How Journalism works
Cameron Laird
recent articles. The details don't matter--I profile a
particular technology for a respected magazine, and the
piece is dated 2001.
All my drafts emphasized that the single best applica-
tion in this domain is a specific Tcl-based one. I also
prepared a Web page of references on how to use other
languages. An editor told me late in the process that
something more "mainstream" was needed. I looked over
what I'd written, puzzled for a bit, then wrote back,
"Sure, of course, no problem, I know exactly what to do",
and wrote an illustration of a programming technique in a
popular different scripting language.
I was surprised and disappointed to find out that all
mention of the best application are gone from what
readers see, along with the references I'd collected.
This is no big deal to me personally; publishing rarely
surprises me these days, and in fact I still rate this
magazine as among the most professional and desirable
of those with which I work. I feel a bit of sorrow for
the people responsible for the Tcl application; I think
they would have welcomed more recognition. What's
hardest to judge is the consequence for readers. I
know the folks at the magazine did this from a sincere
belief that their readers are only ready to read code
written in Perl and Python. Maybe they're right. Maybe
saying anything else is just a waste. In this case, I
believed they could digest a larger truth.
I tell you this just to illustrate a bit of how informa-
tion flows. As I've written several times here recently,
my experience does NOT tell me that particular technical
aspects or decisions constrain Tcl's popular image. What-
ever we do with octal representations or comment
interpretation or Tk look-and-feel will NOT affect the
(business) mass market of sentiment about Tcl.
Remember, everything in the final form of my article is
factual. It's just not the complete or even right
picture, in my view. Keep that in mind, by the way,
next time you're listening or reading or watching the
popular press. Even when you think you've got a direct
connection to reality, and you're watching President
Bush or Derek Jeter or Tom Hanks or Alan Greenspan or
Yasser Arafat speak on his own behalf, at best you're
seeing what a production team has chosen to edit down
for broadcast.
--
Cameron Laird <cla...@NeoSoft.com>
Business: http://www.Phaseit.net
Personal: http://starbase.neosoft.com/~claird/home.html
Andreas Otto
> I just found out about the published form of one of my
> recent articles. The details don't matter--I profile a
> particular technology for a respected magazine, and the
> piece is dated 2001.
Hi,
perhaps you can sent your article to
mailto:zer...@tu-harburg.de
he is doing a good job in writing TclTk based articles
in for german magazines
mfg
aotto
--
================================================================
(C) Compiler-Factory Phone: ++49-(0)8152-399540
Dipl.-Ing Andreas Otto mailto:in...@compiler-factory.com
Business Solutions http://www.compiler-factory.com
Ulmenstrasse 3 => "Compiler", FastWeb, OpMenu
D-34289 Zierenberg => C, C++, Tcl, HTML, database,
=================================================================
Gordon Johnstone
>
> I just found out about the published form of one of my
> recent articles. The details don't matter--I profile a
Can you get the original on to tcl.activestate maybe? Or on this
newsgroup? Just for posterity. Or are there copyright issues.
Gordon
Cameron Laird
My answer: I don't know.
It's an apt question. It's also different from
the way I think. It would take me 'bout a day to
explain just to myself why I didn't think of that
myself. The first layer of an answer is that
everything I have to say to clt on this particular
technology already is in <URL: http://mini.net/tcl >,
and a copy of the draft wouldn't contribute anything
new.
I'll think about it more, though. Thank you, Gordon.
Roy Terry
This is a sad tale indeed and not really
surprising. One question: what's the
point of keeping the details a secret? Are
you bound by a NDA? Do you need to stay
vague to protect future work prospects? If the
reason has to do with your own interests
I can understand it. If not, then how about
the details? The popular prejudices of
industry and marketing hardly need protection
from honest criticism. Further more as the
journalist you also have a trust relationship
with readers that has been violated and
that, I think, is worth struggling for.
It's your reputation too.
Roy Terry
Cameron Laird
Roy Terry <royt...@earthlink.net> wrote:
>Cameron,
>
>This is a sad tale indeed and not really
>surprising. One question: what's the
>point of keeping the details a secret? Are
>you bound by a NDA? Do you need to stay
>vague to protect future work prospects? If the
>reason has to do with your own interests
>I can understand it. If not, then how about
>the details? The popular prejudices of
>industry and marketing hardly need protection
>from honest criticism. Further more as the
>journalist you also have a trust relationship
>with readers that has been violated and
>that, I think, is worth struggling for.
>It's your reputation too.
.
.
More good questions.
No, there's no legal restriction on my actions, 'least
not one I can see. Why don't I, as someone who believes
that "the truth sets us free", supply more details?
That *does* seem paradoxical, doesn't it?
I've got a couple of reasons:
1. I detect enough emotional involvement
in this affair that I don't yet trust
my own judgment. It's simply prudent
for me to keep quiet until I'm more
sure of what to say.
2. 'Best I can tell, it's not *about*
this particular magazine, and I suspect
I'd be contributing to a different kind
of editorial interference by mentioning
it. This is just an example of what
happens constantly in all publications.
Roy Terry
Cameron Laird wrote:
.
.
> I've got a couple of reasons:
> 1. I detect enough emotional involvement
> in this affair that I don't yet trust
> my own judgment. It's simply prudent
> for me to keep quiet until I'm more
> sure of what to say.
Sounds wise to me
> 2. 'Best I can tell, it's not *about*
> this particular magazine, and I suspect
> I'd be contributing to a different kind
> of editorial interference by mentioning
> it. This is just an example of what
> happens constantly in all publications.
Your viewpoint here is a formula for powerlessness:
("I can't do anthing (even protest) a specific grievance
because it's an example of a larger problem and I
can't (singlehandly) solve the larger problem)"
Additionally you might be saying that since
you are no more courageous than the particular
editors you aren't willing to criticize them. If
that is what you're saying then you may be correct
but at the same time you are partronizing and
discounting the character of the editor and
publication. Finally, "editorial interference" means
the management and/or the owners, stockholders
telling editors what they can/can't write. The
term does *not* apply to criticism from outsiders
(if it did then the term would be useless).
BTW, If you choose to do nothing further about
this, I will respect your decision. It's really
a personal matter and I comment simply because
you decided to publish it in the news group.
What were you hoping to achieve by posting?
Roy
Don Porter
> ...Even when you think you've got a direct
> connection to reality, and you're watching President
> Bush or Derek Jeter or Tom Hanks or Alan Greenspan or
> Yasser Arafat speak on his own behalf, at best you're
> seeing what a production team has chosen to edit down
> for broadcast.
There's a difference, though. When I see President Bush,
I see that it's Joe Forehead's report on President Bush.
When I see Tom Hanks, I see that it's Guy Smiley's
interview with Tom Hanks. When I see an article by
Cameron Laird, though, it's not spelled out as explicitly
that it's Cameron Laird (as post-processed by Ed I. Tor).
Thanks for that reminder. Only you can say whether the
end result still deserves to have your name attached to it.
Often your articles include URLs for folks interested in
learning more. Does this one? Perhaps the missing material
can go there?
