E
Espen Vestre
Pillsy said:
Not /jerks/ perhaps, but maybe Asperger Syndrome type personalities?
But that's the case for math and computer science in general.
Not /jerks/ perhaps, but maybe Asperger Syndrome type personalities?
But that's the case for math and computer science in general.
P
Pillsy
I don't think Lisp is worse for this than any other programming
language.
Kinda, yeah. Though I tend to dislike dragging autism spectrum
disorders into conversations like this one....
Cheers,
Pillsy
[...]
L
Lew
eric-and-jane-smith said:
Either that or the book he bought is good enough to convince him not to learn
Lisp.
No, he doesn't want you to guess what book it is. He doesn't care whether you
know what book it is.
Wait, are you saying that you want him to tell you what book it is? Or what?
E
Espen Vestre
Pillsy said:
I agree - I was merely commenting on your general statement, just to
make that clear.
F
fft1976
Actually, LISP is good for impressing potential employers. It's like
Assembly. The unspoken message here is: "I learned thomething that
nobody needth. Imagine how well I might know thingth that people do
need! However, I will not lower mythelf by lithting thothe things"
L
Lew
eric-and-jane-smith said:
No, I really don't. I am in the habit of saying exactly what I mean, for the
most part, and this was one of those times. I really don't think I was at all
unclear in my statement that it would be subject to misinterpretation.
A good book will give an accurate picture of the language, its capabilities
and its shortcomings. Whether it's trying to convince someone not to use the
language or to use it is irrelevant; an honest assessment might well convince
someone not to use it.
I certainly have encountered many informative references for languages, pretty
much all of which were partisan to their respective topics, that convinced me
not to use those languages. No reflection on the languages, necessarily, it's
just that good information allowed me to make an informed assessment whether I
should invest time and energy in them.
R
Robert Maas, http://tinyurl.com/uh3t
From: Patricia Shanahan said:
I don't know of any open-source projects using WebSphere and JBoss. Do you?
I don't know any way to get access to a computer that is running
WebSphere and JBoss, so my "studying" would be just reading
tutorials without any hands-on practice at what I'm reading.
For me, that's a very ineffective way to learn a software system.
I don't know how I could possibly contribute to a WebSphere/JBoss
project without access to any machine that could test the code I
would write. I would not at all be comfortable editing a wiki of
source-code where I have not *ever* tested even one line of code in
WebSphere or JBoss much less the actual made-up code I'm adding to
the wiki. Most likely I'd be banned from the open source project if
I started editing the group-source to have never-tested code,
because it's extremly unlikely that code I just type in without
even a syntax check could possibly be even close to correct.
R
Robert Maas, http://tinyurl.com/uh3t
(After claim that Common Lisp is best language for not getting a paying job
AFAIK Seed7 can't be used to write WebServer applications,
rendering it moot in discussions between Common Lisp or Java or C
or C++ etc. all of which can. There is actually one *new* (unknown)
language which has passed the test of 3-steps-past-hello-CGI (full
HTML FORM decoding), namely Flaming Thunder. For all 7 languages
that pass that test, see:
<http://www.rawbw.com/~rem/HelloPlus/hellos.html#step3>
I'd really like somebody to volunteer demos of JSP and ASP to
include there, but so-far nobody with access to either has offered
a demo.
AFAIK Seed7 can't be used to write WebServer applications,
rendering it moot in discussions between Common Lisp or Java or C
or C++ etc. all of which can. There is actually one *new* (unknown)
language which has passed the test of 3-steps-past-hello-CGI (full
HTML FORM decoding), namely Flaming Thunder. For all 7 languages
that pass that test, see:
<http://www.rawbw.com/~rem/HelloPlus/hellos.html#step3>
I'd really like somebody to volunteer demos of JSP and ASP to
include there, but so-far nobody with access to either has offered
a demo.
L
Lew
Robert said:
Any open-source Java EE server will work for learning; JBoss is a good one,
and the open-source core of WebSphere is Geronimo. Naturally you wouldn't use
both together.
If you have a computer, preferably one with at least 2GB RAM by the way, then
all you have to do is download Geronimo, JBoss, and/or GlassFish, follow the
installation instructions, and Bob's your uncle.
Do you own a computer?
How much RAM does it have?
So test what you write.
So test your code.
So test your code.
