[9fans] Brdline

[9fans] Brdline

[arisawa]

Hello,

The following program falls into endless loop if we execute
	echo -n alice | 8.out
Try.

#include <u.h>
#include <libc.h>
#include <bio.h>

void
main()
{	Biobuf in;
	char *p;
	int lines;
	Binit(&in, 0, OREAD);
	lines = 0;
	for(;;){
		p = Brdline(&in, '\n');
		if(p == nil){
			if(Blinelen(&in) == 0)
				break;
		}else
			lines++;
	}
	print("lines=%d\n",lines);
}



I have been desired to have such a Brdline
that can work even if we execute:
	echo -n alice | 8.out

Current Brdline requires delim as second argument,
and the worse the delim is left untouched *mostly*
at Blinelen -1
We must replace it using Blinelen by '\0' if we want the
returned value to be used as string.
This makes program difficult if input data doesn't have delim.

My imagination of new Brdline (or some other name) is:

void
main()
{	Biobuf in;
	char *p;
	Binit(&in,0,OREAD);
	while((p = Brdline(&in)))
		print("%s", p);
}

'\n' is automatically replaced by '\0',
and '\0' is automatically appended at EOF.
p is nil only at EOF.
some function will be required to detect buffer full.
if buffer full the next Brdline should return the
pointer of the character that is replaced by '\0'
after fixing the value.

Kenji Arisawa

Re: [9fans] Brdline

[Russ Cox]

try using Brdstr instead.



On Wed, 2 Feb 2005 09:37:14 +0900, arisawa@ar.aichi-u.ac.jp
<arisawa@ar.aichi-u.ac.jp> wrote:
> Hello,
>
> The following program falls into endless loop if we execute
>         echo -n alice | 8.out
> Try.
>
> #include <u.h>
> #include <libc.h>
> #include <bio.h>
>
> void
> main()
> {       Biobuf in;
>         char *p;
>         int lines;
>         Binit(&in, 0, OREAD);
>         lines = 0;
>         for(;;){
>                 p = Brdline(&in, '\n');
>                 if(p == nil){
>                         if(Blinelen(&in) == 0)
>                                 break;
>                 }else
>                         lines++;
>         }
>         print("lines=%d\n",lines);
> }
>
> I have been desired to have such a Brdline
> that can work even if we execute:
>         echo -n alice | 8.out
>
> Current Brdline requires delim as second argument,
> and the worse the delim is left untouched *mostly*
> at Blinelen -1
> We must replace it using Blinelen by '\0' if we want the
> returned value to be used as string.
> This makes program difficult if input data doesn't have delim.
>
> My imagination of new Brdline (or some other name) is:
>
> void
> main()
> {       Biobuf in;
>         char *p;
>         Binit(&in,0,OREAD);
>         while((p = Brdline(&in)))
>                 print("%s", p);
> }
>
> '\n' is automatically replaced by '\0',
> and '\0' is automatically appended at EOF.
> p is nil only at EOF.
> some function will be required to detect buffer full.
> if buffer full the next Brdline should return the
> pointer of the character that is replaced by '\0'
> after fixing the value.
>
> Kenji Arisawa
>

Re: [9fans] Brdline

[Bruce Ellis]

or making irrational assumptions about what the compiler does
without ever reading K&R.

brucee

On 1/19/06, Adrian Tritschler <ajft@ajft.org> wrote:
> Ronald G Minnich wrote:
> > read "the art of networking style" for a great rant on standards.
> >
> > Note that film standards are ISO/DIN numbers -- not one number, two.
> > That's how they resolved how to pick the #.
> >
> > 48 byte cells came from standards.
>
> Ah, but think of the elegance that comes from adding an odd byte-length
> of header to result in a prime-number sized packet :-)
>
> > ron
>        Adrian
>
> ---------------------------------------------------------------
> Adrian Tritschler                          mailto:ajft@ajft.org
> Latitude 38°S, Longitude 145°E, Altitude 50m,      Shoe size 44
> ---------------------------------------------------------------
>

Re: [9fans] Brdline

[Adrian Tritschler]

Ronald G Minnich wrote:
> read "the art of networking style" for a great rant on standards.
> 
> Note that film standards are ISO/DIN numbers -- not one number, two.
> That's how they resolved how to pick the #.
> 
> 48 byte cells came from standards.

