without the need for GC) but
>> increasing the reliability and robustness.
>
> what is this?  a language super-collider?  i don't want to see two
> perfectly good languages smashed together, producing yet another
> bastardised piece of junk.
I dont know how to put it but if I am right, Limbo is very good but one
cant use it for low level programming (like GC reducing predicatable
timing). C can be used well for low-level programming but there are
concerns about pitfalls, reliability and robustness (more because of bad
usage). I thought a balance between the two could greatly be useful for
kernel and embedded code.

BTW, I was once reading about profile-guided compilation by Intel for IA-64
and what all it does to check what the pointers actually point to at any
point (some localization). I was wondering how easy it is to do in a
language like Limbo and the opportunities Limbo give for a compiler writer
for optimization. So, why cant there be a better C that is clean too?

Again, I am not an expert in these areas but I am just telling my opinion.

Thanks
dharani

Re: [9fans] 9fans-ego-wars?

[C H Forsyth]

>>I dont know how to put it but if I am right, Limbo is very good but one
>>cant use it for low level programming (like GC reducing predicatable
>>timing). C can be used well for low-level programming but there are

ESPOL was an extended subset of the Burroughs Extended ALGOL language
that was tuned to implement systems programs such as the operating system (MCP).
(i'd hoped the MCP source code was online by now but i couldn't find it;
its use of call-by-name to wait for an event was fascinating.)
anyhow, i thought it would be not too hard to do the same for Limbo,
once i'd decided it was really a sensible idea, that is.
i did think that in practice it might boil down to just preventing
the use of expensive things and limiting it to reference counting,
which could probably be done without actually creating a true dialect.

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> ESPOL was an extended subset of the Burroughs Extended ALGOL language
> that was tuned to implement systems programs such as the operating system
(MCP).

MCP?  i think we had such a beast at basser.  it was a horrible
thing, but it did had some damn interesting ideas.

M[aster] C[ontrol] P[rogram], right?

Re: [9fans] 9fans-ego-wars?

[a]

// M[aster] C[ontrol] P[rogram], right?

yeah, from Tron! what a great movie...
:-)
ア

Re: [9fans] 9fans-ego-wars?

[Brantley Coile]

Ah, youth.  And the Beatles Revolution is the Nike song. :)

 Brantley


On Sat, 27 Sep 2003 12:19:18 -0400, <a@9srv.net> wrote:

> // M[aster] C[ontrol] P[rogram], right?
>
> yeah, from Tron! what a great movie...
> :-)
> ア
>

Re: [9fans] 9fans-ego-wars?

[a]

i actually didn't know Nike used it. that's
kind of... amusing. and the beatles had a *few*
revolutions.
ア

Re: [9fans] 9fans-ego-wars?

[ron minnich]

On Sat, 27 Sep 2003 a@9srv.net wrote:

> // M[aster] C[ontrol] P[rogram], right?
>
> yeah, from Tron! what a great movie...

I was at Burroughs CPU group when Tron came out. At the time, we were told
Burroughs thought about suing Disney, since Burroughs had been using the
term MCP for about 20 years for its OS. They never did sue, thank
goodness.

ron

Re: [9fans] 9fans-ego-wars?

[okamoto]

> On Sat, 27 Sep 2003 a@9srv.net wrote:
>> yeah, from Tron! what a great movie...

I thought it's a OS in Japan. ☺

Kenji

Re: [9fans] 9fans-ego-wars?

[boyd]

oh kenji-kun, you've just given me an evil idea.

Re: [9fans] 9fans-ego-wars?

[okamoto]

> oh kenji-kun, you've just given me an evil idea.

When you are younger than me, I believe it, you should not
call me -kun, instead call -san.  right?   ☺

Kenji

Re: [9fans] 9fans-ego-wars?

[boyd]

i know that.  guess how old i am?

but it's more complicated.  -kun is used with people who
are 'friends'.

Re: [9fans] 9fans-ego-wars?

