“Andrew Thomas” <andrew@cogent.ca> wrote in message
news:ad1chu$d9t$1@inn.qnx.com…
baby-sitting androids? The pie-in-the-sky predictions of the money-hungry
and none too honest university researchers really stained the legitimate
AI
community.
Sort of like cold fusion. There are a few researchers who are still looking
at it and think it might be possible but it got such a bad name that no-one
wants to be associated with it.
Kris
Kris Warkentin <kewarken@qnx.com> wrote:
“Robert Krten” <> nospam88@parse.com> > wrote in message
news:ad2qfs$fvd$> 1@inn.qnx.com> …
Richard Kramer <> rrkramer@kramer-smilko.com> > wrote:
Actually, one of the next projects I was considering is a “resource
manager
builder”. The idea is that it would be a CGI-BIN kind of thing that sits
on my website, and you fill in forms about what kind of resource manager
you want (e.g., “single/multi threaded”, or “directory vs file”, etc).
Once you’ve filled out all that stuff, it generates code, and gives you a
clicky for a .tar.gz that contains the code. It would even generate
prototype-style code for your I/O and Connect functions, maybe generate
your
ISRs, that kind of thing…That’s kind of funny. I had an idea for something like that (not nearly as
complicated though) for us to provide demo’s to people. You would go to our
website, fill in a form which had information about your board, (graphics,
networking, cpu, etc.) and then when you clicked the button, it would build
an image and let you download it to test on your board. Dunno how hard it
would be but it would be a cool way for people to do a quick proof of
concept.
That’s a good concept, but in your case you need something that’s 100%
accurate and working; for mine, I can get away with stuff like “// add your code here” 
Cheers,
-RK
–
Robert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at www.parse.com.
Email my initials at parse dot com.
“Robert Krten” <nospam88@parse.com> wrote in message
news:ad2qfs$fvd$1@inn.qnx.com…
Richard Kramer <> rrkramer@kramer-smilko.com> > wrote:
Actually, one of the next projects I was considering is a “resource
manager
builder”. The idea is that it would be a CGI-BIN kind of thing that sits
on my website, and you fill in forms about what kind of resource manager
you want (e.g., “single/multi threaded”, or “directory vs file”, etc).
Once you’ve filled out all that stuff, it generates code, and gives you a
clicky for a .tar.gz that contains the code. It would even generate
prototype-style code for your I/O and Connect functions, maybe generate
your
ISRs, that kind of thing…
That’s kind of funny. I had an idea for something like that (not nearly as
complicated though) for us to provide demo’s to people. You would go to our
website, fill in a form which had information about your board, (graphics,
networking, cpu, etc.) and then when you clicked the button, it would build
an image and let you download it to test on your board. Dunno how hard it
would be but it would be a cool way for people to do a quick proof of
concept.
Kris
“Robert Krten” <nospam88@parse.com> wrote in message
news:ad2rie$h6c$1@inn.qnx.com…
That’s a good concept, but in your case you need something that’s 100%
accurate and working; for mine, I can get away with stuff like “// add
your code here” >
I know…it’s always so much easier to HAVE and idea than to implement it.
Kris
Cheers,
-RK–
Robert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at > www.parse.com> .
Email my initials at parse dot com.
Kris Warkentin <kewarken@qnx.com> wrote:
“Robert Krten” <> nospam88@parse.com> > wrote in message
news:ad2rie$h6c$> 1@inn.qnx.com> …
That’s a good concept, but in your case you need something that’s 100%
accurate and working; for mine, I can get away with stuff like “// add
your code here” >
I think next time you open “eclipse”, it will have “templates”, with
like “Resrouce Manager”, “Network driver”, “Stand alone application”,
and you choose one, maybe answer some question (a wizard?), it generates
all the necessary code for you. With “// add your code here” of cause
So, Rob, maybe you should consider it “has been done”
-xtang
I know…it’s always so much easier to HAVE and idea than to implement it.
Kris
Cheers,
-RK–
Robert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at > www.parse.com> .
Email my initials at parse dot com.