--
| Don Porter Mathematical and Computational Sciences Division |
| donald...@nist.gov Information Technology Laboratory |
| http://math.nist.gov/~DPorter/ NIST |
|______________________________________________________________________|
Brett Schwarz
I even noticed (looking at the article again), that nothing is even mentioned in the "resources" section. Hummm...
Anyways, I was intending to write a "letter to the editor" about this, asking why Tcl was not mentioned more. I will still go through with it. I hope this is ok with you.
thanks,
--brett
p.s. I have also noticed in this magazine (and other "similar" ones), that alot of articles will reference other languages (i.e. this can be implemented in Per, Python), but 50% of the time, Tcl is left out (even mentioning it). This really fustrates me. I don't know if they are ignorant about Tcl, or if they just detest it enough to not even mention it. It's a pity...
On 24 Oct 2001 09:05:53 -0500
Gerhard Hintermayer
>
> .... a posting with a line length of 574 chars. (my mouse got hot because of the scrolling)
Ever thought of using a news-reader that does a good job ?
--
Gerhard Hintermayer
Cameron Laird
Brett Schwarz <brett_...@yahoo.com> wrote:
.
.
.
>p.s. I have also noticed in this magazine (and other "similar" ones), that alot of articles will reference other languages (i.e. this can be implemented in Per, Python), but 50% of the time, Tcl is left out (even mentioning it). This really fustrates me. I don't know if they are ignorant about Tcl, or if they just detest it enough to not even mention it. It's a pity...
.
.
The editorial staff can be aware of Tcl, and think it
technically appropriate in some sense, but too exotic
for readers or advertisers to swallow. That happens
a lot.
Andrew Tannenbaum
"before" and "after," on the web side by side.
As for how journalism works, I think this is usually true, at least
in my experience with the American press. Any time I've seen a story
reported that I've been involved in, or seen unfold first-hand, the
reporter always edits to convey a message conceived by the news
source, that usually does not reflect plain fact.
On TV, it's usually done by careful selection of "sound bites."
They call this "freedom of the press" - they report what they want to,
and fairness and accuracy are not given high priority.
Andy
<<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>><<>>
In article <3BD6D17D...@space.qinetiq.com>,
Gordon Johnstone <gor...@thornsds.co.uk> wrote:
Kristoffer Lawson
: The editorial staff can be aware of Tcl, and think it
: technically appropriate in some sense, but too exotic
: for readers or advertisers to swallow. That happens
: a lot.
Claims like that are frankly quite ridiculous though (no, I'm not
blaming you!). I believe Tcl is quite clearly the third most popular script
language of the bunch and definitely not one that should be ignored. In
Debian there are still more Tcl-related packages than Python-related ones.
So why do editors do that? Why do they want to ignore Tcl? Why is our image
so poor?
--
- ---------- = = ---------//--+
| / Kristoffer Lawson | www.fishpool.fi|.com
+-> | se...@fishpool.com | - - --+------
|-- Fishpool Creations Ltd - / |
+-------- = - - - = --------- /~setok/
David N. Welton
> I believe Tcl is quite clearly the third most popular script
> language of the bunch and definitely not one that should be ignored.
I would almost say that PHP is. As far as one that new people are
learning, it's higher than Tcl for sure.
--
David N. Welton
Consulting: http://www.dedasys.com/
Free Software: http://people.debian.org/~davidw/
Apache Tcl: http://tcl.apache.org/
Personal: http://www.efn.org/~davidw/
Cameron Laird
Kristoffer Lawson <se...@fishpool.com> wrote:
>Cameron Laird <cla...@starbase.neosoft.com> wrote:
>
>: The editorial staff can be aware of Tcl, and think it
>: technically appropriate in some sense, but too exotic
>: for readers or advertisers to swallow. That happens
>: a lot.
>
>Claims like that are frankly quite ridiculous though (no, I'm not
>blaming you!). I believe Tcl is quite clearly the third most popular script
>language of the bunch and definitely not one that should be ignored. In
>Debian there are still more Tcl-related packages than Python-related ones.
>So why do editors do that? Why do they want to ignore Tcl? Why is our image
>so poor?
.
.
I don't know--not at any satisfyingly deep level.
Kristoffer, both you and I have pondered this,
and I've dismissed a lot of your suggestions in
the past; I *do* respect your interest and intel-
ligence in pursuing the issue, though. It is
absolutely true that Tcl fully deserves attention,
as abundant facts:
* Debian has more Tcl-related packages ...
* Tcl is in IBM's strategically-critical
WSCP
* Tcl is in Oracle's strategically-critical
OEM
* ...
confirm. I also can report that I have abundant
interest that "readers don't buy Tcl": book
sales are mostly risible, magazine editors con-
sistently evaluate Tcl as a fringe subject,
and training classes receive marginal attendance.
I have no more profound conclusion. I can only
thank you for your observations.
Gordon Johnstone
>
> Kristoffer Lawson <se...@fishpool.com> writes:
>
> > I believe Tcl is quite clearly the third most popular script
> > language of the bunch and definitely not one that should be ignored.
>
> I would almost say that PHP is. As far as one that new people are
> learning, it's higher than Tcl for sure.
>
Don't ya just hate it when people cloud the issues with facts.... ;0)
Gordon
Donal K. Fellows
> I also can report that I have abundant
> interest that "readers don't buy Tcl": book
> sales are mostly risible,
I can't buy Tcl books from a bookstore because the bookstore doesn't
carry them (and I'm not going to get multiple copies of the same book
for anyone! :^)
Chicken and egg...
Donal.
--
Donal K. Fellows http://www.cs.man.ac.uk/~fellowsd/ fell...@cs.man.ac.uk
-- There are worse futures that burning in hell. Imagine aeons filled with
rewriting of your apps as WinN**X API will mutate through eternity...
-- Alexander Nosenko <n...@titul.ru>
Kristoffer Lawson
: Kristoffer Lawson <se...@fishpool.com> writes:
:> I believe Tcl is quite clearly the third most popular script
:> language of the bunch and definitely not one that should be ignored.
: I would almost say that PHP is. As far as one that new people are
: learning, it's higher than Tcl for sure.
Good point. I just generally tend to ignore domain-specific languages when
taking part in discussions like this, and I really consider PHP to be that.
Kristoffer Lawson
: I can't buy Tcl books from a bookstore because the bookstore doesn't
: carry them (and I'm not going to get multiple copies of the same book
: for anyone! :^)
Have you talked to them? I find that if shopkeepers are aware of a need,
they will generally try to provide for it.
David N. Welton
> Good point. I just generally tend to ignore domain-specific
> languages when taking part in discussions like this, and I really
> consider PHP to be that.
Yes, but it's so damn popular that when it breaks out of that, despite
the cruft, it will be ahead of Tcl:-( It's maddening to try and
convince people that domain specific languages are a silly idea.
Stefaan A Eeckels
I think the reason is that Tcl is different. Your average
programming Joe or Jane barely grasps the concepts behind 'C',
and when they encounter Tcl they are about as stymied as
when encountering say, Forth.