I would offer different advice from contributing to JBoss or WebSphere,
though. I'd use them to practice Java EE coding and deployment skills, not to
contribute to the platforms themselves. One would need experience relying on
the platforms before daring to modify them.
Your caution about contributing to projects without actual deployment is well
founded. Your concern that you can't download or run a Java EE web server is not.
<http://geronimo.apache.org/>
<http://www.jboss.org/>
<https://glassfish.dev.java.net/>
R
Robert Maas, http://tinyurl.com/uh3t
But it is the best language to stay unemployed.
That's a bogus argument, claiming a brand new language like Seed7
is better because it has a shorter history of unemployment. That's
like claiming that 2-yr-old babies are better software programmers
than I am because they've been unemployed only two years.
Perhaps a better way to calculate this is to go all the way back to
most recent employment, sort of like finding the highest-order bit
of a twos-compliment number. A negative number has an infinite
number of high-order negative bits (extended sign), just like how
Seed7 has an infinitely long extended period of unemployment before
it was invented. But Lisp has *some* employment (17.7 years ago) to
break the long stretch of unemployment, so Lisp is better than
Seed7, having *only* 17.7 years unemployment compared to an
infinite amount of unemployment.
Or we could do a proportional way of calculating merit over the
lifetime of the language. Lisp is 50 years old, and I was earning
money at Lisp for about 10 of those years (and using it for 35 of
those years), so Lisp is 20% good per employment and 70% good per
usefulness, whereas Seed7 is 0% good per employment. (YMMV. Is
there anyone who was using the very first implementation of Lisp
for productive work and has continued to use various versions of
Lisp for the whole 50 years thus getting 100% usefulness out of
it?)
That's a bogus argument, claiming a brand new language like Seed7
is better because it has a shorter history of unemployment. That's
like claiming that 2-yr-old babies are better software programmers
than I am because they've been unemployed only two years.
Perhaps a better way to calculate this is to go all the way back to
most recent employment, sort of like finding the highest-order bit
of a twos-compliment number. A negative number has an infinite
number of high-order negative bits (extended sign), just like how
Seed7 has an infinitely long extended period of unemployment before
it was invented. But Lisp has *some* employment (17.7 years ago) to
break the long stretch of unemployment, so Lisp is better than
Seed7, having *only* 17.7 years unemployment compared to an
infinite amount of unemployment.
Or we could do a proportional way of calculating merit over the
lifetime of the language. Lisp is 50 years old, and I was earning
money at Lisp for about 10 of those years (and using it for 35 of
those years), so Lisp is 20% good per employment and 70% good per
usefulness, whereas Seed7 is 0% good per employment. (YMMV. Is
there anyone who was using the very first implementation of Lisp
for productive work and has continued to use various versions of
Lisp for the whole 50 years thus getting 100% usefulness out of
it?)
R
Robert Maas, http://tinyurl.com/uh3t
From: "eric-and-jane-smith said:
According to most of the experts, getting a job depends almost not
at all on what you can do and almost all on whom you know.
Unfortunately I don't know anyone, so all my skills at writing
software go unnoticed. Starting in early 2001 I put up CGI demos so
that people on the net could look at samples of what I can do, thus
become familiar with my abilities to develop working software
applications, but I don't know how to find anyone in the local area
willing to look at my demos and discuss them with me. Several total
strangers I've never met, living thousands of miles from me, have
tried my demos and said they liked them, but they were of no help
in setting me up with local employers. The one local person who was
willing to experience a demo of my very first CGI demo in early
2001, a recruiter at Volt in Mountain View, really liked my demo,
but said his agency recruits *only* for MicroSoft, and they aren't
hiring. When I contacted him again in more recent years, he said
that MicroSoft still wasn't hiring. I haven't yet found even one
person with connections in industry in the local area to let me
show a demo of my more recent and serious CGI applications.
T
thomas.mertes
(After claim that Common Lisp is best language for not getting a paying job
AFAIK Seed7 can't be used to write WebServer applications, ...
Well, Seed7 can be used for WebServer applications, but I admit
that the release does not contain CGI examples. I have to polish
my CGI examples a little bit before I release them. Additionally
I need to write documentation how to write CGI programs in Seed7.
I also plan to extend Comanche, the WebServer for static pages,
released with seed7_05_20090510 such that it is capable to
execute CGI programs.