On 29 May 2002 15:01:16 GMT, nospam88@parse.com (Robert Krten) wrote:
…
Actually, one of the next projects I was considering is a “resource manager
builder”. The idea is that it would be a CGI-BIN kind of thing that sits
on my website, and you fill in forms about what kind of resource manager
you want (e.g., “single/multi threaded”, or “directory vs file”, etc).
Once you’ve filled out all that stuff, it generates code, and gives you a
clicky for a .tar.gz that contains the code. It would even generate
prototype-style code for your I/O and Connect functions, maybe generate your
ISRs, that kind of thing…
Speaking of things to be done-I think a graphical embedded system builder
will also be useful.
e.g. Pick and click interface for utilities & options to go into the image build file.
Sorta capturing the “Building embedded systems” book in s/w…
“Robert Krten” <nospam88@parse.com> wrote in message
news:ad0his$p52$1@inn.qnx.com…
The point of this tirade (and I did warn you-all it would be one >
is that I’m really kinda bored – what’s going to be
the “next” exciting thing in the software field that you can hold in awe
and terror (like OS’s and compilers used to be)?Implement the DWIM machine instruction.
About the only thing that jumps out at me is the field of AI – and this
is
because of two things: my own general ignorance of the state-of-the-art
in
the field, and my perception that it’s all just the same old crap – do
they
still use LISP? Are they any closer to making a truly artificial
intelligence?
Well, I programmed for 13 years on Wall St. We actually had a system that
DID put out real-time buy and sell recommendations and DID make you money IF
you added only a tiny amount of intelligence yourself. And this was 15
years ago.
First, lots of people wrote software to tell you when to take a position and
most of it is right most of the time. The first problem is that just
because the stock last traded at X or that the bid/ask is X doesn’t mean
that you can buy it at X. The difference between where the software gets
excited over a stock and where you could actually go out and buy it is often
the difference between profit and loss.
The second problem is that once you take a position (long or short) you also
need to decide when to get out of it. (Either take a profit or admit your
wrong and take a small loss rather than a large loss.) Our system (the
company is now out of business [STOP LAUGHING!]) had the best getting out
logic of any system out there always maximizing profit.
But the real problem that required human intervention is that all of the
data was originally typed in by hard at the stock exchanges and typos were
extremely common. If a single trade came down the pike that made the stock
look interesting someone had call someone verbally and confirm that the
trade was real and valid. It was quite likely that all validation
indicators made the trade look valid when it wasn’t. Also, many trades that
were valid fell outside of the specs of validity because of slight delays in
the reporting of the trades. Now that a MUCH higher percentage of trades
are completely automated (no human intervention at all) I assume that the
reliability of the data should be much better. And Oh, BTW, many of the
wrong trades were actually sent down the line on purpose by the specialists.
They aren’t allowed to do that but they did it a hundred times a day.
Still, the stock market is the ultimate AI application. Plug the computer
in and let it print money.
And for the record, the reason that company went out of business was because
the bastard that ran the company (my old boss) was a corrupt thief and
robbed and stole from his own investors (who were his friends) till there
wasn’t a dime left in the company and they wouldn’t support him anymore.
“Kris Warkentin” <kewarken@qnx.com> wrote in message
news:ad0svq$ogm$1@nntp.qnx.com…
I remember that bit about four lines but I believe it was four lines of
CODE, not a four line equation
Did his compiler have a limit on line length?
“Robert Krten” <nospam88@parse.com> wrote in message
news:ad2p8v$f8t$1@inn.qnx.com…
I guess my “practical definition” of a particular class of “useful”
AI/MI would be a language translator that did better than the horror
stories you
see on the web where some English->X->English translation of “my mother
gave me lunch” comes out as “the ate princess radish green wedding” >I haven’t heard that one, but, …
20-30 years ago they had a supposedly smart English → Russian → English
translator that translated the biblical phrase “The spirit is willing but
the flesh is weak.” to Russian and back to English and it came out “The wine
is good but the meat has spoiled.”
“Kris Warkentin” <kewarken@qnx.com> wrote in message
news:ad2sam$aqt$1@nntp.qnx.com…
“Robert Krten” <> nospam88@parse.com> > wrote in message
news:ad2rie$h6c$> 1@inn.qnx.com> …
That’s a good concept, but in your case you need something that’s 100%
accurate and working; for mine, I can get away with stuff like “// add
your code here” >I know…it’s always so much easier to HAVE and idea than to implement
it.I disagree. Give me legs and I’ll make a computer dance like Fred Astair.