I've given quite a few "Introduction to Tcl/Tk" talks, and
the simple fact is that people who survive as VB and
C++ (as generated by Visual C++) programmers don't grok Tcl.
When I show the audience some really impressive Tcl/Tk programs,
I hear "Hey, this is kinda neat -- it only took you _two_ days?"
but when I show them the code all interest is gone. I get the
same reaction when I show people some of my PostScript tricks.
They are suitably impressed, but it doesn't convince them
PostScript (or Forth) is worthwhile learning.
My pet theory is that languages that're not recognisably
algebraic in nature and notation are doomed to have a
niche audience.
--
Stefaan (GPG Fingerprint 25D8 551B 4C0F BF73 3283 21F1 5978 D158 7539 76E4)
--
"Technically, Windows is an 'operating system,' which means that it
supplies your computer with the basic commands that it needs to
suddenly, with no warning whatsoever, stop operating." -Dave Barry
Cameron Laird
Stefaan A Eeckels <Stefaan...@ecc.lu> wrote:
.
.
.
>When I show the audience some really impressive Tcl/Tk programs,
>I hear "Hey, this is kinda neat -- it only took you _two_ days?"
>but when I show them the code all interest is gone. I get the
>same reaction when I show people some of my PostScript tricks.
>They are suitably impressed, but it doesn't convince them
>PostScript (or Forth) is worthwhile learning.
>
>My pet theory is that languages that're not recognisably
>algebraic in nature and notation are doomed to have a
>niche audience.
.
.
That sounds plausible.
I feel as though Tcl was more acceptable at times
in the past--but perhaps that only confirms your
proposition. It might be that the programming
population was more intellectually adventurous a
decade ago.
Cameron Laird
>algebraic in nature and notation are doomed to have a
>niche audience.
.
.
Oh! Oh! I should add that this must be understood as
"... recognisably algebraic (in a programming sense) in
nature ..."
Real algebraists grok Forth and Tcl far easier than they
get Perl and C.
Don Porter
> I feel as though Tcl was more acceptable at times
> in the past--but perhaps that only confirms your
> proposition.
Tcl is selling to a different crowd now.
5+ years ago, Tcl was sold to C programmers as a better way
of programming than doing everything in C.
Now it's offered as an alternative scripting language to
folks who already know Perl or Python, or those who know
none, but want the first one they learn to be the most popular
one. Tcl's every bit as good as those other scripting
languages, but it's hard to come up with example where it's
just miles and miles ahead of them, and that kind of clear
superiority is necessary to go against the flow.
Helmut Giese
>I think the reason is that Tcl is different. Your average
>programming Joe or Jane barely grasps the concepts behind 'C',
>and when they encounter Tcl they are about as stymied as
>when encountering say, Forth.
>
>I've given quite a few "Introduction to Tcl/Tk" talks, and
>the simple fact is that people who survive as VB and
>C++ (as generated by Visual C++) programmers don't grok Tcl.
>
>When I show the audience some really impressive Tcl/Tk programs,
>I hear "Hey, this is kinda neat -- it only took you _two_ days?"
>but when I show them the code all interest is gone. I get the
>same reaction when I show people some of my PostScript tricks.
>They are suitably impressed, but it doesn't convince them
>PostScript (or Forth) is worthwhile learning.
programmers seem to be different. I am often astonished of (what I
presume to be) the lack of basics and general concepts - they know a
little C, and (so it seems to me) a couple of years later they don't
know much more.
Reasons?
In the last years the need for programmers has so much grown, that
many more people took such kind of formation ('cuse me, that's French,
is that English, too? 'education' maybe?) and / or got the respective
jobs.
Back in the good ol' days :) I somehow believe that a programmer had
to have another attitude: Most of the stuff had to be programmed 'in
house' - there weren' t those fantastic, ready made front / back ends,
where you do a little clicking and everything works. All of us had to
do more 'grunt work' - but learned a lot by doing so.
Today, you can click together an app in a real short time - which is
*good* (I just imagine doing GUI with the raw Windows API - shudder).
And as long as everything works from the beginning, all is fine. But
beware of the moment, when something goes wrong: At this moment some
deeper understanding is required - bof.
>
>My pet theory is that languages that're not recognisably
>algebraic in nature and notation are doomed to have a
>niche audience.
But what you said above makes sense to me. Should I ever be in need to
present Tcl to an audience, I will think twice, how I'd go about it.
Best regards
Helmut
Acacio Cruz
> Tcl's every bit as good as those other scripting
> languages, but it's hard to come up with example where it's
> just miles and miles ahead of them, and that kind of clear
> superiority is necessary to go against the flow.
Tcl is a *gazillion* miles ahead of Perl in maintainability (new English word, I'm sure! :) )
Anyone who ever had to return to an already made Perl script to change will understand what I mean.
-Acacio
Phil Ehrens
apps are as stable as Tcl apps, and they are just plain
wrong. The Tcl event loop allows you to safely do very
complex things that, when you do them in other languages,
become hazards. This is why Perl and Python both embed
Tcl and Tk to do their dirty work ;^)
Has anybody looked at Sketch? Brilliant, but as far as
I can tell Python is just there as a wrapper for Tcl/Tk.
http://sketch.sourceforge.net/index.html
I still think all we need is someone a bit rougher than
Dr. O to go one-on-two with Larry and Guido and make
them say WHY they need Tcl and Tk if their languages are
so perfectly wonderful.
I can't say I ever think "gosh, I wish I could put some
Perl in here because Tcl just can't handle this". But
apparently both Larry and Guido said that about Tcl,
since there it is.
And I'll say it one more time: all we need is for
Brent Welch to appear on stage with his footy bag
and a beer, and all the geeks will dump Larry and
Guido in a second.
Phil
Don Porter wrote:
> ... Tcl's every bit as good as those other scripting
Cameron Laird
.
[hearty, fact-
filled rant]
.
.
>I can't say I ever think "gosh, I wish I could put some
>Perl in here because Tcl just can't handle this". But
>apparently both Larry and Guido said that about Tcl,
>since there it is.
.
.
Well, kind of. It's not clear Larry is
aware of graphical stuff, and he certainly
isn't responsible for putting Tcl in Perl.
Perl/Tk is remarkably marginal to his
concerns.
--
Cameron Laird <Cam...@Lairds.com>
Phil Ehrens
since he wrote the broken Tk examples that come with
the Python source ;^)
Cameron Laird wrote:
>>I can't say I ever think "gosh, I wish I could put some
>>Perl in here because Tcl just can't handle this". But
>>apparently both Larry and Guido said that about Tcl,
>>since there it is.
> .
Les Cargill
Absolutem obsoletum - that which works is considered to be
obsolete.
Pehaps as the focus on profitability and other signs of
sanity occur ( however painfully) to the techie sphere,
Tcl will get its due. FWIW, the whole of tech journalism is
atrocious. They don't even gots Bob Pease in EE Times no more :)
You never see news reports that say "A missile was launched today in
which nothing went wrong, noone was hurt and cost was contained.".