Ah, but think of the elegance that comes from adding an odd byte-length
of header to result in a prime-number sized packet :-)

> ron
	Adrian

---------------------------------------------------------------
Adrian Tritschler                          mailto:ajft@ajft.org
Latitude 38°S, Longitude 145°E, Altitude 50m,      Shoe size 44
---------------------------------------------------------------

Re: [9fans] Brdline

[Bruce Ellis]

what cricket is on the tv today?  when do i have to get to pub?
why is life so complex?  ken once commented that a simple
RCA connector was not standardized but everyone agreed
to make them the same.  there are zillions in the world.
maybe it was standandized post factum.

brucee

On 1/19/06, andrey mirtchovski <andrey@lanl.gov> wrote:
> why not, it's wednesday after all and we're pretty far off topic:
>
> ...None of these facts,  however  strange  or  inexplicable,  is  as
> strange  or  inexplicable  as  the  rules of the game of Brockian
> Ultra-Cricket, as played in the higher dimensions. A full set  of
> rules  is  so  massively complicated that the only time they were
> all  bound  together  in  a   single   volume,   they   underwent
> gravitational collapse and became a Black Hole.
>
> A brief summary, however, is as follows:
>
> Rule One: Grow at least three extra legs. You  won't  need  them,
> but it keeps the crowds amused.
>
> Rule Two: Find one good Brockian Ultra-Cricket player. Clone  him
> off  a  few  times.  This  saves  an  enormous  amount of tedious
> selection and training.
>
> Rule Three: Put your team and the opposing team in a large  field
> and build a high wall round them.
>
> The reason for this is that, though the game is a major spectator
> sport,  the  frustration  experienced  by  the  audience  at  not
> actually being able to see what's going on leads them to  imagine
> that it's a lot more exciting than it really is. A crowd that has
> just watched a rather humdrum game  experiences  far  less  life-
> affirmation  than  a  crowd  that believes it has just missed the
> most dramatic event in sporting history.
>
> Rule Four: Throw lots of assorted  items  of  sporting  equipment
> over  the  wall for the players. Anything will do - cricket bats,
> basecube bats, tennis guns, skis, anything you  can  get  a  good
> swing with.
>
> Rule Five: The players should now lay about  themselves  for  all
> they are worth with whatever they find to hand. Whenever a player
> scores a "hit" on another player, he should immediately run  away
> and apologize from a safe distance.
>
> Apologies should be concise, sincere and, for maximum clarity and
> points, delivered through a megaphone.
>
> Rule Six: The winning team shall be the first team that wins.
>
> [and elsewhere]
>
> "Let's be blunt, it's a nasty  game"  (says  The  Hitch  Hiker's
> Guide  to the Galaxy) "but then anyone who has been to any of the
> higher dimensions will know that they're a pretty  nasty  heathen
> lot  up  there  who should just be smashed and done in, and would
> be, too, if anyone could work out a way  of  firing  missiles  at
> right-angles to reality."
>
>
>

Re: [9fans] Brdline

[Joel Salomon]

On 1/18/06, Skip Tavakkolian <9nut@9netics.com> wrote:
> That's nothing compared to the Brokian Ultra-Cricket rule book, that,
> according to the Guide (h2g2), was so massive that it underwent a
> gravitational collapse under its own weight, causing a massive
> blackhole.

In the process of listening to the original radio series — I believe
it's h2g3 that talks of BUC.

> Rule Three: Put your team and the opposing team in a large  field
> and build a high wall round them.
>
> The reason for this is that, though the game is a major spectator
> sport,  the  frustration  experienced  by  the  audience  at  not
> actually being able to see what's going on leads them to  imagine
> that it's a lot more exciting than it really is. A crowd that has
> just watched a rather humdrum game  experiences  far  less  life-
> affirmation  than  a  crowd  that believes it has just missed the
> most dramatic event in sporting history.

I've been lurking on a standard revision committee (IE³ 754R). 
Interesting exposure to numerics, but little "life affirmation".  ;)

--Joel

Re: [9fans] Brdline

[andrey mirtchovski]

why not, it's wednesday after all and we're pretty far off topic:

...None of these facts,  however  strange  or  inexplicable,  is  as
strange  or  inexplicable  as  the  rules of the game of Brockian
Ultra-Cricket, as played in the higher dimensions. A full set  of
rules  is  so  massively complicated that the only time they were
all  bound  together  in  a   single   volume,   they   underwent
gravitational collapse and became a Black Hole.

A brief summary, however, is as follows:

Rule One: Grow at least three extra legs. You  won't  need  them,
but it keeps the crowds amused.

Rule Two: Find one good Brockian Ultra-Cricket player. Clone  him
off  a  few  times.  This  saves  an  enormous  amount of tedious
selection and training.

Rule Three: Put your team and the opposing team in a large  field
and build a high wall round them.