I can implement any well defined project. And if it’s not well defined I
will define it and tell you the pros, cons and alternatives to each
decision. But my problem is being creative. I’m lousy at coming up with
ideas.
“Bill Caroselli (Q-TPS)” <QTPS@EarthLink.net> wrote in message
news:ad3llo$5v8$1@inn.qnx.com…
“Kris Warkentin” <> kewarken@qnx.com> > wrote in message
news:ad2sam$aqt$> 1@nntp.qnx.com> …
“Robert Krten” <> nospam88@parse.com> > wrote in message
news:ad2rie$h6c$> 1@inn.qnx.com> …
That’s a good concept, but in your case you need something that’s 100%
accurate and working; for mine, I can get away with stuff like “// add
your code here” >I know…it’s always so much easier to HAVE and idea than to implement
it.I disagree. Give me legs and I’ll make a computer dance like Fred Astair.
I can implement any well defined project. And if it’s not well defined I
will define it and tell you the pros, cons and alternatives to each
decision. But my problem is being creative. I’m lousy at coming up with
ideas.
Well defined project? Sure, that’s easy. But how many projects are really
that well defined? I’m including the definition as part of the
implementation. Me having a cool idea is not the same as defining it. For
example, how are you going to represent the ‘sum total of all embedding QNX
knowledge’ in such a way that an algorithm can take some parameters and
generate an optimal buildfile doing everything you need? Once you’ve
defined your data representation, algorithms and interfaces, the code often
writes itself but getting the correct definition is non-trivial.
Kris
On 28 May 2002 18:17:00 GMT, nospam88@parse.com (Robert Krten) wrote:
Well, there are still two or three immortal questions, from
unde malum to software crisis, that have hardly been “done”, and the
“Encyclopaedia of Ignorance: Everything You Ever Wanted to Know About
the Unknown”, being 25 years old, needs a new edition.
For a couple of things that haven’t been done in our field see
for example EWD1304, “The end of computing science” (and EWD1305,
“Answers to questions from students of Software Engineering” (“the
approximate reconstruction of the questions is left as an exercise to
the reader”)):
http://www.cs.utexas.edu/users/EWD/ewd13xx/EWD1304.PDF
http://www.cs.utexas.edu/users/EWD/ewd13xx/EWD1305.PDF
I haven’t seen quantum (not Quantum) computers yet, but you
can download QCL 0.4.3, a programming language for those beasts to run
on conventional computers. Interesting claims about computational
complexity of quantum algorithms are made, and funny measures of
complexity are proposed (not time, not space, but energy for example).
Hope it won’t end like the “5th generation computers” project…
ako
Remember the Simpson’s episode with Homer on top of Moe’s bar with his band?
Limo pulls up, George Harrison pokes his head out and says, “It’s been done”.Well, that’s kind of the way I feel about the current state of affairs in
the software field.(…)
Anyway, I’m done. Anyone have any ideas on what would be “fun”? (apart from
restoring old computers, of course >Cheers,
-RKRobert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at > www.parse.com> .
Email my initials at parse dot com.
Great insight Robert.
I think there’s another result of “It’s Been Done”:
As a relative newcomer to the software industry, my feeling is that because
“it’s been done”, developers generally try to find new ways to do the same
thing making very simple code very complex. I’m a believer in the philosophy
of KISS (Keep It Simple Stupid) or as our politically correct Cub Scout
leader in our company says KISMIF (Keep It Simple, Make It Fun).
We all want to be innovators but sometimes simple is the way to go.
Doug
“Robert Krten” <nospam88@parse.com> wrote in message
news:ad0his$p52$1@inn.qnx.com…
Remember the Simpson’s episode with Homer on top of Moe’s bar with his
band?
Limo pulls up, George Harrison pokes his head out and says, “It’s been
done”.Well, that’s kind of the way I feel about the current state of affairs in
the software field.Remember when an “operating system” was something to fear and hold in awe?