> --
>
> Cameron Laird <cla...@NeoSoft.com>
> Business: http://www.Phaseit.net
> Personal: http://starbase.neosoft.com/~claird/home.html
Les Cargill
Stefaan A Eeckels wrote:
>
<snip>
> I think the reason is that Tcl is different. Your average
> programming Joe or Jane barely grasps the concepts behind 'C',
> and when they encounter Tcl they are about as stymied as
> when encountering say, Forth.
>
> I've given quite a few "Introduction to Tcl/Tk" talks, and
> the simple fact is that people who survive as VB and
> C++ (as generated by Visual C++) programmers don't grok Tcl.
>
> When I show the audience some really impressive Tcl/Tk programs,
> I hear "Hey, this is kinda neat -- it only took you _two_ days?"
> but when I show them the code all interest is gone. I get the
> same reaction when I show people some of my PostScript tricks.
> They are suitably impressed, but it doesn't convince them
> PostScript (or Forth) is worthwhile learning.
>
> My pet theory is that languages that're not recognisably
> algebraic in nature and notation are doomed to have a
> niche audience.
Tcl is far more algebraic in nature than aby other language I can think
of except Lisp, in the general sense of "algebra" as opposed to the
class you took in 9th grade.
>
> --
> Stefaan (GPG Fingerprint 25D8 551B 4C0F BF73 3283 21F1 5978 D158 7539 76E4)
> --
> "Technically, Windows is an 'operating system,' which means that it
> supplies your computer with the basic commands that it needs to
> suddenly, with no warning whatsoever, stop operating." -Dave Barry
Cameron Laird
>Okay, given that, it's hard to say the same about Guido,
>since he wrote the broken Tk examples that come with
>the Python source ;^)
.
.
I think you're complimenting Guido. He, like
several others of the greats, designed a
language that's so good he's not the best
stylist in it. That implies deep creativity
and taste.
You might also be making a small technical
observation about Tkinter. That we'll settle
another time.
Don Porter
> And I'll say it one more time: all we need is for
> Brent Welch to appear on stage with his footy bag
> and a beer, and all the geeks will dump Larry and
> Guido in a second.
Might as well take on the Apache Team too...
...able to build an entire web server with his bare hands...
Stefaan A Eeckels
> Stefaan A Eeckels wrote:
> >
> > My pet theory is that languages that're not recognisably
> > algebraic in nature and notation are doomed to have a
> > niche audience.
>
> Tcl is far more algebraic in nature than aby other language I can think
> of except Lisp, in the general sense of "algebra" as opposed to the
> class you took in 9th grade.
Actually, I was using "algebraic" as a shorthand for the
FORTRAN/Algol class of languages. I'm pretty sure I'm not
the first one to use it. Maybe my brain's addled from writing
too many status reports.
Stefaan A Eeckels
> In article <20011025212143.7fbf...@ecc.lu>,
> Stefaan A Eeckels <Stefaan...@ecc.lu> wrote:
> .
> .
> .
> >My pet theory is that languages that're not recognisably
> >algebraic in nature and notation are doomed to have a
> >niche audience.
> .
> .
> .
> Oh! Oh! I should add that this must be understood as
> "... recognisably algebraic (in a programming sense) in
> nature ..."
That's what I meant - the languages that say things like
i = i + 1
and descend from FORTRAN/Algol.
>
> Real algebraists grok Forth and Tcl far easier than they
> get Perl and C.
I remember the few epiphanies in my career. The first and most
important was simply programming (on a TI-56), which got me into
real microprocessors and assembly coding in general.
The second one was the first time I used the Unix shell.
The final one was Forth. I actually struggled to really grok 'C'
and Perl gives me urticaria.
Donal K. Fellows
> That's what I meant - the languages that say things like
> i = i + 1
You want assignment operators added to [expr]? (If we do one, we do the
whole lot of C ones.) It's a bunch of work, but fundamentally feasable
if someone wants to TIP it up (though it is definitely not something I'd
encourange for 8.4; it feels like too big a change to me. I'd also much
rather have ':=' for assignment than just plain '=' as keeping the form
which can be mixed up as a syntax error avoids sooo much trouble! :^)
Donal.
--
Donal K. Fellows http://www.cs.man.ac.uk/~fellowsd/ fell...@cs.man.ac.uk
make: *** No rule to make target `war'. Stop.
David N. Welton
> Pehaps as the focus on profitability and other signs of sanity occur
> ( however painfully) to the techie sphere, Tcl will get its due.
I think that this should be one of the focuses when talking with
people about the Tcl Renaissance.
Chris Nelson
> ...
> and Perl gives me urticaria.
In _The Belle of Amherst_, Emily Dickenson says, "Now _there's_ a word
you can take your hat off to!" Twice this week, I've had to look up a
word -- an english word, not some jargon -- that I read on c.l.t. First
that was Cameron's "apothegmatizing", now "urticaria." :-)
Chris
--
Given infinite time, everything will happen.
David N. Welton
> Stefaan A Eeckels wrote:
> > ...
> > and Perl gives me urticaria.
>
> In _The Belle of Amherst_, Emily Dickenson says, "Now _there's_ a word
> you can take your hat off to!" Twice this week, I've had to look up a
> word -- an english word, not some jargon -- that I read on c.l.t. First
> that was Cameron's "apothegmatizing", now "urticaria." :-)
Using gnus to read news, you can just do:
(defun dict-word-at-point ()
"Runs 'dict' command on word at point"
(interactive)
(shell-command (concat "dict" " " (thing-at-point 'word))))
M-x dict-word-at-point
and get the definition. Quite handy, although the standard dict
server doesn't seem to have "apothegmatizing".
Chris Nelson
> ...
> You never see news reports that say "A missile was launched today in
> which nothing went wrong, noone was hurt and cost was contained.".
I don't know; Mars Odyssey seemed to make the news. ;-)
--
The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw
Max Ischenko
Stefaan A Eeckels wrote:
> I think the reason is that Tcl is different. Your average programming
> Joe or Jane barely grasps the concepts behind 'C', and when they
> encounter Tcl they are about as stymied as when encountering say,
> Forth.
> I've given quite a few "Introduction to Tcl/Tk" talks, and the simple
> fact is that people who survive as VB and C++ (as generated by Visual
> C++) programmers don't grok Tcl.
I think you're basically right, based on my own experience. I played with
Tcl several years ago and found it "strange" and "clumsy". Then I learned
Python and was excited about the language. Time pass by. I was introduced
to Ruby, LISP, Smalltalk, browse c2.com, read a lot.
Then I came back to TCL and now I'm able to do much more
powerful/elegant programming in tcl. It seems that I have `grok' it at
least.
--
Do not underestimate the value of print statements for debugging.
Andrew Kuchling
> I still think all we need is someone a bit rougher than
> Dr. O to go one-on-two with Larry and Guido and make
> them say WHY they need Tcl and Tk if their languages are
> so perfectly wonderful.
Because Tk was, and pretty much still is, the only GUI toolkit out
there that's portable across all of Windows, Unix, MacOS. That's the
only real reason. See the python-dev summary for October 2000 at
http://www.amk.ca/python/dev/2000-10-2.html for a summary of the last
discussion about this.
wxWindows is equally portable, but for some reason has never really
caught on. With the release of Qt/Mac, Qt now covers the new MacOS
but not the old one, so perhaps the PyQt binding is the way forward.