The reason for this is that, though the game is a major spectator
sport,  the  frustration  experienced  by  the  audience  at  not
actually being able to see what's going on leads them to  imagine
that it's a lot more exciting than it really is. A crowd that has
just watched a rather humdrum game  experiences  far  less  life-
affirmation  than  a  crowd  that believes it has just missed the
most dramatic event in sporting history.

Rule Four: Throw lots of assorted  items  of  sporting  equipment
over  the  wall for the players. Anything will do - cricket bats,
basecube bats, tennis guns, skis, anything you  can  get  a  good
swing with.

Rule Five: The players should now lay about  themselves  for  all
they are worth with whatever they find to hand. Whenever a player
scores a "hit" on another player, he should immediately run  away
and apologize from a safe distance.

Apologies should be concise, sincere and, for maximum clarity and
points, delivered through a megaphone.

Rule Six: The winning team shall be the first team that wins.

[and elsewhere]

"Let's be blunt, it's a nasty  game"  (says  The  Hitch  Hiker's
Guide  to the Galaxy) "but then anyone who has been to any of the
higher dimensions will know that they're a pretty  nasty  heathen
lot  up  there  who should just be smashed and done in, and would
be, too, if anyone could work out a way  of  firing  missiles  at
right-angles to reality."

Re: [9fans] Brdline

[Skip Tavakkolian]

That's nothing compared to the Brokian Ultra-Cricket rule book, that,
according to the Guide (h2g2), was so massive that it underwent a
gravitational collapse under its own weight, causing a massive
blackhole.

>>> How about all standards committees advising one individual, the 
>>> standards czar, your brilliant expert, with a background in law and 
>>> social science as well as technology, who is able to apply duly 
>>> constituted public authority to a standard. He/she cannot have any 
>>> alliances with anyone but the ITU.
>>
>> And where will you find any brilliant expert willing to do that job? 
>> It's guaranteed thankless and without innovation.
> 
> Big house with expansive lawn sloping down to Lake Geneva. Lots of 
> ITU-paid servants and readers of papers at your beck and call. Maybe a 
> string quartet too. And a jester to deliver bad news.
> 
>> Nothing is as soul-draining as standards arguments.
> 
> Let the serfs argue. You rule.

Re: [9fans] Brdline

[Wes Kussmaul]

Paul Lalonde wrote:

>> How about all standards committees advising one individual, the 
>> standards czar, your brilliant expert, with a background in law and 
>> social science as well as technology, who is able to apply duly 
>> constituted public authority to a standard. He/she cannot have any 
>> alliances with anyone but the ITU.
>
> And where will you find any brilliant expert willing to do that job? 
> It's guaranteed thankless and without innovation.

Big house with expansive lawn sloping down to Lake Geneva. Lots of 
ITU-paid servants and readers of papers at your beck and call. Maybe a 
string quartet too. And a jester to deliver bad news.

> Nothing is as soul-draining as standards arguments.

Let the serfs argue. You rule.

-- 
Wes Kussmaul
CIO
The Village Group
738 Main Street
Waltham, MA 02451

781-647-7178


My uncle likes to say that the world’s biggest troubles started when the serpent said, “Try this fruit, and by the way if a bunch of people collectively calling themselves Arthur Andersen signs something it’s the same as if a person named Arthur Andersen signed it.” I don’t get the serpent and fruit part. Must be some Swiss mythology thing. He can be a bit obscure. 

                         P.K. Iggy
                         _How I Like Fixed The Internet_
                           (Tales from the Great Infodepression of 2009
                           and the prosperity that followed)

Re: [9fans] Brdline

[Sape Mullender]

> how about someone (or two) experts write the standard?
> worked for K&R.
> 
> brucee

That worked.  The UMTS standard, in contrast, was done by 4000 people
and, trust me, it shows.  I ran a received configuration message (30 bytes or
so) through the ASN-1 decoder and ended up with a 5 megabyte C struct.
Amazing.