I thought this is not a language comparison discussion.
After looking at your page for 10 seconds I found the following:
There is a switch to bold (with <b>), probably to highlight
a Warning, but it never switches back to normal. Maybe you
should fix that bug.
Greetings Thomas Mertes
Seed7 Homepage: http://seed7.sourceforge.net
Seed7 - The extensible programming language: User defined statements
and operators, abstract data types, templates without special
syntax, OO with interfaces and multiple dispatch, statically typed,
interpreted or compiled, portable, runs under linux/unix/windows.
T
thomas.mertes
The dialog above uses a concept which might be unknown
to you. Look at http://en.wikipedia.org/wiki/Humour
for an explanation of this concept.
[snip paragraph because keyword 'Lisp' was used]
And now for something completely different.
Greetings Thomas Mertes
Seed7 Homepage: http://seed7.sourceforge.net
Seed7 - The extensible programming language: User defined statements
and operators, abstract data types, templates without special
syntax, OO with interfaces and multiple dispatch, statically typed,
interpreted or compiled, portable, runs under linux/unix/windows.
R
Robert Maas, http://tinyurl.com/uh3t
From: (e-mail address removed)
What you describe there is not so much a lack of a people skill but
the presence of a distinctly un-skill. Fortunately I don't have
that un-skill. My approach is to try to find when a person has time
to listen to my idea, then to present my idea and express my belief
that it's a great idea and ask what they think of it. Then I try to
reasonably deal with any problems they see, for example by
proposing a possible solution to the problem. Unfortunately even
the several people who claimed to really like my idea, don't have
any time to spend working with me on developing the idea, and don't
have time to proofread my preliminary idea-specs, etc., so their
liking the idea doesn't turn out to be of any practical use to me.
As for the several pitches I've made in various threads in several
newsgroups, and on Twitter, for my tinyurl.com/NewEco and its
various individual services, not one person has sent me e-mail
expressing interest much less willingness to work with me on
developing the idea. But when I checked my tinyurl.com/Portl1
shortly after I got the connection-logger running, I saw that
connections had come in from several places around the world:
Cablecom GmbH, Zuerich, CH
NTL Infrastructure, Leeds, GB
Milanese, Italy
Hamburg, Germany
Buenos Aires, AR
Oslo, NO
Barnaul, RU
Jazztel, ES
Tampere, FI
as well as various places in the USA:
Northwestern University, Evanston, IL
PaeTec, Fairport, NY / Saffron Technologies, Morrisville, NC
University of New Mexico / CS
Activision, Santa Monica, CA
Google Inc.
Comcast
The hit from Google was obviously an attempt to spider my site,
which isn't possible beyond the first stage of logging portal
because of the Turing test to protect all access beyond that point,
so Google probably didn't get much to put in their index. I haven't
checked back since May.03 to see if any new IP numbers connected
since then, from even more places around the world. Hmm, would it
be perverse to compare the number of countries connecting to my
tinyurl.com/Portl1 to the number of countries with confirmed cases
of Mexican A-H1N1 flu, and notice that as of May.03 they were
approximately the same number although different in details? (For
example, my site had Italy Argentina and Norway, which the flu had
not yet reached, but lacked New Zealand France and most notably
Mexico!)
Hmm, I did a Google search for text in the contents of the PHP
script that was spidered via tinyurl.com/Portl1, and there were no
matches, so why did Google spider it but not yet include it in
their index nearly a week later? Was it deemed not worth including?
Or was the logged connection *not* their spider but actually some
real live person at Google interested in my NewEco portal?
Yes. Unfortuately in RL I have been unable to find anybody with
enough intelligence and presence of mind to understand my ideas yet
enough spare time to seriously discuss them with me.
Reading a book is just about the worst possible way to become
enlightened as to the value of *totally* interactive development of
computer software made possible by the R-E-P loop together with
identical seantics between typed-in-REP-code an compiled code
(unlike Java, where BeanShell REP is grossly different in semantics
from compiled code).
