Zemax software
Myriam TUR
Currently, I initiate myself into ZEMAX software. I wonder wether somebody
knows this optical software and if she/he followed Optima-Research courses ?
Thank you very much.
Yours faithfully,
Myriam TUR
James R (Jim) Lynch III
copy of: "Introduction to Lens Design (With Practical ZEMAX Examples)", by
Joseph M. Geary, Willmann-Bell, Inc.(www.wilbell.com), ISBN 0-943396-75-1.
James R (Jim) Lynch III
Laurent d'Mascarenhas
"James R (Jim) Lynch III" wrote on his magical keyboard:
> For teaching yourself ZEMAX, I don't think you can do better than to get a
> copy of: "Introduction to Lens Design (With Practical ZEMAX Examples)", by
> Joseph M. Geary, Willmann-Bell, Inc.(www.wilbell.com), ISBN 0-943396-75-1.
Clever trick indeed to fool Mr. WCE and maybe redirect his rants somewhere
else, who knows.
But seriously, try also:
They have books on many other topics besides the dreaded Z...x software.
Ld'M XVII Himself
John Savard
<jrlyn...@comcast.net> wrote, in part:
>For teaching yourself ZEMAX, I don't think you can do better than to get a
>copy of: "Introduction to Lens Design (With Practical ZEMAX Examples)", by
>Joseph M. Geary, Willmann-Bell, Inc.(www.wilbell.com), ISBN 0-943396-75-1.
THERE is a reason for using Zemax if you don't need Code V.
Not only do many other people use it (but nearly all other packages
out there import Zemax lens designs), making them practically the
Microsoft of lens design software, but because of its popularity,
excellent textbooks in its use are out there.
http:/zemax.com/
But OSLO also has a company behind it, and there are companies
teaching courses with it.
http://www.sinopt.com/
http://www.lambdares.com/
And Code V is a very full-featured program, for large companies doing
very serious optical design work.
There are definitely other lens design programs out there. I've come
across three which are sold by businesses which at least don't admit
to being one-man shows:
OpTaLiX used to come with a free version that had an unlimited number
of surfaces, but which didn't optimize; that is no longer true, its
evaluation version is still like that, but now it runs for only a
limited time.
WinLens does have a similar evaluation version, WinLens 4.3, but that
version will not export lenses to any other format, unlike the OpTiX
evaluation version.
VOB is another little-known optical design program. It is offered in
versions billed as an inexpensive basic version, a standard version,
and a professional version. When I saw that only the professional
version did optimization, I was shocked. When I saw that the
professional version was an inexpensive $300, I stopped being shocked.
Yes, you *can* purchase a program that optimizes and is full-featured
for well under $1,000. However, they do fairly strongly advise people
to try the trial version first before ordering; whether that's just
because optical software is inherently complicated, or because it is
lacking something people expect to find on optical software due to
having heard of Zemax, whether trivial or important, I wouldn't know.
They do note that the optimization is local, not global, but global
optimization is sufficiently hard to pull off that I don't think too
many other programs offer that either.
I believe I may have heard of Dr. Volker Schmidt somewhere else.
And in my web searching, I've turned up yet another one, for $600,
which does damped least-squares and simulated annealing (this latter
one allows optimization to be other than local, so I've heard), and
it's even non-sequential:
Then there are the ones from independent programmers.
There is Don Dilworth's program SYNOPSYS (he is also the designer of a
very good wide-angle eyepiece, although it is about the size of a
Speers-Waler):
http://home.gwi.net/OSD/home.htm
And, of course, I hardly need to tell you about Enterprise, from Jim
Klein, a frequent participant in this newsgroup.
http://www.westcoastengineering.com/
And, finally, I will mention three examples of free optical design
software.
One of them is, I am afraid, not actually available. In a web search
on Seidel aberrations, I came across a defunct grad student website in
Germany, pointing to a nonexistent grad student website in Mexico,
referring to DASO II, a lens design program for Windows.
Looking again at the site, though, it is possible it was only a
third-order lens design program.