How did those guys do it? What made it work? How did their memory system
work? Nowadays, it’s “been done”. POSIX says “this is the set of
functions
you will have”. QSSL, to their credit, has done an excellent job in
implementing
the POSIX specs, leaving very little to the imagination. Sure, QNX 4 and
Neutrino
are really “clean” implementations of operating systems, taking tons of
talent
and so on to do, but… “it’s been done”.The separation between filesystem, TCP/IP stack, device drivers, and core
OS
is again so well done that it leaves little to “wonder” at.Remember compilers? This was something that was almost “godlike” in its
ability to “understand” C code and generate low level machine code.
Nowadays, “it’s been done”.There are tons of other examples.
The database guys haven’t progressed – a database is still not some kind
of
wonderful, futuristic AI-like replacement for memory, it’s just a set of
indexes
into files. Wheee… “it’s been done”.The point of this tirade (and I did warn you-all it would be one >
is that I’m really kinda bored – what’s going to be
the “next” exciting thing in the software field that you can hold in awe
and terror (like OS’s and compilers used to be)?About the only thing that jumps out at me is the field of AI – and this
is
because of two things: my own general ignorance of the state-of-the-art
in
the field, and my perception that it’s all just the same old crap – do
they
still use LISP? Are they any closer to making a truly artificial
intelligence?
I don’t need something that looks and acts like a human – if it is truly
“artificial” it could have its own way of looking at the world and
interacting
with it – that’s fine. It doesn’t need to love, or express emotion, or
be
able to compose music (necessarily). It just needs to be able to do
something
that shows its intelligence.Anyway, I’m done. Anyone have any ideas on what would be “fun”? (apart
from
restoring old computers, of course >Cheers,
-RKRobert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at > www.parse.com> .
Email my initials at parse dot com.
Doug Rixmann <rixmannd@rdsdata.com> wrote:
Great insight Robert.
I think there’s another result of “It’s Been Done”:
As a relative newcomer to the software industry, my feeling is that because
“it’s been done”, developers generally try to find new ways to do the same
thing making very simple code very complex. I’m a believer in the philosophy
of KISS (Keep It Simple Stupid) or as our politically correct Cub Scout
leader in our company says KISMIF (Keep It Simple, Make It Fun).We all want to be innovators but sometimes simple is the way to go.
That’s a good point; I hadn’t really considered the “innovation through complexity”
angle We all know companies that have done that
Cheers,
-RK
Doug
“Robert Krten” <> nospam88@parse.com> > wrote in message
news:ad0his$p52$> 1@inn.qnx.com> …Remember the Simpson’s episode with Homer on top of Moe’s bar with his
band?
Limo pulls up, George Harrison pokes his head out and says, “It’s been
done”.Well, that’s kind of the way I feel about the current state of affairs in
the software field.Remember when an “operating system” was something to fear and hold in awe?
How did those guys do it? What made it work? How did their memory system
work? Nowadays, it’s “been done”. POSIX says “this is the set of
functions
you will have”. QSSL, to their credit, has done an excellent job in
implementing
the POSIX specs, leaving very little to the imagination. Sure, QNX 4 and
Neutrino
are really “clean” implementations of operating systems, taking tons of
talent
and so on to do, but… “it’s been done”.The separation between filesystem, TCP/IP stack, device drivers, and core
OS
is again so well done that it leaves little to “wonder” at.Remember compilers? This was something that was almost “godlike” in its
ability to “understand” C code and generate low level machine code.
Nowadays, “it’s been done”.There are tons of other examples.
The database guys haven’t progressed – a database is still not some kind
of
wonderful, futuristic AI-like replacement for memory, it’s just a set of
indexes
into files. Wheee… “it’s been done”.The point of this tirade (and I did warn you-all it would be one >
is that I’m really kinda bored – what’s going to be
the “next” exciting thing in the software field that you can hold in awe
and terror (like OS’s and compilers used to be)?About the only thing that jumps out at me is the field of AI – and this
is
because of two things: my own general ignorance of the state-of-the-art
in
the field, and my perception that it’s all just the same old crap – do
they
still use LISP? Are they any closer to making a truly artificial
intelligence?
I don’t need something that looks and acts like a human – if it is truly
“artificial” it could have its own way of looking at the world and
interacting
with it – that’s fine. It doesn’t need to love, or express emotion, or
be
able to compose music (necessarily). It just needs to be able to do
something
that shows its intelligence.Anyway, I’m done. Anyone have any ideas on what would be “fun”? (apart
from
restoring old computers, of course >Cheers,
-RKRobert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at > www.parse.com> .