If and when I get tinyurl.com/NewEco up&running, whereby anyone can
get an account and log in and browse Requests For Bids (RFBs) and
post bids in response and the lowest bidder then works for "pay"
(credit on the system that can't be cashed out at present but can
be used to purchase services from the system or from other
laborers), would you be willing to try your hand at bidding on some
Lisp software contracts (tasks I could do myself but I have far too
much "on my plate" to have time to write *all* the code myself),
little tasks that take a few seconds or up to five minutes, where
you write just a couple lines of Lisp code, or maybe you write one
complete function per spec. Alternately would you be willing to bid
on contracts for writing Seed7 code? Or is Seed7 incapable of
easily being programmed to do most of the data-processing tasks I'd
be wanting done? Or are you collecting $50/hr for Seed7 coding
already so you wouldn't waste your valuable time writing Seed7 code
for anything less than $50/hr?
Here are examples of small tasks I might submit for bids (in Seed7 or whatever):
- Given a large integer, break it into byte-fields per some spec,
such as some fixed-width fields and some UTF-8 variable-length
values. (Inverse of the next task.)
- Given various fields as small integers, express each as
fixed-width or UTF-8 code and append together to make a large
integer. (Inverse of the previous task.)
- Given an associative array containing the decoded HTML FORM contents,
check whether a particular field is present, and if not go down
one branch (stub for the present), else check several specific
possible values for that field and dispatch to appropriate branch
(each stub for present).
- Given the timestamp A3 when a particular "card" was last
finished, and the timestamp B1 when that "card" was scheduled to
be re-started, and the timestamp B2 when that "card" was
actually re-started (sometime after schedule if the system was
busy with something else so it had to be queued for later),
calculate the delta-time from A3 to each of B1 and B2 and
compute the geometric mean ABG of those two delta-times.
- Given the timestamp B3 when the "card" was just now finished
again, and the ABG computed earlier, multiple ABG times a
randomly generated scale factor per a formula I provide, and add
that to B3, to generate the expected time C1 when this "card" is
scheduled to be started *next* time.
- Given a set of "cards", each with scheduled time to be processed
next, all in one table within a MySQL database, determine which
are already (over-)due and which are not yet due, count the due
set, and choose the "card" in the due set that is most overdue.
(2 return values: count, and most-overdue card)
Do you think it would take longer than 5 minutes to write and test
and submit the code for any of those single tasks in Seed7?
This would probably take longer than 5 minutes to code, so I'd need
to break it down into smaller tasks before posting RFBs for each,
or else wait until I have procedures that would support such larger
tasks and a user-base large enough that I'd get more than one bid
for such a large task:
- Given two strings (of text), find the largest matching
sub-string, and from what remains (no overlap allowed)
recursively find the next-largest matching sub-string, providing
that each match is at least two characters long. Return list of
triples showing sub-string and location within each of the two
given strings.
How long do you think it would take to program that in Seed7?
How efficient would it be for moderate-sized strings (appx. 50 to
200 characters in each of the given strings, with matching segments
of all possible lengths, sometimes most of the strings match as one
huge single sub-string, sometimes there are lots of tiny
3-character and 2-character pieces matching but the sequence of
them is scrambled from one input string to the other)?
What you describe there is not so much a lack of a people skill but
the presence of a distinctly un-skill. Fortunately I don't have
that un-skill. My approach is to try to find when a person has time
to listen to my idea, then to present my idea and express my belief
that it's a great idea and ask what they think of it. Then I try to
reasonably deal with any problems they see, for example by
proposing a possible solution to the problem. Unfortunately even
the several people who claimed to really like my idea, don't have
any time to spend working with me on developing the idea, and don't
have time to proofread my preliminary idea-specs, etc., so their
liking the idea doesn't turn out to be of any practical use to me.
As for the several pitches I've made in various threads in several
newsgroups, and on Twitter, for my tinyurl.com/NewEco and its
various individual services, not one person has sent me e-mail
expressing interest much less willingness to work with me on
developing the idea. But when I checked my tinyurl.com/Portl1
shortly after I got the connection-logger running, I saw that
connections had come in from several places around the world:
Cablecom GmbH, Zuerich, CH
NTL Infrastructure, Leeds, GB
Milanese, Italy
Hamburg, Germany
Buenos Aires, AR
Oslo, NO
Barnaul, RU
Jazztel, ES
Tampere, FI
as well as various places in the USA:
Northwestern University, Evanston, IL
PaeTec, Fairport, NY / Saffron Technologies, Morrisville, NC
University of New Mexico / CS
Activision, Santa Monica, CA
Google Inc.
