denigrate people if they do not disagree with you.

"Richard Heathfield" <dontmail@address.co.uk.invalid> wrote in message
news:bsjfmf$fjh$1@titan.btinternet.com...
Tom St Denis wrote:

In the context of this thread and the OP
recommending textbooks when freely available useful resources exist is
just stupid IMO.

In the context of this thread, then, you think it stupid to recommend
Knuth?

Actually, yes I do.

I find it quite hard to believe that you believe this, given that Knuth
is just about the most solid and comprehensive computer science reference
work
available, bar none.

This is not the issue. The issue is whether spending 65$ to learn MP
division is a good idea.

I say no. But that's just me.

Sure textbooks are useful [I own a few dozen on various subjects]. I'd
just
rather see an emerging trend of openly available resources. Call me
the idealist.

If a free alternative to Knuth that is of the same quality were to exist,
fine, we'd all like a copy. It doesn't follow that recommending Knuth is
a bad idea.

This again doesn't follow. The OP wanted to learn MP division [as an
aside I bet the OP has already learned what he wants so us bickering about
this
achieves nothing].

Buying a textbook which covers other things just to
learn one Algorithm D is not smart.

If I recall the OP was to implement the division in a program. Hmm, well
my textbook is called, oh, um... "Implementing Multiple Precision
Arithmetic". Not only do I cover MP division in abstract, pseudo-code and
source code but I also do so for free.

You may not like me, whatever,

but the reference does exist. Saying the
OP shouldn't use it because I'm some talkative newbie is just plain
arrogant.

I just like calling him on stuff like that
to see if he will ever support anything he writes.

I don't think it should be necessary to support a statement like "read
Knuth". Those who have a copy of Knuth (or who have read it in the
library)
already know that it's the best book (or, rather, set of books!) on CS in
existence, and those who haven't are not qualified to offer an opinion.

Actually I would. Knuth is great for theory. It sucks royal ass if you
have
to implement it. In fact I used HAC as my reference for division for a
reason. And I'm sorry but MIX hardly translates easily to C source
code...

"Richard Heathfield" <dontmail@address.co.uk.invalid> wrote in message
news:bskb5t$b5$1@hercules.btinternet.com...
I find it quite hard to believe that you believe this, given that
Knuth is just about the most solid and comprehensive computer science
reference
work
available, bar none.

This is not the issue. The issue is whether spending 65$ to learn MP
division is a good idea.

But on the other hand, seeking the best information source /is/ a good
idea.

Well by this argument why not suggest the OP get a full PhD in EECS? That
would be the best course of action in this case.

And anyway, Knuth is available in any decent public library.

Um maybe university library. The Nepean city library [closest to my
house] not only doesn't carry journals but doesn't carry CS books [that
are actually useful and not crap for dummies] from 1970 on.

To be fair I agree that Knuth is a good resource but a lot of what it
covers
is available elsewhere for free. The benefit of Knuth is a common writing
style that covers a lot of different topics in a few texts. Knuth still
doesn't cover quite a bit that you need in the real world [e.g. realtime
programming, DSP theory, coding theory, etc...]

And are you offering $2.56 per error found? Are you that sure that your
information is correct? If not, then you are in no position to argue that
your stuff is as good as Knuth's stuff. No, I don't offer $2.56 per error
either, so I'm not trying to claim your stuff is valueless. I'm just
pointing out that Knuth's reputation is extremely high, and justifiably
so.

Again, nothing to disagree with here. I'm just trying to point out that a
free resource which was written EXACTLY FOR THIS INSTANCE is available.
Knuth is more theory than practice. My book is more practice than theory.

For the record though, I do welcome all bug reports and I give credit for
those who find them [except in one case where I forgot but post-credited
the person].

And you think that is useless?

Tom St Denis wrote:

I never said there aren't smarter people then me around. I could name a
half dozen off the top of my head. Watch. Contini, Fisher, Fluhrer,
Rose,
Rosing and Wagner. [and those came out in alphabetical order... figure
that out!].

You missed a couple of obvious contenders: Doug Gwyn and Donald Knuth.

Tom St Denis wrote:

In the context of this thread and the OP
recommending textbooks when freely available useful resources exist is
just stupid IMO.

In the context of this thread, then, you think it stupid to recommend
Knuth?

I find it quite hard to believe that you believe this, given that Knuth is
just about the most solid and comprehensive computer science reference
work
available, bar none.

Sure textbooks are useful [I own a few dozen on various subjects]. I'd
just
rather see an emerging trend of openly available resources. Call me the
idealist.

If a free alternative to Knuth that is of the same quality were to exist,
fine, we'd all like a copy. It doesn't follow that recommending Knuth is a
bad idea.

I just like calling him on stuff like that
to see if he will ever support anything he writes.

I don't think it should be necessary to support a statement like "read
Knuth". Those who have a copy of Knuth (or who have read it in the
library)
already know that it's the best book (or, rather, set of books!) on CS in
existence, and those who haven't are not qualified to offer an opinion.

No book is complete for all cases of all readers. I've never
encountered
anyone claiming that Knuth was not worth buying until now. Either way,
you simply cannot let anyone entertain the notion of using a resource
that has not been recommended by you. Why is that exactly?