(I certainly hope this is the case, as Qt is really clean and
attractive.)
--amk
Cameron Laird
Andrew Kuchling <akuc...@mems-exchange.org> wrote:
.
.
.
>wxWindows is equally portable, but for some reason has never really
>caught on. With the release of Qt/Mac, Qt now covers the new MacOS
>but not the old one, so perhaps the PyQt binding is the way forward.
>(I certainly hope this is the case, as Qt is really clean and
>attractive.)
>
>--amk
>
And well-documented.
Why is wxWindows so little-known? I've looked into that
before; I think part of the reason is its portability.
During the mid-'90s, USAicans had trouble taking seriously
a portable product supported by British academics. Such a
construct couldn't possibly look good, according to Western
Hemisphere beliefs.
wxWindows' C++ implementation was also out-of-step with
several audiences.
Kristoffer Lawson
: I feel as though Tcl was more acceptable at times
: in the past--but perhaps that only confirms your
: proposition. It might be that the programming
: population was more intellectually adventurous a
: decade ago.
I never found Tcl's programming method to be that strange. In many
ways it's pretty much the same as everywhere, except that everything's
a command. OTOH I always love concise specification and logic so maybe
it just felt "right" immediately..
Larry Smith
> In article <20011025212143.7fbf...@ecc.lu>,
> Stefaan A Eeckels <Stefaan...@ecc.lu> wrote:
> >My pet theory is that languages that're not recognisably
> >algebraic in nature and notation are doomed to have a
> >niche audience.
.
> That sounds plausible.
> I feel as though Tcl was more acceptable at times
> in the past--but perhaps that only confirms your
> proposition. It might be that the programming
> population was more intellectually adventurous a
> decade ago.
I think most people were, but today we are at the
tail end of a decades-long business cycle that has
shortened so drastically that nowadays management
only looks at the quarterly bottom line. Sure Tcl
can save them 50% of their effort or more - but it
takes a month to get comfortable with it, and we
need things done NOW, not a month from now. And
so our entire culture, top to bottom, has become
the ultimate example of penny wise and pound foolish.
No one - and I mean NO one - outside of the open
source world is even remotely interested in long
term issues, maintainability, or stability. All
they care about is ship time.
And there is another evil aspect lurking in all
this: it has been turned into a business staple.
You NEED bugs and instability to keep people on
the upgrade escalator and keep the money flowing
in. If this stuff just worked, was stable, and
only needed a new module every now and again to
add a feature - well, there's not much call for
that...from the software companies. Just because
every user in the world is begging for it doesn't
mean it's a priority or anything...
--
.-. .-. .---. .---. .-..-. | Wild Open Source Inc.
| |__ / | \| |-< | |-< > / | "Making the bazaar just a
`----'`-^-'`-'`-'`-'`-' `-' | little more commonplace."
home: www.smith-house.org | work: www.wildopensource.com
Steve Blinkhorn
: As for how journalism works, I think this is usually true, at least
: in my experience with the American press. Any time I've seen a story
: reported that I've been involved in, or seen unfold first-hand, the
: reporter always edits to convey a message conceived by the news
: source, that usually does not reflect plain fact.
: On TV, it's usually done by careful selection of "sound bites."
: They call this "freedom of the press" - they report what they want to,
: and fairness and accuracy are not given high priority.
Although I don't write for the technical computing press, I do have
many dealings with journalists and broadcasters, and I do write not
infrequently for Nature. My experience, which may be UK parochial,
is that journalists listen if you are a) very up front and open, and
b) authoritative and explicit about the core of what you have to say.
For instance, last year I did a TV interview for Newsnight, which is
the intellectual heavyweight BBC bulletin, about issues concerning
university entrance practices in the UK. The journalist came with a
lot of preconceived ideas about what I ought to think, and a camera
crew. As is my normal practice, I told them just to set the camera
rolling, we recorded about 40 minutes worth, and then I told her that
she could cut the material any way she liked *so long as the
impression conveyed was (.... detail of my position on the matter,
which I don't need to bore you with)*. I have to say that the end
result was a beautifully-cut 30 seconds.
I've always insisted on seeing proofs of anything I've written before
it goes to press, and on journalists checking back with me if they
are saying anything technical, or if they want to manufacture a quote.
A lot of people may be shy of doing it, perhaps because of the use of
terms like "submitted for publication". Don't submit: if you're the
expert, you may have to work hard to express things in simple enough
language, but in the end you have the right not to have your material
mistreated. In UK law, there is a statutory 'moral right' to be
identified as author, and to prevent 'derogatory treatment' of what
you write. If what happened to Cameron had been done to me I would
have gone into ton of bricks mode: editors aren't doing you a favour
by publishing, you're providing them with the material they need to
make a living.
The instances where stuff I have written has been heavily edited have
always been when a new and inexperienced copy editor has had a rush of
blood to the head. I've always, absolutely always, been able to
reverse such changes. And one trick I discovered early on: always
write as nearly exactly to the required length as possible, then if
they cut stuff, they're short of copy.
--
Steve Blinkhorn <st...@prd.co.uk>
Victor Wagner
: in the past--but perhaps that only confirms your
: proposition. It might be that the programming
: population was more intellectually adventurous a
: decade ago.
: --
There is same amount of adveturous people in trade now.
There is just a lot more people who call themeselves programmers,
so people who want to learn are no more majority.
--
This is a test of the Emergency Broadcast System. If this had been an
actual emergency, do you really think we'd stick around to tell you?
Victor Wagner
: Don Porter wrote:
:> Tcl's every bit as good as those other scripting
:> languages, but it's hard to come up with example where it's
:> just miles and miles ahead of them, and that kind of clear
:> superiority is necessary to go against the flow.
: Tcl is a *gazillion* miles ahead of Perl in maintainability
:(new English word, I'm sure! :) )
size of project. In interval between 10 and 1000 lines you
might be true, but if you have larger projects, even rudimentary
ways of static checking which present in Perl, become a huge advantage.
Unfortunately, Tcl has no equivalents of use strict, -wc and -T.
: Anyone who ever had to return to an already made Perl script
:to change will understand what I mean.
I have 40000 lines project to maintain on Perl. But I disagree with
you.
: -Acacio
--
'Course, that doesn't work when 'a' contains parentheses.
-- Larry Wall in <1997102116...@wall.org>
Les Cargill
Chris Nelson wrote:
>
> Les Cargill wrote:
> > ...
> > You never see news reports that say "A missile was launched today in
> > which nothing went wrong, noone was hurt and cost was contained.".
>
> I don't know; Mars Odyssey seemed to make the news. ;-)
>
True! Breath of fresh air, that.
> --
> The reasonable man adapts himself to the world; the unreasonable one
> persists in trying to adapt the world to himself. Therefore all
> progress depends on the unreasonable man. -- George Bernard Shaw
Acacio Cruz
Victor Wagner wrote:
> Acacio Cruz <acaci...@eur.autodesk.com> wrote:
> : Don Porter wrote:
>
> :> Tcl's every bit as good as those other scripting
> :> languages, but it's hard to come up with example where it's
> :> just miles and miles ahead of them, and that kind of clear
> :> superiority is necessary to go against the flow.