Comcast
The hit from Google was obviously an attempt to spider my site,
which isn't possible beyond the first stage of logging portal
because of the Turing test to protect all access beyond that point,
so Google probably didn't get much to put in their index. I haven't
checked back since May.03 to see if any new IP numbers connected
since then, from even more places around the world. Hmm, would it
be perverse to compare the number of countries connecting to my
tinyurl.com/Portl1 to the number of countries with confirmed cases
of Mexican A-H1N1 flu, and notice that as of May.03 they were
approximately the same number although different in details? (For
example, my site had Italy Argentina and Norway, which the flu had
not yet reached, but lacked New Zealand France and most notably
Mexico!)
Hmm, I did a Google search for text in the contents of the PHP
script that was spidered via tinyurl.com/Portl1, and there were no
matches, so why did Google spider it but not yet include it in
their index nearly a week later? Was it deemed not worth including?
Or was the logged connection *not* their spider but actually some
real live person at Google interested in my NewEco portal?
Yes. Unfortuately in RL I have been unable to find anybody with
enough intelligence and presence of mind to understand my ideas yet
enough spare time to seriously discuss them with me.
By pure coincidence I just bought a book about Lisp a few days
ago. I had some Lisp knowledge already and wanted more
information. From what I have seen so far I can tell you:
I will not turn into a Lisp fan.
Reading a book is just about the worst possible way to become
enlightened as to the value of *totally* interactive development of
computer software made possible by the R-E-P loop together with
identical seantics between typed-in-REP-code an compiled code
(unlike Java, where BeanShell REP is grossly different in semantics
from compiled code).
If and when I get tinyurl.com/NewEco up&running, whereby anyone can
get an account and log in and browse Requests For Bids (RFBs) and
post bids in response and the lowest bidder then works for "pay"
(credit on the system that can't be cashed out at present but can
be used to purchase services from the system or from other
laborers), would you be willing to try your hand at bidding on some
Lisp software contracts (tasks I could do myself but I have far too
much "on my plate" to have time to write *all* the code myself),
little tasks that take a few seconds or up to five minutes, where
you write just a couple lines of Lisp code, or maybe you write one
complete function per spec. Alternately would you be willing to bid
on contracts for writing Seed7 code? Or is Seed7 incapable of
easily being programmed to do most of the data-processing tasks I'd
be wanting done? Or are you collecting $50/hr for Seed7 coding
already so you wouldn't waste your valuable time writing Seed7 code
for anything less than $50/hr?
Here are examples of small tasks I might submit for bids (in Seed7 or whatever):
- Given a large integer, break it into byte-fields per some spec,
such as some fixed-width fields and some UTF-8 variable-length
values. (Inverse of the next task.)
- Given various fields as small integers, express each as
fixed-width or UTF-8 code and append together to make a large
integer. (Inverse of the previous task.)
- Given an associative array containing the decoded HTML FORM contents,
check whether a particular field is present, and if not go down
one branch (stub for the present), else check several specific
possible values for that field and dispatch to appropriate branch
(each stub for present).
- Given the timestamp A3 when a particular "card" was last
finished, and the timestamp B1 when that "card" was scheduled to
be re-started, and the timestamp B2 when that "card" was
actually re-started (sometime after schedule if the system was
busy with something else so it had to be queued for later),
calculate the delta-time from A3 to each of B1 and B2 and
compute the geometric mean ABG of those two delta-times.
- Given the timestamp B3 when the "card" was just now finished
again, and the ABG computed earlier, multiple ABG times a
randomly generated scale factor per a formula I provide, and add
that to B3, to generate the expected time C1 when this "card" is
scheduled to be started *next* time.
- Given a set of "cards", each with scheduled time to be processed
next, all in one table within a MySQL database, determine which
are already (over-)due and which are not yet due, count the due
set, and choose the "card" in the due set that is most overdue.
(2 return values: count, and most-overdue card)
Do you think it would take longer than 5 minutes to write and test
and submit the code for any of those single tasks in Seed7?
This would probably take longer than 5 minutes to code, so I'd need
to break it down into smaller tasks before posting RFBs for each,
or else wait until I have procedures that would support such larger
tasks and a user-base large enough that I'd get more than one bid
for such a large task:
- Given two strings (of text), find the largest matching
sub-string, and from what remains (no overlap allowed)
recursively find the next-largest matching sub-string, providing
that each match is at least two characters long. Return list of
triples showing sub-string and location within each of the two
given strings.