[okamoto]

> i know that.  guess how old i am?
 I have confidence that you are younger than me from your
many writings.

> but it's more complicated.  -kun is used with people who
> are 'friends'.

friends?  I don't know you. ☺

Kenji

Re: [9fans] 9fans-ego-wars?

[boyd]

    I don't know you

QED

Re: [9fans] 9fans-ego-wars?

[Vester Thacker]

On Mon, Sep 29, 2003 at 12:01:29PM +0900, okamoto@granite.cias.osakafu-u.ac.jp wrote:
> > oh kenji-kun, you've just given me an evil idea.
>
> When you are younger than me, I believe it, you should not
> call me -kun, instead call -san.  right?   ???
>
> Kenji


Right, Kanji-san would be more appropriate in this case. If both of
you don't know the other well, then -san should be used.

-- Vester Thacker

Re: [9fans] 9fans-ego-wars?

[boyd]

    Right, Kanji-san would be more appropriate in this case. If both of
    you don't know the other well, then -san should be used.

so, i should stick -san on everone's name or call them Mr/Ms/Mrs/Sir/Madam ... ?

Re: [9fans] 9fans-ego-wars?

[Vester Thacker]

On Mon, Sep 29, 2003 at 05:40:51AM -0400, boyd@sdgm.net wrote:
>     Right, Kanji-san would be more appropriate in this case. If both of
>     you don't know the other well, then -san should be used.
>
> so, i should stick -san on everone's name or call them Mr/Ms/Mrs/Sir/Madam ... ?

No, but if you are going to use -kun you should use it correctly don't
you think.

-- Vester Thacker

Re: [9fans] 9fans-ego-wars?

[boyd]

i'm calling nobody sir et al or -san, except when i'm in bullshit business situation.

liberté, égalité fraternité ...

Re: [9fans] 9fans-ego-wars?

[Vester Thacker]

On Mon, Sep 29, 2003 at 06:11:42AM -0400, boyd@sdgm.net wrote:
> i'm calling nobody sir et al or -san, except when i'm in bullshit business situation.
>
> libert??, ??galit?? fraternit?? ...
Yeah, yeah, yeah. Take it off the list and to irc.freenode.org #plan9
if you got time to kill.

-- Vester Thacker

Re: [9fans] 9fans-ego-wars?

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 88 bytes --]

Judging from the last few weeks, boyd's got time to kill and
the weapons to do it with.

[-- Attachment #2: Type: message/rfc822, Size: 2549 bytes --]

From: Vester Thacker <vthacker@0xfe.org>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] 9fans-ego-wars?
Date: Mon, 29 Sep 2003 19:28:19 +0900
Message-ID: <20030929102819.GA2684@0xfe.org>

On Mon, Sep 29, 2003 at 06:11:42AM -0400, boyd@sdgm.net wrote:
> i'm calling nobody sir et al or -san, except when i'm in bullshit business situation.
>
> libert??, ??galit?? fraternit?? ...
Yeah, yeah, yeah. Take it off the list and to irc.freenode.org #plan9
if you got time to kill.

-- Vester Thacker

Re: [9fans] 9fans-ego-wars?

[boyd]

Ooo-Rah!!

Re: [9fans] 9fans-ego-wars?

[bs]

C H Forsyth wrote:
>>>I dont know how to put it but if I am right, Limbo is very good but one
>>>cant use it for low level programming (like GC reducing predicatable
>>>timing). C can be used well for low-level programming but there are
>
>
> ESPOL was an extended subset of the Burroughs Extended ALGOL language
> that was tuned to implement systems programs such as the operating system (MCP).
> (i'd hoped the MCP source code was online by now but i couldn't find it;
> its use of call-by-name to wait for an event was fascinating.)
> anyhow, i thought it would be not too hard to do the same for Limbo,
> once i'd decided it was really a sensible idea, that is.
> i did think that in practice it might boil down to just preventing
> the use of expensive things and limiting it to reference counting,
> which could probably be done without actually creating a true dialect.
Truly the Burroughs ALGOL derivatives are excellent:
NEWP(New Prog Lang) for the OS
DMALGOL for Databases
DCALGOL for DataComm
each have extensions to suite their purpose.
IIRC ESPOL was also implemented on Burroughs systems, but I cannot
recall specific references.