	Sape

Re: [9fans] Brdline

[Bruce Ellis]

I think - just do it.  The ozinferno "standard" is defined by the
implementation - plus whatever documentation that i get time
to write.  This does not solve larger issues (like standards
for cell phones) but it works for me.  I didn't go to a meeting
to add function pointers, i just did it.

brucee

On 1/19/06, David Leimbach <leimy2k@gmail.com> wrote:
>
>
> On 1/18/06, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> > how about someone (or two) experts write the standard?
> > worked for K&R.
>
>
> What's worse is standards with no reference implementation. Both C an C99
> seem to have suffered from this disease.
>
> C++ hasn't been implemented as 1998's spec has erm... specified to my
> knowledge without any bugs.  [EDG comes closest and Intel and other compiler
> vendors are just using their front end, and paying for it as a result].
>
> C99 has implementation issues like tgmath.h that are actually impossible to
> implement in just C99.  You absolutely will need compiler extensions to
> implement that header properly.
>
> Perhaps the best way to specify a standard is to define it in a reference
> implementation then talk about it.  Not specify on paper and dream about how
> it should work then find out how far off you were when you start trying to
> prototype it.
>
> Dave
>
> > brucee
> >
> > On 1/19/06, Wes Kussmaul < wes@village.com> wrote:
> > > Paul Lalonde wrote:
> > >
> > > > Standards are for when there are too many cooks in the kitchen. By
> > > > their very nature they have to compromise.
> > > > Give me the work of a standards committee before the the work of a
> > > > single idiot; but most of all give me the work of a brilliant expert
> > > > before that of the committee. And for God's sake, please don't turn my
> > > > expert into an idiot by throwing him onto a committee!
> > >
> > > How about all standards committees advising one individual, the
> > > standards czar, your brilliant expert, with a background in law and
> > > social science as well as technology, who is able to apply duly
> > > constituted public authority to a standard. He/she cannot have any
> > > alliances with anyone but the ITU.
> > >
> > > --
> > > Wes Kussmaul
> > > CIO
> > > The Village Group
> > > 738 Main Street
> > > Waltham, MA 02451
> > >
> > > 781-647-7178
> > >
> > >
> > > My uncle likes to say that the world's biggest troubles started when the
> serpent said, "Try this fruit, and by the way if a bunch of people
> collectively calling themselves Arthur Andersen signs something it's the
> same as if a person named Arthur Andersen signed it." I don't get the
> serpent and fruit part. Must be some Swiss mythology thing. He can be a bit
> obscure.
> > >
> > >                         P.K. Iggy
> > >                         _How I Like Fixed The Internet_
> > >                           (Tales from the Great Infodepression of 2009
> > >                           and the prosperity that followed)
> > >
> > >
> > >
> > >
> >
>
>

Re: [9fans] Brdline

[David Leimbach]

[-- Attachment #1: Type: text/plain, Size: 2557 bytes --]

On 1/18/06, Bruce Ellis <bruce.ellis@gmail.com> wrote:
>
> how about someone (or two) experts write the standard?
> worked for K&R.



What's worse is standards with no reference implementation. Both C an C99
seem to have suffered from this disease.

C++ hasn't been implemented as 1998's spec has erm... specified to my
knowledge without any bugs.  [EDG comes closest and Intel and other compiler
vendors are just using their front end, and paying for it as a result].

C99 has implementation issues like tgmath.h that are actually impossible to
implement in just C99.  You absolutely will need compiler extensions to
implement that header properly.

Perhaps the best way to specify a standard is to define it in a reference
implementation then talk about it.  Not specify on paper and dream about how
it should work then find out how far off you were when you start trying to
prototype it.

Dave

brucee
>
> On 1/19/06, Wes Kussmaul <wes@village.com> wrote:
> > Paul Lalonde wrote:
> >
> > > Standards are for when there are too many cooks in the kitchen. By
> > > their very nature they have to compromise.
> > > Give me the work of a standards committee before the the work of a
> > > single idiot; but most of all give me the work of a brilliant expert
> > > before that of the committee. And for God's sake, please don't turn my
> > > expert into an idiot by throwing him onto a committee!
> >
> > How about all standards committees advising one individual, the
> > standards czar, your brilliant expert, with a background in law and
> > social science as well as technology, who is able to apply duly
> > constituted public authority to a standard. He/she cannot have any
> > alliances with anyone but the ITU.
> >
> > --
> > Wes Kussmaul
> > CIO
> > The Village Group
> > 738 Main Street
> > Waltham, MA 02451
> >
> > 781-647-7178
> >
> >
> > My uncle likes to say that the world's biggest troubles started when the
> serpent said, "Try this fruit, and by the way if a bunch of people
> collectively calling themselves Arthur Andersen signs something it's the
> same as if a person named Arthur Andersen signed it." I don't get the
> serpent and fruit part. Must be some Swiss mythology thing. He can be a bit
> obscure.
> >
> >                         P.K. Iggy
> >                         _How I Like Fixed The Internet_
> >                           (Tales from the Great Infodepression of 2009
> >                           and the prosperity that followed)
> >
> >
> >
> >
>

[-- Attachment #2: Type: text/html, Size: 3698 bytes --]

Re: [9fans] Brdline

[Charles Forsyth]

> Nothing is as  soul-draining as standards arguments.

in the mid 80s i was on just one committee for a short time and when i left
my dept i put the resulting many big boxes of papers in the dept library
as a Warning to later generations.  it was originally only intended to
add four or five simple things (one type and a few functions) to an
existing language (perhaps an afternoon to implement).
it didn't end up that way.