How long do you think it would take to program that in Seed7?
How efficient would it be for moderate-sized strings (appx. 50 to
200 characters in each of the given strings, with matching segments
of all possible lengths, sometimes most of the strings match as one
huge single sub-string, sometimes there are lots of tiny
3-character and 2-character pieces matching but the sequence of
them is scrambled from one input string to the other)?
T
thomas.mertes
You hit the spot. I want objective (neutral point of view)
information.
Why someone thinks that reading a book about Lisp or learning
it, would automatically turn me into a Lisp fan, is beyond my
comprehension (In other words: Is the pope catholic?).
BTW. I took several lectures about Lisp and AI at the university.
That way I am already familiar with car, cdr, caaadr, cons, setq,
dotted pairs, propertys, s-expressions, lambda, deffun, defmacro,
programs as data and other Lisp concepts. I just don't get
enthusiastic about this concepts because I miss other concepts
which IMHO also have some value. I will not start a discussion
about this concepts since I already know how Lisp fanatics
react:
- Either a concept is present in Lisp and therefore considered
great.
- Or a concept is missing in Lisp which renders it unnecessary
and stupid.
My answer to "How should a language look like?" is Seed7.
Seed7 is far from perfect and needs improvement at many
places, but at least some basic concepts are done in a way
I think they should be done. If you find places where Seed7
should be improved, please tell me.
I hope you invest some time to look at Seed7.
Greetings Thomas Mertes
Seed7 Homepage: http://seed7.sourceforge.net
Seed7 - The extensible programming language: User defined statements
and operators, abstract data types, templates without special
syntax, OO with interfaces and multiple dispatch, statically typed,
interpreted or compiled, portable, runs under linux/unix/windows.
R
Robert Maas, http://tinyurl.com/uh3t
From: Lew said:
I agree. But in the case of Lisp I respectfully submit that any
such book might recommend not using Lisp specifically to earn lots
of money as "software engineer" but would surely recommend using
Lisp for writing your own software applications, taking advantage
of rapid development of code in a wide variety of data-processing
areas. Any book that flat out advised never using Lisp for any
purpose couldn't possibly be "good" as you described.
Please look at:
<http://www.rawbw.com/~rem/HelloPlus/hellos.html#s4outl>
and scroll down to Lesson 6 (comparison of languages) and tell me
if you disagree with my main point that Lisp is great overall,
except for specific areas where one of the other languages might be
better. IMO the "good book" you described ought to say something
approximately like what I say there.
Note, if you have suggestions for additions to what I wrote there,
such as additional problem domains where some other language is
much better than Lisp, please tell me!!
Note that for embedded systems, C used to be the only mainstream
language available (with Forth being better for some purposes
although not "mainstream" IMO), but nowadays Java is used a lot in
embedded systems such as cell-phones. I'm not sure what if anything
I should say about this topic in Lesson 6.
R
Robert Maas, http://tinyurl.com/uh3t
But it is the best language to stay unemployed.
Oh no, you're mistaken. I am quite familiar with humor. One of my
favorite TV programs this past season is intensely full of humor.
"The Big Bang Theory"
It was just that after all these months of you hawking Seed7 as
better than Lisp, seriously, not as a joke, it was such a surprise
for you to have given up that and now just make a *joke* about
Seed7 being better than Lisp, not to be taken seriously, that such
an unexpected event completely missed being treated by my mind as a
likely event. Well, then, I'm very pleased you are now treating
Seed7 as a joke rather than a serious programming language.
Note: You have not yet provided an online demo of how Seed7 can be
used as a CGI application, many months after I challenged you to do
so, so at this point I am quite sure either Seed7 isn't a serious
language for such purposes or you are ashamed to show the ugly code
in Seed7 that would be required for such a demo. So maybe it's just
as well that you now accept Seed7 as just a joke.
By the way, what do you think of the Flaming Thunder CGI demo that
was posted sometime last year (or maybe the year before)? Flaming
Thunder succeeds where Seed7 failed.
BTW, if you change your mind again, want to treat Seed7 as serious
again, that CGI-demo challenge remains, so any time you want to
demonstrate that Seed7 can do CGI, feel free to contact me to show
me your demo. Until then, enjoy your humor.