MCP sources were never available online, however, any site could have
the sources to MCP, DMSII, COMS if they wanted. In fact many sites had
their own communications systems (before COMS, after GEMCOS), and a lot
of sites did submit patches to all the above.

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> I dont know how to put it but if I am right, Limbo is very good but one
> cant use it for low level programming (like GC reducing predicatable
> timing). C can be used well for low-level programming but there are
> concerns about pitfalls, reliability and robustness (more because of bad
> usage). I thought a balance between the two could greatly be useful for
> kernel and embedded code.

use the right tool for the job.  i'll stick with the M-9, dan can stick with
K-bar, but i might have a backup plan too:

    http://www.insultant.net/barstow/MVC-764F.JPG

Re: [9fans] 9fans-ego-wars?

[Skip Tavakkolian]

> I used to think a new C derivative with Limbo features into it would be
> very nice (and this should have been the actual sequel to C instead of
> C++). There are nice features in Limbo (like modules, adts, slicing,
> automatic declaration(?), etc) that could as well be in C while retaining
> C's core style (without the need for GC) but increasing the reliability and
> robustness.

Take a look at Pike (the language, not the man).  It is the C like
language that the Roxen web server is written in.  It extends lists,
arrays, etc to first class types, has GC, adts, slicing, etc.  It is
expressive enough, but it is inconsistent and perhaps the most tedious
are the type declarations like:

array(array(string)) x;		// Pike

compare this to:

x : array of array of string;	# Limbo

Also the notation := in Limbo is pretty convenient and lacking in Pike.

FYI, Pike homepage is:

http://pike.ida.liu.se/

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> Limbo and all that things that come with it seem quite neat.

limbo is the business.  no more having to hack C in user mode
or write in 'orrible python or in various other disgusting hacks,
err, programming languages.

Re: [9fans] 9fans-ego-wars?

[Charles Forsyth]

>>than willing to accept whatever is offered.  I'll start by
>>getting a new emu (mu old one just crashes and burns)
>>and running charon again.

wait a day or so and the newest one with the current
charon will be there.

Re: [9fans] 9fans-ego-wars?

[Andrew Simmons]

At 19:02 25/09/2003 -0400, you wrote:
>Firebird has turned into the browser of choice in our group also.  It
>is less bloated, and can navigate most of our internal, built expressly
>for Exploder, web pages.  I haven't looked at the code but ehg has
>gotten it to compile for µsoft I believe.  I use it by vncing
>to a linux box.

There are also versions for Windows and MacOSX to be had at:

http://www.mozilla.org/products/firebird/

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

    http://www.mozilla.org/products/firebird/

when i hear the word mozilla i reach for my MP5.

Re: [9fans] 9fans-ego-wars?

[Andrew Simmons]

>
>when i hear the word mozilla i reach for my MP5.

I think it might narrow the field down more if you told us under what
circumstances you wouldn't reach for your MP5.

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> I think it might narrow the field down more if you told us under what
> circumstances you wouldn't reach for your MP5.

yes, a cache miss cache :)

Re: [9fans] 9fans-ego-wars?

[D. Brownlee]

could a cache miss cache be classified as a "feature
request/enhancement" bug?

boyd, rounin wrote:

>>I think it might narrow the field down more if you told us under what
>>circumstances you wouldn't reach for your MP5.
>
>
> yes, a cache miss cache :)
>
>

Re: [9fans] 9fans-ego-wars?

[boyd, rounin]

> could a cache miss cache be classified as a "feature
> request/enhancement" bug?

iirc kre wrote one for the 4.2 BSD namei, so typos wouldn't
cause a zillion instructions to get executed just to return ENOENT.