Re: [9fans] Brdline

[Paul Lalonde]

>
> How about all standards committees advising one individual, the  
> standards czar, your brilliant expert, with a background in law and  
> social science as well as technology, who is able to apply duly  
> constituted public authority to a standard. He/she cannot have any  
> alliances with anyone but the ITU.

And where will you find any brilliant expert willing to do that job?   
It's guaranteed thankless and without innovation.
You need a brilliant domain expert, and most of those are very much  
interested in their work, not in farting around being advised by  
committees.
My last employer killed my sense of worth by pilling me on standards  
committees, evaluation committees, coffee-cup-washing committees ad  
infinitum.  I may not be a brilliant expert, but they put me there as  
an expert, and killed the expertise simultaneously.  Nothing is as  
soul-draining as standards arguments.

Paul

Re: [9fans] Brdline

[Bruce Ellis]

how about someone (or two) experts write the standard?
worked for K&R.

brucee

On 1/19/06, Wes Kussmaul <wes@village.com> wrote:
> Paul Lalonde wrote:
>
> > Standards are for when there are too many cooks in the kitchen. By
> > their very nature they have to compromise.
> > Give me the work of a standards committee before the the work of a
> > single idiot; but most of all give me the work of a brilliant expert
> > before that of the committee. And for God's sake, please don't turn my
> > expert into an idiot by throwing him onto a committee!
>
> How about all standards committees advising one individual, the
> standards czar, your brilliant expert, with a background in law and
> social science as well as technology, who is able to apply duly
> constituted public authority to a standard. He/she cannot have any
> alliances with anyone but the ITU.
>
> --
> Wes Kussmaul
> CIO
> The Village Group
> 738 Main Street
> Waltham, MA 02451
>
> 781-647-7178
>
>
> My uncle likes to say that the world's biggest troubles started when the serpent said, "Try this fruit, and by the way if a bunch of people collectively calling themselves Arthur Andersen signs something it's the same as if a person named Arthur Andersen signed it." I don't get the serpent and fruit part. Must be some Swiss mythology thing. He can be a bit obscure.
>
>                         P.K. Iggy
>                         _How I Like Fixed The Internet_
>                           (Tales from the Great Infodepression of 2009
>                           and the prosperity that followed)
>
>
>
>

Re: [9fans] Brdline

[Wes Kussmaul]

Paul Lalonde wrote:

> Standards are for when there are too many cooks in the kitchen. By 
> their very nature they have to compromise.
> Give me the work of a standards committee before the the work of a 
> single idiot; but most of all give me the work of a brilliant expert 
> before that of the committee. And for God's sake, please don't turn my 
> expert into an idiot by throwing him onto a committee!

How about all standards committees advising one individual, the 
standards czar, your brilliant expert, with a background in law and 
social science as well as technology, who is able to apply duly 
constituted public authority to a standard. He/she cannot have any 
alliances with anyone but the ITU.

-- 
Wes Kussmaul
CIO
The Village Group
738 Main Street
Waltham, MA 02451

781-647-7178


My uncle likes to say that the world’s biggest troubles started when the serpent said, “Try this fruit, and by the way if a bunch of people collectively calling themselves Arthur Andersen signs something it’s the same as if a person named Arthur Andersen signed it.” I don’t get the serpent and fruit part. Must be some Swiss mythology thing. He can be a bit obscure. 

                         P.K. Iggy
                         _How I Like Fixed The Internet_
                           (Tales from the Great Infodepression of 2009
                           and the prosperity that followed)

Re: [9fans] Brdline

[Bruce Ellis]

well put sir, and i have a small kitchen.