Email my initials at parse dot com.
–
Robert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at www.parse.com.
Email my initials at parse dot com.
There’s a bunch of things you can aim for in your code - readability,
efficiency (time and space), reliability, maintainability and the ever
elusive ‘elegance’.
I think elegance is one of the coolest concepts in programming, the closest
to art. It’s not even simple to define. I heard it once defined as, “the
shortest length of code to accomplish the desired task”. By that standard
though, some pretty obfuscated perl code would be considered elegant. Maybe
it’s like the old saw, “I don’t know art but I know what I like.” Quicksort
is elegant. Dijkstra’s algorithm is elegant. I think microkernel
architectures are elegant. Does the beauty lie in simplicity of form and
utility of function? Perhaps our job is complete not when there’s nothing
more to add but when there’s nothing left to take away.
Kris
“Doug Rixmann” <rixmannd@rdsdata.com> wrote in message
news:adnsos$65a$1@inn.qnx.com…
Great insight Robert.
I think there’s another result of “It’s Been Done”:
As a relative newcomer to the software industry, my feeling is that
because
“it’s been done”, developers generally try to find new ways to do the same
thing making very simple code very complex. I’m a believer in the
philosophy
of KISS (Keep It Simple Stupid) or as our politically correct Cub Scout
leader in our company says KISMIF (Keep It Simple, Make It Fun).We all want to be innovators but sometimes simple is the way to go.
Doug
“Kris Warkentin” <kewarken@qnx.com> wrote in message
news:adnud7$bqc$1@nntp.qnx.com…
Perhaps our job is complete not when there’s nothing
more to add but when there’s nothing left to take away.
Geez Kris, I’ve never heard something so ‘elegant’ come out of your mouth.
In fact, I still haven’t. But your fingers–wow–they’ve got something to
say to the world!
Jerry
Nah, he’s just ripping off Michelangelo… 
Geez Kris, I’ve never heard something so ‘elegant’ come out of your mouth.
In fact, I still haven’t. But your fingers–wow–they’ve got something to
say to the world!
Or Antoine de Saint Exupery.
Chris Wiebe <cwiebe@qnx.com> wrote:
: Nah, he’s just ripping off Michelangelo…
:> Geez Kris, I’ve never heard something so ‘elegant’ come out of your mouth.
:> In fact, I still haven’t. But your fingers–wow–they’ve got something to
:> say to the world!
Yeah. I knew I had heard it somewhere before. I probably should have said,
“Someone once said…” but it ruined the flow of the prose.
Perhaps now I should make another quote:
"It is better to keep your mouth closed and let people think you are a fool
than to open it and remove all doubt. "
Mark Twain (1835 - 1910)
cheers,
Kris
“Sean Boudreau” <seanb@qnx.com> wrote in message
news:adog70$ooc$1@nntp.qnx.com…
Or Antoine de Saint Exupery.
Chris Wiebe <> cwiebe@qnx.com> > wrote:
: Nah, he’s just ripping off Michelangelo… >:> Geez Kris, I’ve never heard something so ‘elegant’ come out of your
mouth.
:> In fact, I still haven’t. But your fingers–wow–they’ve got something
to
:> say to the world!
Sean Boudreau <seanb@qnx.com> wrote:
Or Antoine de Saint Exupery.
Or, as discussed at the Ottawa QNX Users’ Group we decided this
was somewhat akin to Einstein’s famous quote (paraphrased) “You should
make things as simple as possible, but no simpler”
Cheers,
-RK
Chris Wiebe <> cwiebe@qnx.com> > wrote:
: Nah, he’s just ripping off Michelangelo… >:> Geez Kris, I’ve never heard something so ‘elegant’ come out of your mouth.
:> In fact, I still haven’t. But your fingers–wow–they’ve got something to
:> say to the world!
–
Robert Krten, PARSE Software Devices +1 613 599 8316.
Realtime Systems Architecture, Books, Video-based and Instructor-led
Training and Consulting at www.parse.com.
Email my initials at parse dot com.