Re: FORTH levels
- From: Aleksej Saushev <asau@xxxxxxxx>
- Date: Fri, 01 Feb 2008 20:57:30 +0300
Elizabeth D Rather <erather@xxxxxxxxx> writes:
Aleksej Saushev wrote:
Elizabeth D Rather <erather@xxxxxxxxx> writes:...
How can I test your words?You can learn a great deal about my company and its customers
from our web site, www.forth.com.
I explained, why I don't trust you. You're interested in presenting
everything is good in your area too much. I want _independent_ source.
Anton provides metrics on USENET groups, and Forth is beaten by Lisp
alone, TCL alone, and Python alone, John Doty refers to conferences,
I can try to perform search myself, so it is verifiable source too.
Your site is not.
And I know examples of good looking companies, which are corrupt inside.
I don't believe you at this distance, provide verifiable source
of information.
Of course we present favorable information about our products,
services, and customers. But we do not exaggerate or lie (nor
do other reputable companies). You are the first person who has
ever accused me of lying, and I resent it.
Product information may be verified by trying out free
downloads. If you wish, you can contact customers and verify
that they use our products, although for large companies it may
be hard to locate the actual group doing the development. On a
limited basis, our sales dept may put you in touch with
individuals if you have a legitimate need (e.g. you're
considering a similar project), but not to refute wild
accusations of lying.
I don't talk about lying, but the picture you present differs
from what I see and from what others can see, too much.
I can believe in the explanation, that you have found a unique
place, your little beautiful island ("the zoo", like John says)
where you can survive in relative isolation from other world,
you still have old custormers, everything runs fine. Yet your
island is not the whole world, and "pandas are endangered in the
wild."
Counting folks on USENET groups is an utterly useless means of
estimating the number of folks using a technology. Most
engineers are too busy doing their work to hang out on
newsgroups, and some of the most active groups are active
because their technology is so problematic that folks
desperately need help that they can't get from their providers.
It is _very_ useful, and it is perfectly _valid_.
When people do use programming language, they produce tons of code,
which is spread in various samples. If newcomers come and ask for
examples and receive code snippets or references to ready to run
libraries, it is good indication that the programming language
lives pretty well. If Forth libraries or code is deeply hidden
in commercial products, it indicates, that the code is too hard
to write to give it to others gratis.
Conferences are also unreliable metrics, because most attendees
are primarily concerned with results of their research rather
than programming languages used in instruments. Also, they
primarily reflect academic communities rather than the far
larger and more numerous companies developing hardware and
software products.
Right, attendees are interested in results of actual work, and
they impose strict limitations on the programming languages they
use. Almost noone will use a language, which is hard to use.
And if Forth has died out, so worse to it. That clearly indicates,
that Forth is a bad choice to perform tasks. To add to this,
researches can afford higher risks in choosing tools, if they
stopped using the tool, it means the latter is too hard to use,
and much better alternatives exist.
Some magazines, such as Embedded Systems Programming, take
occasional polls of languages in use by their readers, but they
also reflect a tiny subset of actual users of anything. How
many people here read ESP? And of those that do, how many
respond to surveys?
Surveys are _less_ reliable, indeed. Much less than conferences
and USENET.
Bottom line: there are no useful independent metrics.
It is obvious, you're wrong here. You can use any sensible measure,
the latter doesn't need to be ideal. And any realistic one shows
Forth is dead in comparison to C, C++, Java, Lua, TCL, Python.
C is
significantly more prevalent in embedded systems than Forth: I
have never denied that. However, we maintain a thriving
business with a large number of customers who express
considerable satisfaction with our products and services. In my
opinion, so long as companies such as FORTH, Inc. and MPE and
independent suppliers such as Marcel, Anton, and others thrive
and remain active it's wildly premature to declare Forth "dead".
Sorry? I don't see any real projects of independent suppliers mentioned,
last Gforth release was more than 4 years ago, the referenced projects
are more of academic interest. And I can't say any good word about iForth,
though printouts look impressive sometimes, and I can't deny, that Marcel
seems to be among rear ones here, who have sense of reality.
--
CKOPO CE3OH...
.
- Prev by Date: Re: .fs extension endangered
- Next by Date: Re: FORTH levels
- Previous by thread: Re: paul grahams arc, a new lisp, and his words about makign code shorter, vs forth?
- Next by thread: Re: FORTH levels
- Index(es):
Relevant Pages
|