brucee

On 1/19/06, Paul Lalonde <plalonde@telus.net> wrote:
> Standards are for when there are too many cooks in the kitchen.  By
> their very nature they have to compromise.
> Give me the work of a standards committee before the the work of a
> single idiot; but most of all give me the work of a brilliant expert
> before that of the committee.  And for God's sake, please don't turn
> my expert into an idiot by throwing him onto a committee!
>
> Paul
>
> On 17-Jan-06, at 10:11 PM, Bruce Ellis wrote:
>
> > standards are weird.  i wish someone would standardize standards.
> > oh no that would involve a committee of uninformed experts.
> >
> > i have never had a problem with kenc, both for user and kernel stuff.
> > but i don't use putchar() or some other recalcitrant macro.
> >
> > the proof is in the pudding.
> >
> > brucee
> >
> > On 1/18/06, Paul Lalonde <plalonde@telus.net> wrote:
> >>
> >> On 17-Jan-06, at 9:38 PM, Simon Williams wrote:
> >>
> >>> The nice thing about standards is that there are so many to choose
> >>> from
> >>> Cheers
> >>>        Simon ( who cant remember who said this first )
> >>
> >> Andy Tannenbaum, wasn't it?
> >>
> >> Paul
> >>
> >>
>
>

Re: [9fans] Brdline

[Paul Lalonde]

:-)  Must learn to proof my spelling of names in languages I don't  
speak :-)

On 18-Jan-06, at 7:25 AM, Brantley Coile wrote:

>> Andy Tannenbaum, wasn't it?
>
> No.  It was Andy Tanenbaum.  Not Andy Tannenbaum.  That's a  
> different fellow.
>> From his book, Computer Networks.
>

Re: [9fans] Brdline

[Paul Lalonde]

Standards are for when there are too many cooks in the kitchen.  By  
their very nature they have to compromise.
Give me the work of a standards committee before the the work of a  
single idiot; but most of all give me the work of a brilliant expert  
before that of the committee.  And for God's sake, please don't turn  
my expert into an idiot by throwing him onto a committee!

Paul

On 17-Jan-06, at 10:11 PM, Bruce Ellis wrote:

> standards are weird.  i wish someone would standardize standards.
> oh no that would involve a committee of uninformed experts.
>
> i have never had a problem with kenc, both for user and kernel stuff.
> but i don't use putchar() or some other recalcitrant macro.
>
> the proof is in the pudding.
>
> brucee
>
> On 1/18/06, Paul Lalonde <plalonde@telus.net> wrote:
>>
>> On 17-Jan-06, at 9:38 PM, Simon Williams wrote:
>>
>>> The nice thing about standards is that there are so many to choose
>>> from
>>> Cheers
>>>        Simon ( who cant remember who said this first )
>>
>> Andy Tannenbaum, wasn't it?
>>
>> Paul
>>
>>

Re: [9fans] Brdline

[Ronald G Minnich]

read "the art of networking style" for a great rant on standards.

Note that film standards are ISO/DIN numbers -- not one number, two. 
That's how they resolved how to pick the #.

48 byte cells came from standards.

ron

Re: [9fans] Brdline

[Brantley Coile]

> Andy Tannenbaum, wasn't it?

No.  It was Andy Tanenbaum.  Not Andy Tannenbaum.  That's a different fellow.
>From his book, Computer Networks.

Re: [9fans] Brdline

[Bruce Ellis]

ahhh - just woken so here is a mildly related rant while the
coffee revs up.

standards?  when i was staying with skip at Club Vashon
we had a power outage (some bloody tree falling) and when
the candles and flashlights started to fail i trundled upstairs
and grabbed by video camera which has night vision, spotlight
and a 5 hour battery (footage of target - the wonder dog - catching
frisbees in the dark will appear).

so the next day i go to costco with pat ('cause my luggage was
lost and then busted) and the first thing in the door was a huge
flashlight.  brucee thinks 10 million candle-power.  10 million of
anything must be good and indeed it has been used more than
recreationally, to help stranded souls on the sound.  so i grabbed
it as a present (of course she had cooked me a good breakfast).

but back to the point.  i was at "bunnings" the other day (well
actually most days because i'm in renovation hell) and the same
flashlight was there - but repackaged (because candlepower is
not SI and not legal in australia) - bloody lumens.

then i bought a water blaster (good toy) and it's clearly marked
in PSI (pounds per square inch) - not very SI.

so there are rules and standards but they just get broken.

stay safe, throw the frisbee.

brucee

On 1/18/06, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> standards are weird.  i wish someone would standardize standards.
> oh no that would involve a committee of uninformed experts.
>
> i have never had a problem with kenc, both for user and kernel stuff.
> but i don't use putchar() or some other recalcitrant macro.
>
> the proof is in the pudding.
>
> brucee
>
> On 1/18/06, Paul Lalonde <plalonde@telus.net> wrote:
> >
> > On 17-Jan-06, at 9:38 PM, Simon Williams wrote:
> >
> > > The nice thing about standards is that there are so many to choose
> > > from
> > > Cheers
> > >        Simon ( who cant remember who said this first )
> >
> > Andy Tannenbaum, wasn't it?
> >
> > Paul
> >
> >
>