I didn't say TAOCP is a bad investment. I was simply calling Gwyn on his
suggestion. If he thinks his suggestion was a good one why not back it
up?

E.g. show some flaw in the open sources I quoted that is not in TAOCP.
[Again this goes back to how Gwyn behaves in the group].

I find it quite hard to believe that you believe this, given that Knuth
is just about the most solid and comprehensive computer science
reference
work
available, bar none.

This is not the issue. The issue is whether spending 65$ to learn MP
division is a good idea.

But on the other hand, seeking the best information source /is/ a good
idea.

And anyway, Knuth is available in any decent public library.

This again doesn't follow. The OP wanted to learn MP division [as an
aside I bet the OP has already learned what he wants so us bickering
about
this
achieves nothing].

The discussion may no longer be relevant to the OP. That doesn't mean it
isn't worth having, if it clarifies how people can best be helped by this
newsgroup.

Buying a textbook which covers other things just to
learn one Algorithm D is not smart.

(a) Nobody said he has to buy it; library copies contain precisely the
same
text as bookshop copies.
(b) If he /does/ decide to buy it, he almost certainly won't regret it.
Not
if he's serious about programming, anyway.

If I recall the OP was to implement the division in a program. Hmm,
well
my textbook is called, oh, um... "Implementing Multiple Precision
Arithmetic". Not only do I cover MP division in abstract, pseudo-code
and
source code but I also do so for free.

And are you offering $2.56 per error found? Are you that sure that your
information is correct? If not, then you are in no position to argue that
your stuff is as good as Knuth's stuff. No, I don't offer $2.56 per error
either, so I'm not trying to claim your stuff is valueless. I'm just
pointing out that Knuth's reputation is extremely high, and justifiably
so.

Actually I would. Knuth is great for theory. It sucks royal ass if you
have
to implement it. In fact I used HAC as my reference for division for a
reason. And I'm sorry but MIX hardly translates easily to C source
code...

I sympathise with this view to some extent, because Knuth can be rather
terse; but on the other hand, I found Knuth's explanation of Miller-Rabin
to be far clearer than either Schneier or anything I could find on the
Web.

This is not the issue. The issue is whether spending 65$ to learn MP
division is a good idea.

But on the other hand, seeking the best information source /is/ a good
idea.

Well by this argument why not suggest the OP get a full PhD in EECS? That
would be the best course of action in this case.

The Nepean city library [closest to my house] not only doesn't carry
journals but doesn't carry CS books [that are actually useful and not
crap for dummies] from 1970 on.

Somehow I don't think that is coming up. You guys disagree with me
questioning Gwyn. Gwyn doesn't care [he probably has me killfiled] so why
do you guys care? Let it be. I never said he was wrong.

The benefit of Knuth is a common writing
style that covers a lot of different topics in a few texts. Knuth still
doesn't cover quite a bit that you need in the real world [e.g. realtime
programming, DSP theory, coding theory, etc...]

Again, nothing to disagree with here. I'm just trying to point out that a
free resource which was written EXACTLY FOR THIS INSTANCE is available.
Knuth is more theory than practice. My book is more practice than theory.

"Richard Heathfield" <dontmail@address.co.uk.invalid> wrote in message
news:bsjet1$ec8$2@titan.btinternet.com...
Tom St Denis wrote:

I never said there aren't smarter people then me around. I could name a
half dozen off the top of my head. Watch. Contini, Fisher, Fluhrer,
Rose,
Rosing and Wagner. [and those came out in alphabetical order... figure
that out!].

You missed a couple of obvious contenders: Doug Gwyn and Donald Knuth.

I have yet to have Gwyn teach me anything other than citing published
authors is not considered valid.

Knuth never emailed me about anything
neither.

All six people I named have at one point taught me something out
of their own spare time.

I'd say if Greg Rose came in here talking smack about me that would carry
more weight than some two-bit loser nobody that dropped out of the wind to
post some flame in usenet [Sorry Greg, I know you'd never do that, just
trying to make an example]

I don't disparage people.

I may discourage them from acting like f'ing
retards but that's totally different.

If I wanted to put people down I wouldn't reply to MKS the way I would do.

I don't even know. I know that the University of Ottawa has public access
but I don't know if they let anyone off the street without buying an id
first [last I checked it was cheap but that's not the point].

Tom St Denis wrote:

And anyway, Knuth is available in any decent public library.

Um maybe university library. The Nepean city library [closest to my
house] not only doesn't carry journals but doesn't carry CS books [that
are actually useful and not crap for dummies] from 1970 on.

Not even by order? I had no idea the library system was so bad in Canada. In
the UK, it's generally possible to order a book if the library doesn't
carry its own copy.

Everyone is entitled to an education, not just those willing to be $50K in
debt to go into the workplace and earn $28K a year... [or whatever the
situation].

He doesn't [and afaik hasn't]. I just like calling him on stuff like that
to see if he will ever support anything he writes.

[For reference, see older arguments involving padding rounds in block
ciphers. I cited numerous published well respected cryptographers, gwyn
simply said I was wrong without making an argument].