Oh no, you're mistaken. I am quite familiar with humor. One of my
favorite TV programs this past season is intensely full of humor.
"The Big Bang Theory"
It was just that after all these months of you hawking Seed7 as
better than Lisp, seriously, not as a joke, it was such a surprise
for you to have given up that and now just make a *joke* about
Seed7 being better than Lisp, not to be taken seriously, that such
an unexpected event completely missed being treated by my mind as a
likely event. Well, then, I'm very pleased you are now treating
Seed7 as a joke rather than a serious programming language.
Note: You have not yet provided an online demo of how Seed7 can be
used as a CGI application, many months after I challenged you to do
so, so at this point I am quite sure either Seed7 isn't a serious
language for such purposes or you are ashamed to show the ugly code
in Seed7 that would be required for such a demo. So maybe it's just
as well that you now accept Seed7 as just a joke.
By the way, what do you think of the Flaming Thunder CGI demo that
was posted sometime last year (or maybe the year before)? Flaming
Thunder succeeds where Seed7 failed.
BTW, if you change your mind again, want to treat Seed7 as serious
again, that CGI-demo challenge remains, so any time you want to
demonstrate that Seed7 can do CGI, feel free to contact me to show
me your demo. Until then, enjoy your humor.
N
Nicolas Neuss
Hmm, I consider myself a Lisp fan and would not want to choose between
these two statements. Maybe you should refine your simplistic world
view a little qbit?
Nicolas
R
Robert Maas, http://tinyurl.com/uh3t
From: (e-mail address removed)
I'm looking forward to when you do that. But note that the CGI
examples in my helloCGI+steps document are very simple, completely
toy, trivial, just to demonstrate the one principle without mixing
in a bunch of Red Herrings. If you can just translate my demos
directly to Seed7, that would be best.
For more serious CGI examples, showing major things a CGI
application needs in addition to just decoding the HTML FORM
contents, such as validating contents of textfield to be an
integer, or handling cookies, see:
<http://www.rawbw.com/~rem/HelloPlus/CookBook/h4s.html>
After you get the super-basic toy demo of CGI FORM decoding, I hope
you go ahead to also translate those more serious cookbook examples
to Seed7.
IMO, once a student knows how to do general D/P software in Seed7
(or any language), all it takes to learn CGI programming is a set
of trivial demos of major aspects (HTML FORM decoding, validating
various kinds of fields from the form, handling cookies, generating
output of HTML either by inline spewing or by DOM, etc.) with very
little additional text needed to explain what each unit of code is
doing. There's a PHP tutorial I like, see tinyurl.com/phptut, which
is one possible amount/style of documentation for how to do various
things. It's divided into just-starting tutorials about each
different *kind* of thing being handled, and corresponding manual
listing *every* function available for such kind of thing being
handled. I'm not saying that's the best way to do such a tutorial,
but it's a good starting point for thinking of how you want to do
it. I seem to recall your Seed7 tutorial is actually somewhat
similar in verbosity/style. Sun's J2SE JavaDoc is another good
example.
If and when I find time to re-do my cookbook/matrix, it'll be
somewhat different, heavily emphasizing the concept of intentional
datatypes, making clear distinction between the intentional
datatype that expresses what you need and the actual datatype that
affects speed of various operations. For example, Lisp uses binary
integers, which make bit/byte-extraction/building fast but decimal
I/O and digit-extracting slow, whereas PHP uses decimal-digit
strings to represent big integers, making bit/byte
extraction/building rather slow but decimal I/O and digit
extraction fast. Same intentional data type (unsigned integer), two
different actual datatypes that make different operations fast. For
another example, a *set* can be implemented as array or linked list
or hash table or self-balancing binary-search tree, each with its
own particular sets of operations that are fast vs. slow. Finally,
there's UTF-8 (usually compact, but variable length) vs.
fixed-length 32-bit UniCodes (takes more space, but is easier to
scan and count length of string etc.).
Thanks for the heads up. I ran the W3C validator on that WebPage,
and found that (missing 'b' character so I had </> instead of </b>)
as well as two other errors (<br> in wrong place just after </li>
instead of earlier, because the </li> was misplaced). Fixed all
three, so it now validates fine. In all the months since I
previously edited that file and screwed up those three places, you
are the first to find one of them and tell me. If you happen to
notice any other of my active Web pages that fail validation,
please let me know.