Re: [9fans] Brdline

[Bruce Ellis]

standards are weird.  i wish someone would standardize standards.
oh no that would involve a committee of uninformed experts.

i have never had a problem with kenc, both for user and kernel stuff.
but i don't use putchar() or some other recalcitrant macro.

the proof is in the pudding.

brucee

On 1/18/06, Paul Lalonde <plalonde@telus.net> wrote:
>
> On 17-Jan-06, at 9:38 PM, Simon Williams wrote:
>
> > The nice thing about standards is that there are so many to choose
> > from
> > Cheers
> >        Simon ( who cant remember who said this first )
>
> Andy Tannenbaum, wasn't it?
>
> Paul
>
>

Re: [9fans] Brdline

[Paul Lalonde]

On 17-Jan-06, at 9:38 PM, Simon Williams wrote:

> The nice thing about standards is that there are so many to choose  
> from
> Cheers
>        Simon ( who cant remember who said this first )

Andy Tannenbaum, wasn't it?

Paul

Re: [9fans] Brdline

[Bruce Ellis]

it was either ken or chez if my memory survives.

brucee

On 1/18/06, Simon Williams <wyman.silliams@gmail.com> wrote:
> The nice thing about standards is that there are so many to choose from
> Cheers
>        Simon ( who cant remember who said this first )
>
>
> On 1/18/06, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> > wait for the next standard.  confusion will reign supreme!
> >
> > brucee
> >
> > On 1/17/06, Charles Forsyth < forsyth@terzarima.net> wrote:
> > > > i'll be wearing the dope sack over my head this week.
> > >
> > > to be fair, that honour should be reserved for ansi.
> > > they took a confusing thing and made it even more confusing.
> >
>
>

Re: [9fans] Brdline

[Simon Williams]

[-- Attachment #1: Type: text/plain, Size: 531 bytes --]

The nice thing about standards is that there are so many to choose from
Cheers
       Simon ( who cant remember who said this first )

On 1/18/06, Bruce Ellis <bruce.ellis@gmail.com> wrote:
>
> wait for the next standard.  confusion will reign supreme!
>
> brucee
>
> On 1/17/06, Charles Forsyth <forsyth@terzarima.net> wrote:
> > > i'll be wearing the dope sack over my head this week.
> >
> > to be fair, that honour should be reserved for ansi.
> > they took a confusing thing and made it even more confusing.
>

[-- Attachment #2: Type: text/html, Size: 920 bytes --]

Re: [9fans] Brdline

[Bruce Ellis]

wait for the next standard.  confusion will reign supreme!

brucee

On 1/17/06, Charles Forsyth <forsyth@terzarima.net> wrote:
> > i'll be wearing the dope sack over my head this week.
>
> to be fair, that honour should be reserved for ansi.
> they took a confusing thing and made it even more confusing.

Re: [9fans] Brdline

[Charles Forsyth]

> i'll be wearing the dope sack over my head this week.

to be fair, that honour should be reserved for ansi.
they took a confusing thing and made it even more confusing.

Re: [9fans] Brdline

[Russ Cox]

> | but that won't actually have any effect on systems where
> | int is 32 bits but long is 64.  the problem is that ((p)[3]<<24)
> | is being (correctly) treated as a signed int, and then
> | (vlong)(...|((p)[3]<<24)) sign extends.  Casting the (p)[0]
> | to (ulong) has the effect of making the whole 32-bit expression
> | unsigned on 32-bit systems, but if ulong is 64 bits, then
> | you'll still sign-extend ((p)[3]<<24) during the convertsion
> | from int to ulong.
>
> i'll be wearing the dope sack over my head this week.

this took me forever to puzzle through, by the way.
the only reason i'm pretty sure it's true is that i tried it!  ;-)

russ

Re: [9fans] Brdline

[erik quanstrom]

Russ Cox <rsc@swtch.com> writes

| but that won't actually have any effect on systems where
| int is 32 bits but long is 64.  the problem is that ((p)[3]<<24)
| is being (correctly) treated as a signed int, and then
| (vlong)(...|((p)[3]<<24)) sign extends.  Casting the (p)[0]
| to (ulong) has the effect of making the whole 32-bit expression
| unsigned on 32-bit systems, but if ulong is 64 bits, then
| you'll still sign-extend ((p)[3]<<24) during the convertsion
| from int to ulong.

i'll be wearing the dope sack over my head this week.