>
> : Tcl is a *gazillion* miles ahead of Perl in maintainability
> :(new English word, I'm sure! :) )
>
> No, not at all. It depends on your programming habits and
> size of project. In interval between 10 and 1000 lines you
> might be true, but if you have larger projects, even rudimentary
> ways of static checking which present in Perl, become a huge advantage.
>
> Unfortunately, Tcl has no equivalents of use strict, -wc and -T.
>
> : Anyone who ever had to return to an already made Perl script
> :to change will understand what I mean.
>
> I have 40000 lines project to maintain on Perl. But I disagree with
> you.
We agree that we disagree.
You, individually, may feel at ease with your huge perl project.
I sincerely doubt that many other people would too.
What works best for an individual is not necessarily the best fir a group,
enterprise, team, ... And, at that level, Perl's inherent crypticness is a fault.
As the Camel book says, "there is not *one* way of doing something; there's many"
and this "feature" truly is a problem when maintenance time comes.
Of course, good practices ease the pain but still the pain baseline for Perl is higher
than Tcl's, for instance. That's what attracted me to Tcl from Perl.
-Acacio
Les Cargill
Larry Smith wrote:
>
> Cameron Laird wrote:
>
> > In article <20011025212143.7fbf...@ecc.lu>,
> > Stefaan A Eeckels <Stefaan...@ecc.lu> wrote:
>
> > >My pet theory is that languages that're not recognisably
> > >algebraic in nature and notation are doomed to have a
> > >niche audience.
> .
> > That sounds plausible.
>
> > I feel as though Tcl was more acceptable at times
> > in the past--but perhaps that only confirms your
> > proposition. It might be that the programming
> > population was more intellectually adventurous a
> > decade ago.
>
> I think most people were, but today we are at the
> tail end of a decades-long business cycle that has
> shortened so drastically that nowadays management
> only looks at the quarterly bottom line.
The number of problems that can be solved in 90
days is painfully short.
> Sure Tcl
> can save them 50% of their effort or more - but it
> takes a month to get comfortable with it, and we
> need things done NOW, not a month from now. And
> so our entire culture, top to bottom, has become
> the ultimate example of penny wise and pound foolish.
The pennies will eventually dry up and those with
the pounds win in the long term. Evolution is really
hard to argue with.
> No one - and I mean NO one - outside of the open
> source world is even remotely interested in long
> term issues, maintainability, or stability. All
> they care about is ship time.
>
I respectfully diagree. It's a focus in open source
because that's a cultural norm of those in open
source, but the industry has to get back to realistic
business models or it will simply stop being.
One of the great things about Tcl is that the objects
built under it really can become commoditizable.
Download, unzip, "package require" and you're off .
Because of the amount of discipline built into the
interface structure, it really is pretty much plug
and play. Might take some tweaking, but much less
than other paradigms I've seen.
> And there is another evil aspect lurking in all
> this: it has been turned into a business staple.
> You NEED bugs and instability to keep people on
> the upgrade escalator and keep the money flowing
> in.
This is not new. Only the rate has changed. I once
worked for somebody who anticipated the WinXP
business model by at least two decades...
> If this stuff just worked, was stable, and
> only needed a new module every now and again to
> add a feature - well, there's not much call for
> that...from the software companies. Just because
> every user in the world is begging for it doesn't
> mean it's a priority or anything...
>
If you're saying the present business model is broken,
I'll agree. I don't know what will replace it.
> --
> .-. .-. .---. .---. .-..-. | Wild Open Source Inc.
> | |__ / | \| |-< | |-< > / | "Making the bazaar just a
> `----'`-^-'`-'`-'`-'`-' `-' | little more commonplace."
> home: www.smith-house.org | work: www.wildopensource.com
David N. Welton
> One of the great things about Tcl is that the objects built under it
> really can become commoditizable. Download, unzip, "package
> require" and you're off . Because of the amount of discipline built
> into the interface structure, it really is pretty much plug and
> play. Might take some tweaking, but much less than other paradigms
> I've seen.
I find, that, unfortunately, a lot of good Tcl code is wrapped up in
applications. Often times very tightly. A lot of people seem to
start with a 'small project' that uses Tcl and Tk, and then just keep
going, without properly breaking the code into segments.
Ron Zajac
>
> Andrew Tannenbaum (t...@world.std.DELETE-THIS-TOKEN.com) wrote:
>
> : [snip]
>
> : On TV, it's usually done by careful selection of "sound bites."
> : They call this "freedom of the press" - they report what they want to,
> : and fairness and accuracy are not given high priority.
>
> Although I don't write for the technical computing press, I do have
> many dealings with journalists and broadcasters, and I do write not
> SNIPped]
...and many thanks for the hot tips, Steve!
-raz
--
Can't do you no harm
To feel your own pain. -John Lennon
Jeffrey Hobbs
>
> Les Cargill <lcar...@worldnet.att.net> writes:
>
> > One of the great things about Tcl is that the objects built under it
> > really can become commoditizable. Download, unzip, "package
> > require" and you're off . Because of the amount of discipline built
> > into the interface structure, it really is pretty much plug and
> > play. Might take some tweaking, but much less than other paradigms
> > I've seen.
>
> I find, that, unfortunately, a lot of good Tcl code is wrapped up in
> applications. Often times very tightly. A lot of people seem to
> start with a 'small project' that uses Tcl and Tk, and then just keep
> going, without properly breaking the code into segments.
Ah, but that is part of the unseen beauty of Tcl's greatness,
that can lead down the wrong path only because Tcl's true power
and strengths aren't seen at the outset.
Manager: I need this impossible task done in 2 weeks, or else!
Programmer: Uh, that's crazy, but I'll do my best. It'll be
command line only and not debugged, but I might be able to do it.
[Minor panic ensues while P does some soul-searching and net
searching, and finally remembers Tcl/Tk that they heard about
in passing]
P: Hmmm, this looks like it could save me a lot of time...
[3 days later, after learning a sufficient amount of Tcl/Tk and
programming up a solution]...
P: OK, here's what you asked for. I even managed to put a slick
GUI on it. It should even run on Windows without changes.
M: Wow. This will really advance my career. Now that you did
that so quick, why don't you add this, that, this, the other ...
...
A familiar story, with only minor modification, for many I'm sure.
So many projects start that way with Tcl, and rarely does one get
the time to reevaluate the code base, rewriting it for modularity
or maintainability (not that Tcl has a real problem there anyway),
etc. The lucky ones that get lead design time and/or are more
proficient with Tcl's abilities can do this, but I would bet that
is less than half the cases. So many Tcl projects start with
someone learning the language, that you can't expect them to
understand the Tao of Tcl before they start. I wonder what it
would be like to revisit my original Tcl work...
--
Jeff Hobbs The Tcl Guy
Senior Developer http://www.ActiveState.com/
Tcl Support and Productivity Solutions
Chang LI
In my experience Tcl especially Tk codes are difficult to
make them module and reuse. Design a megawidget is difficult.