Re: [9fans] 9fans-ego-wars?

[Ishwar Rattan]

Amen

-ishwar


On Thu, 25 Sep 2003, Sam wrote:

> I'm as guilty as any about shooting spitwads
> from the back of the room, but if the ego
> battles must continue with such fervency
> perhaps we could create another list so
> people can insult each other and still feel
> like they have an audience?
>
> Or maybe all that wasted energy could be
> redirected at writing a web browser?
>
> Sam
>
>
>

Re: [9fans] 9fans-ego-wars?

[Dan Cross]

> I used to think a new C derivative with Limbo features into it would be
> very nice (and this should have been the actual sequel to C instead of
> C++). There are nice features in Limbo (like modules, adts, slicing,
> automatic declaration(?), etc) that could as well be in C while retaining
> C's core style (without the need for GC) but increasing the reliability and
> robustness.

I've become a fan of Cyclone in the last year or so.  I sort of think
of it as a C on steroids, and frankly, I'd much rather be programming
in it for many common programming tasks than programming in strict
ANSI C.  http://www.research.att.com/projects/cyclone/

	- Dan C.

Re: [9fans] 9fans-ego-wars?

[Dennis Ritchie]

Forsyth writes,

 > ESPOL was an extended subset of the Burroughs Extended ALGOL language
 > that was tuned to implement systems programs such as the operating system (MCP).
 > (i'd hoped the MCP source code was online by now but i couldn't find it;
 > its use of call-by-name to wait for an event was fascinating.)

I haven't found MCP, but ESPOL and extended ALGOL manuals
are to be found at Al Kossow's site:

  http://www.spies.com/~aek/pdf/burroughs/B5500/

		Dennis

Re: [9fans] 9fans-ego-wars?

[Douglas A. Gwyn]

Dennis Ritchie wrote:
> I haven't found MCP, but ESPOL and extended ALGOL manuals
> are to be found at Al Kossow's site:

I may have some of the "internals" manuals (blue cover
as opposed to the usual red cover) buried in my archives;
they would be left over from my days at Rice in the late
1960s.  Unfortunately it will be quite some time before
I can look for them.  If I find them I'll try to get
copies to Al.  By the way, I sure hope that somebody is
making an archival backup of his site in case it goes
away some day.

Re: [9fans] 9fans-ego-wars?

[bs]

C H Forsyth wrote:

 >>> I dont know how to put it but if I am right, Limbo is very good but one
 >>> cant use it for low level programming (like GC reducing predicatable
 >>> timing). C can be used well for low-level programming but there are
 >
 >
 >
 > ESPOL was an extended subset of the Burroughs Extended ALGOL language
 > that was tuned to implement systems programs such as the operating
system (MCP).
 > (i'd hoped the MCP source code was online by now but i couldn't find it;
 > its use of call-by-name to wait for an event was fascinating.)
 > anyhow, i thought it would be not too hard to do the same for Limbo,
 > once i'd decided it was really a sensible idea, that is.
 > i did think that in practice it might boil down to just preventing
 > the use of expensive things and limiting it to reference counting,
 > which could probably be done without actually creating a true dialect.

Truly the Burroughs ALGOL derivatives are excellent:
NEWP(New Prog Lang) for the OS
DMALGOL for Databases
DCALGOL for DataComm
each have extensions to suite their purpose.
IIRC ESPOL was also implemented on Burroughs systems, but I cannot
recall specific references.

MCP sources were never available online, however, any site could have
the sources to MCP, DMSII, COMS if they wanted. In fact many sites had
their own communications systems (before COMS, after GEMCOS), and a lot
of sites did submit patches to all the above.

Re: [9fans] 9fans-ego-wars?

[Charles Forsyth]

>>MCP sources were never available online, however, any site could have

i thought it might have ended up being made visible by its
owner(s) in a similar way to the old unix sources but apparently not.