Never mind checking the HTML in my PHP scripts for now, because
calls to 'die' don't add the </body></html> that they really
should, so there's no way a PHP script that uses die can pass
validation. And on the free PHP/MySQL hosting sites I'm using, it's
even worse: JavaScript and advertisements are added by the server
in ways that totally break the almost-valid HTML that my PHP script
is generating, for example:
<http://validator.w3.org/check?uri=h...(detect+automatically)&doctype=Inline&group=0>
I'm looking forward to when you do that. But note that the CGI
examples in my helloCGI+steps document are very simple, completely
toy, trivial, just to demonstrate the one principle without mixing
in a bunch of Red Herrings. If you can just translate my demos
directly to Seed7, that would be best.
For more serious CGI examples, showing major things a CGI
application needs in addition to just decoding the HTML FORM
contents, such as validating contents of textfield to be an
integer, or handling cookies, see:
<http://www.rawbw.com/~rem/HelloPlus/CookBook/h4s.html>
After you get the super-basic toy demo of CGI FORM decoding, I hope
you go ahead to also translate those more serious cookbook examples
to Seed7.
IMO, once a student knows how to do general D/P software in Seed7
(or any language), all it takes to learn CGI programming is a set
of trivial demos of major aspects (HTML FORM decoding, validating
various kinds of fields from the form, handling cookies, generating
output of HTML either by inline spewing or by DOM, etc.) with very
little additional text needed to explain what each unit of code is
doing. There's a PHP tutorial I like, see tinyurl.com/phptut, which
is one possible amount/style of documentation for how to do various
things. It's divided into just-starting tutorials about each
different *kind* of thing being handled, and corresponding manual
listing *every* function available for such kind of thing being
handled. I'm not saying that's the best way to do such a tutorial,
but it's a good starting point for thinking of how you want to do
it. I seem to recall your Seed7 tutorial is actually somewhat
similar in verbosity/style. Sun's J2SE JavaDoc is another good
example.
If and when I find time to re-do my cookbook/matrix, it'll be
somewhat different, heavily emphasizing the concept of intentional
datatypes, making clear distinction between the intentional
datatype that expresses what you need and the actual datatype that
affects speed of various operations. For example, Lisp uses binary
integers, which make bit/byte-extraction/building fast but decimal
I/O and digit-extracting slow, whereas PHP uses decimal-digit
strings to represent big integers, making bit/byte
extraction/building rather slow but decimal I/O and digit
extraction fast. Same intentional data type (unsigned integer), two
different actual datatypes that make different operations fast. For
another example, a *set* can be implemented as array or linked list
or hash table or self-balancing binary-search tree, each with its
own particular sets of operations that are fast vs. slow. Finally,
there's UTF-8 (usually compact, but variable length) vs.
fixed-length 32-bit UniCodes (takes more space, but is easier to
scan and count length of string etc.).
Thanks for the heads up. I ran the W3C validator on that WebPage,
and found that (missing 'b' character so I had </> instead of </b>)
as well as two other errors (<br> in wrong place just after </li>
instead of earlier, because the </li> was misplaced). Fixed all
three, so it now validates fine. In all the months since I
previously edited that file and screwed up those three places, you
are the first to find one of them and tell me. If you happen to
notice any other of my active Web pages that fail validation,
please let me know.
Never mind checking the HTML in my PHP scripts for now, because
calls to 'die' don't add the </body></html> that they really
should, so there's no way a PHP script that uses die can pass
validation. And on the free PHP/MySQL hosting sites I'm using, it's
even worse: JavaScript and advertisements are added by the server
in ways that totally break the almost-valid HTML that my PHP script
is generating, for example:
<http://validator.w3.org/check?uri=h...(detect+automatically)&doctype=Inline&group=0>
T
thomas.mertes
As I already said: This is not my simplisic world view.
Proof: Just tell me about great language features that are
missing in Lisp.
Greetings Thomas Mertes
Seed7 Homepage: http://seed7.sourceforge.net
Seed7 - The extensible programming language: User defined statements
and operators, abstract data types, templates without special
syntax, OO with interfaces and multiple dispatch, statically typed,
interpreted or compiled, portable, runs under linux/unix/windows.