It must have some other ways to write reuseable Tcl codes.
OXTcl is a solution for reuse?
Chang
Donal K. Fellows
> I wonder what it would be like to revisit my original Tcl work...
If mine is anything to go by, "embarrassing"... :^)
Donal.
--
Donal K. Fellows http://www.cs.man.ac.uk/~fellowsd/ fell...@cs.man.ac.uk
-- Naturally I will not be justifying this traditional argument with evidence
and rationale. -- Richard Thrippleton <re...@cam.ac.uk>
John E. Perry
> Chris Nelson wrote:
> >
> > Les Cargill wrote:
> > > ...
> > > You never see news reports that say "A missile was launched today in
> > > which nothing went wrong, noone was hurt and cost was contained.".
> >
> > I don't know; Mars Odyssey seemed to make the news. ;-)
> >
>
> True! Breath of fresh air, that.
>
Hmm. If there hadn't just been two consecutive spectacular and costly Mars
failures, I wonder how much attention it would have attracted...
John Perry
AS&M
Jeff Hobbs
> In my experience Tcl especially Tk codes are difficult to
> make them module and reuse. Design a megawidget is difficult.
> It must have some other ways to write reuseable Tcl codes.
Some people that learned CS from the OO model at the outset
find that an OO system is essential for such tasks. However,
I find that disciplined, designed coding can do just as well
to build large systems with reusable components. True, it
might be alleviated with something like xotcl, but I do not
think that is a requirement. There are many case examples
that support that (like Tcl itself).
That said, I'd still like to see OO become a more natural
component of Tcl, via XOTcl and/or [incr Tcl], so users that
prefer that don't have the desire to look elsewhere based on
an initial superficial analysis of what we all know is the
best language going. :)
Chris Nelson
Not much, I'm sure. :-(
--
Never doubt that a small group of thoughtful, concerned citizens
can change the world. Indeed, it's the only thing that ever has.
--Margaret Mead
Steve Blinkhorn
: Jeffrey Hobbs wrote:
: > I wonder what it would be like to revisit my original Tcl work...
: If mine is anything to go by, "embarrassing"... :^)
Or, just occasionally "was I really as clever as that once upon a time?"
The worst shock of such a kind was when I was checking some old video
cassettes to find something I recorded as teaching material in the mid
'80s. To my considerable alarm, in a programme I had forgotten I had
made, there was me 15 years younger with only the merest flecks of
grey, 20 pounds thinner, far more articulate.... doesn't bear thinking
about. Then there was the very neat variable-metric numerical
optimisation Fortran subroutine....
--
Steve Blinkhorn <st...@prd.co.uk>
Larry Smith
>
> Larry Smith wrote:
> > I think most people were, but today we are at the
> > tail end of a decades-long business cycle that has
> > shortened so drastically that nowadays management
> > only looks at the quarterly bottom line.
>
> The number of problems that can be solved in 90
> days is painfully short.
Oh, true enough. As a culture we no longer have
the ability to solve harder problems, though. The
limit is cultural, and appears in various forms in
many situations in America. Land of the Quick Fix,
that's us.
> > Sure Tcl
> > can save them 50% of their effort or more - but it
> > takes a month to get comfortable with it, and we
> > need things done NOW, not a month from now. And
> > so our entire culture, top to bottom, has become
> > the ultimate example of penny wise and pound foolish.
>
> The pennies will eventually dry up and those with
> the pounds win in the long term. Evolution is really
> hard to argue with.
Without question. Luckily, many of the countries that
would be most likely to be able to take advantage of it
have problems of their own, some of them self-inflicted.
But you are correct in that sooner or later some country
is going to manage to get the blinders off long enough
to steal a march on us and we'll go back to being a
banana republic. In a sense, that's good, since we run
the country like a banana republic anyway.
> > No one - and I mean NO one - outside of the open
> > source world is even remotely interested in long
> > term issues, maintainability, or stability. All
> > they care about is ship time.
>
> I respectfully diagree. It's a focus in open source
> because that's a cultural norm of those in open
> source, but the industry has to get back to realistic
> business models or it will simply stop being.
I agree with you about the consequences to the industry,
but I think I disagree that they "have" to get back to
realistic business models. When the entire culture is
handicapped, broken business models can remain viable
on a short-term horizon until something upsets the whole
applecart (they can be compared with "water monopoly"
empires). And even when that happens, well, in the US
major companies are not allowed to fail, the gov't will
find some way to pump money into them and keep them going
in the interests of "national security". Remember, this
is also the country that has raised deficit spending to
a high art. If we don't have the money, we'll just pretend
we do. It works, so long as most people don't understand
the chicanery and are willing to let their ignorance under-
write the scheme. It's a shell game, but it works. But
no, it's not smart. It's inherently unstable, and giving it
any semblence of long-term stability requires the sacrifice
of resources that could've been used for better things.
But when you no longer aspire to better things, hey, why not?
A fool and his money...
> If you're saying the present business model is broken,
> I'll agree. I don't know what will replace it.
We do agree. As to what will replace it, well, I don't
know. WE can speculate about what a post-capitalistic
economy might look like only because as open-source fans
we live in something like what might evolve (given the
technology and a way to deploy it, and with a sufficiently
educated population, some form of gift-based economy with
some capitalistic trimmings to deal with physical resources)
- but I tend to doubt we (as a culture) will ever achieve
something like that without vision and commitment. We
never will by doing what's expedient.
regards,
Bob Techentin
"Victor Wagner" <vi...@wagner.rinet.ru> wrote
>
> ...but if you have larger projects, even rudimentary
> ways of static checking which present in Perl, become a huge advantage.
>
> Unfortunately, Tcl has no equivalents of use strict, -wc and -T.
I do my static checking with procheck, instead of forcing the
interpreter/compiler to do that for me every time somebody runs my script.
:-)
Bob
--
Bob Techentin techenti...@mayo.edu
Mayo Foundation (507) 538-5495
200 First St. SW FAX (507) 284-9171
Rochester MN, 55901 USA
http://www.mayo.edu/sppdg/
Andreas Kupries
dav...@dedasys.com (David N. Welton) writes:
> I find, that, unfortunately, a lot of good Tcl code is wrapped up in
> applications. Often times very tightly. A lot of people seem to
> start with a 'small project' that uses Tcl and Tk, and then just
> keep going, without properly breaking the code into segments.
So, whoever is out there and has some in an application which is
generally usable, consider to extract the code out of your application.
Also look at tcllib if there is an existing module your code fits
into. If yes, register a patch in the tcllib SF project containing
your code. If not, register a patch containing your code and propose a
module for it.
--
Sincerely,
Andreas Kupries <akup...@shaw.ca>
Developer @ <http://www.activestate.com/>
Private <http://www.purl.org/NET/akupries/>
-------------------------------------------------------------------------------
}
lvi...@yahoo.com
According to Jeff Hobbs <Je...@ActiveState.com>:
:That said, I'd still like to see OO become a more natural
:component of Tcl, via XOTcl and/or [incr Tcl], so users that
:prefer that don't have the desire to look elsewhere based on
:an initial superficial analysis of what we all know is the
:best language going. :)
Speaking of this - what is the status of the TIP, submitted I believe July
2001, proposing that incr tcl become a part of the distributed tcl tar
file...