- erik

Re: [9fans] Brdline

[Russ Cox]

> likewise, seek takes a vlong "where"so that a -1 "don't care" value can be used.
> this was the source of the sign-extension issue p9p's GBIT64 macro.

seek takes a vlong so that you can seek backwards in a file.
if you replace seek with pread, then you're right, except that -1 doesn't
at all mean "don't care".

> this was the source of the sign-extension issue p9p's GBIT64 macro.

it wasn't.  the original GBIT64 says:

#define	GBIT64(p)	((vlong)((p)[0]|...|((p)[3]<<24)) |\
				((vlong)((p)[4]|...|((p)[7]<<24)) << 32))

and you suggested:

#define	GBIT64(p)	((vlong)((ulong)(p)[0]|...|((p)[3]<<24)) |\
				((vlong)((ulong)(p)[4]|...|((p)[7]<<24)) << 32))

but that won't actually have any effect on systems where
int is 32 bits but long is 64.  the problem is that ((p)[3]<<24)
is being (correctly) treated as a signed int, and then
(vlong)(...|((p)[3]<<24)) sign extends.  Casting the (p)[0]
to (ulong) has the effect of making the whole 32-bit expression
unsigned on 32-bit systems, but if ulong is 64 bits, then
you'll still sign-extend ((p)[3]<<24) during the convertsion
from int to ulong.

the only way i see to fix this is to explicitly cast away the
top bits:

#define	GBIT64(p)	((u32int)((p)[0]|...|((p)[3]<<24)) |\
				((vlong)((p)[4]|...|((p)[7]<<24)) << 32))

this is now fixed on sources and cvs.
russ

Re: [9fans] Brdline

[erik quanstrom]

likewise, seek takes a vlong "where"so that a -1 "don't care" value can be used.
this was the source of the sign-extension issue p9p's GBIT64 macro.

if you know you're stuck with gcc, uvlong and ~0ULL may well have be a less error-
prone option.

- erik

Charles Forsyth <forsyth@terzarima.net> writes

| 
| > What about Brdstr? it seems to me a superset of Brdline and it returns
| > a char *....
| 
| perhaps it was added later and they didn't think of that?
| i'm not sure it makes a big difference.
| the whole char*/uchar* interaction is bad though.
| uchar* is important to ensure no sign-extension,
| but it isn't compatible with the str* functions, and explicit casts
| can mask mistakes.

Re: [9fans] Brdline

[Bruce Ellis]

i believe that is why limbo just has "byte" which is unsigned,
no options.  get's those bugs out before they get in.

brucee

On 1/17/06, Charles Forsyth <forsyth@terzarima.net> wrote:
> > What about Brdstr? it seems to me a superset of Brdline and it returns
> > a char *....
>
> perhaps it was added later and they didn't think of that?
> i'm not sure it makes a big difference.
> the whole char*/uchar* interaction is bad though.
> uchar* is important to ensure no sign-extension,
> but it isn't compatible with the str* functions, and explicit casts
> can mask mistakes.

Re: [9fans] Brdline

[Charles Forsyth]

> What about Brdstr? it seems to me a superset of Brdline and it returns
> a char *....

perhaps it was added later and they didn't think of that?
i'm not sure it makes a big difference.
the whole char*/uchar* interaction is bad though.
uchar* is important to ensure no sign-extension,
but it isn't compatible with the str* functions, and explicit casts
can mask mistakes.

Re: [9fans] Brdline

[Gorka guardiola]

I am still not happy with the explanation of it giving something you may
want to converto to uchar *.
What about Brdstr? it seems to me a superset of Brdline and it returns
a char *....

On 1/16/06, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> beware that Brdline gives you a pointer to stuff that might change
> when you do another Bio act.  it has bitten me.  use or save!
>
I already expected that to happen, after all it is a pointer to the buffer...

--
- curiosity sKilled the cat

Re: [9fans] Brdline

[Bruce Ellis]

beware that Brdline gives you a pointer to stuff that might change
when you do another Bio act.  it has bitten me.  use or save!

brucee

On 1/16/06, Charles Forsyth <forsyth@terzarima.net> wrote:
> >Is there any reason why Brdline, bio(2) returns void * instead of char *
>
> perhaps to allow it to be assigned to char* or uchar*

Re: [9fans] Brdline

[Charles Forsyth]

>Is there any reason why Brdline, bio(2) returns void * instead of char *

perhaps to allow it to be assigned to char* or uchar*

[9fans] Brdline

[Gorka guardiola]

Is there any reason why Brdline, bio(2) returns void * instead of char *
which I think it should return?. Just curious.
--
- curiosity sKilled the cat