--
"I know of vanishingly few people ... who choose to use ksh." "I'm a minority!"
<URL: mailto:lvi...@cas.org> <URL: http://www.purl.org/NET/lvirden/>
Even if explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.
lvi...@yahoo.com
According to Andreas Kupries <akup...@home.com>:
:Also look at tcllib if there is an existing module your code fits
:into. If yes, register a patch in the tcllib SF project containing
:your code. If not, register a patch containing your code and propose a
:module for it.
So , what is the process of moving from the feature requests submission
to the code arriving in a tcllib tar near us? I see, looking at
http://tcllib.sf.net/, quite a number of feature requests, bugs, etc.
registered at tcllib. Is there perhaps a need for some more bodies from
the Tcl community to help get things done?
Mo
>
> According to Jeff Hobbs <Je...@ActiveState.com>:
> :That said, I'd still like to see OO become a more natural
> :component of Tcl, via XOTcl and/or [incr Tcl], so users that
> :prefer that don't have the desire to look elsewhere based on
> :an initial superficial analysis of what we all know is the
> :best language going. :)
>
> Speaking of this - what is the status of the TIP, submitted I believe July
> 2001, proposing that incr tcl become a part of the distributed tcl tar
> file...
TIP 50 passed but has yet to be implemented. Getting the Itcl build
system integrated with the Tcl build system is harder than it sounds.
http://tcl.activestate.com/cgi-bin/tct/tip/50.html
cheers
Mo DeJong
Victor Wagner
: "Victor Wagner" <vi...@wagner.rinet.ru> wrote
:>
:> ...but if you have larger projects, even rudimentary
:> ways of static checking which present in Perl, become a huge advantage.
:>
:> Unfortunately, Tcl has no equivalents of use strict, -wc and -T.
: I do my static checking with procheck, instead of forcing the
: interpreter/compiler to do that for me every time somebody
runs my script.
You have the point.
But note that these options do dynamic checks as well.
And these dynamic checks are far more neccessary for Tcl than for
Perl, becouse runtime code generation is way more usable in Tcl.
--
It's all magic. :-)
-- Larry Wall in <72...@jpl-devvax.JPL.NASA.GOV>
Kevin Kenny
> Unfortunately, Tcl has no equivalents of use strict, -wc and -T.
No *direct* equivalents, granted, but [::safe::interpCreate] comes closer
to my view of the world than -wc and -T do. Moreover, Safe Base supports
a configurable security model, so I can have exceptionally fine-grained
control over exactly what untrusted code is doing.
I'm not sure what a Tcl counterpart of 'use strict' might be. Tcl already
feels stricter than Perl.
--
73 de ke9tv/2, Kevin KENNY GE Corporate R&D, Niskayuna, New York, USA
Peter Lewerin
> to my view of the world than -wc and -T do.
Agreed. Also, IANAPH, but my understanding is that the tainting model
in Perl has non-obvious loopholes (I believe a foreach loop untaints
variables in the looped-over list -- don't quote me :-) ) and that it's
very tempting to untaint variables without properly checking their
content. Of course, if you know what you're doing it should do what
it's supposed to do.
> I'm not sure what a Tcl counterpart of 'use strict' might be. Tcl already
> feels stricter than Perl.
Thinking aloud...
strict refs: well, symbolic references is an integral part of Tcl
programming.
strict vars: nothing stops you from using undeclared variables, but Tcl
scoping is indeed stricter than Perl's.
strict subs: again, Tcl is stricter by default.
Cameron Laird
Steve Blinkhorn <st...@sole.prd.co.uk> wrote:
.
.
I'm catching on to a pattern here. Instead of complaining
about difficult deployment situations, Steve engineers and
shares his own installation solution. Rather than moaning
about the travails of publication, he offers useful advice
on ensuring successful outcomes. These are the sorts of
symptoms often seen in a man who actually accomplishes what
he sets out to do.
--
Cameron Laird <Cam...@Lairds.com>
Business: http://www.Phaseit.net
Personal: http://starbase.neosoft.com/~claird/home.html
Steve Blinkhorn
: In article <9r8mff$qnl$1...@fastnet.prd.co.uk>,
: Steve Blinkhorn <st...@sole.prd.co.uk> wrote:
: .
: .
: .
: >Although I don't write for the technical computing press, I do have
: >many dealings with journalists and broadcasters, and I do write not
(stuff removed)
: I'm catching on to a pattern here. Instead of complaining
: about difficult deployment situations, Steve engineers and
: shares his own installation solution. Rather than moaning
: about the travails of publication, he offers useful advice
: on ensuring successful outcomes. These are the sorts of
: symptoms often seen in a man who actually accomplishes what
: he sets out to do.
That's all deeply flattering, but so far as programming is concerned
I frequently stand on the shoulders of better brains than mine (I've
got my own MRI scan on one of my machines, so I know that *my* brain
has got shoulders, even if yours hasn't ;-)
Since I first got started with ARPAnet in the '70s I have been
consistently impressed by the willingness of fine brains to share
their wisdom, and I have to say that clt is a star performer, almost
invariably both erudite and kindly to beginners. And the interesting
stuff we're doing now is built on the ingenuity of Brent Welch and
Jean-Claude Wippler, and Vince Darley and Jeff Hobbs etc. etc. etc.
This is not a plug (it's not really meant for you lot) but
www.careerdemon.com instantiates a browserless secure client-server
model for individual assessment and guidance wrt careers, using
scripted documents, a modified tclkit and tclhttpd. (The
installation falls over on ME, but we've just found a fix for that).
--
Steve Blinkhorn <st...@prd.co.uk>
Jeff Hobbs
...
> This is not a plug (it's not really meant for you lot) but
> www.careerdemon.com instantiates a browserless secure client-server
> model for individual assessment and guidance wrt careers, using
> scripted documents, a modified tclkit and tclhttpd.
Yeah, I wasn't sure if this was a tclkit plug, or a recommendation
that we all need assessment and career guidance. ;^)
Chris Nelson
> ...
> The instances where stuff I have written has been heavily edited have
> always been when a new and inexperienced copy editor has had a rush of
> blood to the head. I've always, absolutely always, been able to
> reverse such changes. And one trick I discovered early on: always
> write as nearly exactly to the required length as possible, then if
> they cut stuff, they're short of copy.
On a greatly different scale, my Tcl/Tk Programmer's Reference was
contracted to be a maximum of, I think, 384pp. Part way through, it was
clear I was running long. There were a couple of suggestions on how we
could cut a subsection of each entry throughout the book and get to the
desired length. I resisted and said it was easier for me to write the
book "as long as it needed to be" and edit down later. Happily, while I
was wrapping up, they upped the cover price of the series and, with it,
the maximum page count. I ended up cutting nothing for length! :-)
Chris
--
Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety. -- Ben Franklin