But if anyone knows someone over at INAOE who can say if either DASO
II or the original FORTRAN program DASO is available for distribution,
do let us know.
(You can find this URL on Google, but it isn't much use...)
The second doesn't appear to be finished yet. It draws great Bertele
curves, lens diagrams, and spot diagrams, but it doesn't appear
possible to edit a lens design by clicking on the spreadsheet cells
yet, nor could I find an optimization feature, but perhaps it does
work from the command line.
Still, its original author did a lot. There's even a group on Yahoo!
Groups about this, and another one about the BASIC compiler by the
same author. The BASIC compiler has an active development community.
Sadly, the optical design program does not. A few people expressed an
interest, but decided that it was a better idea to start a new
object-oriented and non-sequential program... of which nothing came.
http://www.maxreason.com/software/optics/opus.html
And the third is a DOS program that really does optimize. It seems to
be limited in the number of surfaces it can handle because of the fact
that it offers no way to advance past the last surface on the screen
rather than a program limitation.
It *was* at
http://members.aol.com/drowesmi/trace.html
but that URL is no longer valid, and Google turns up no other. It
showed three spot diagrams in three windows, and I suspect that its
optimization algorithm was simplistic in nature. But it still likely
would come in handy to amateur telescope makers, for example.
However, another program by Dave Rowe is available at
http://www.ceravolo.com/fringe/index.htm
I should warn you; the page displayed very badly in my browser, I had
to select the text to make it readable.
Still, since Mr. Rowe has been generous to the amateur astronomy
community, I expect that his software will again surface on the web
soon.
Also, on
http://www.considine.net/mac/vacpan.html
is a JavaScript calculator for Schmidt corrector plate making (the
secret sauce of the Schmidt-Cassegrain wars) based on formulas by Dave
Rowe.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
look into. One of them I had intended to mention; MODAS.
http://members.kabsi.at/i.krastev/about.html
http://members.kabsi.at/i.krastev/features.html
http://members.kabsi.at/i.krastev/download.html
http://members.kabsi.at/i.krastev/register.html
... the intended page,
http://members.kabsi.at/i.krastev/modas.html
has a blank colored frame on the left which is supposed to allow
navigation between those four pages, but which does not, and no other
links are provided.
Does he really think that anyone still surfs the Web with JavaScript
enabled any more?
I had thought that this program was shareware, with a fee to purchase
it, but it appears that I was mistaken, since the registration page
doesn't say anything about money.
But when I go to the download page, I see that the DOS version is
limited to 25 surfaces, and the Windows version to 4.
Presumably there IS a version of the Windows version that works with
more than 4 services, but for which there is a charge.
Sadly, it defeated my Internet skills to find information about it.
Then there's ATMOS
which I mistakenly thought only handled systems with mirrors... I
presumably confused it with a free optical design program that I
encountered a long time ago.
But this program is a very reasonable $120; the screenshots do show,
though, that its interface is different from that of most other
optical design programs.
And there's also the Beam 2, 3, and 4 series, from
http://www.stellarsoftware.com/
and many others.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
Laurent d'Mascarenhas
It's still listed on their website.
Ld'M
John Savard
(John Savard) wrote, in part:
>Sadly, it defeated my Internet skills to find information about it.
There _is_ a page which mentions that, at one time, the price for a
version of MODAS was $35, but it is unclear if that applies to the
current version, or something completely different. (It's at the
bottom of the news.html page. Higher up, there seem to be ordering
instructions, but I suppose one should E-mail first to check that his
address is still valid...)
Nor can I determine if the full version of MODAS for Windows goes to
25 surfaces like the DOS version billed as a full version, or if it is
more powerful.
I tried the DOS version, to see what it is like.
Hints:
The program won't read in lenses from the EXPORT directory, only the
EXAMPLES directory - and the lenses are all in the EXPORT directory,
so you will have to copy them over.
If you ask for a graphic display of the lens system, use the ESC key
to get back to the program.
Also, the DOS version installs by means of InstallShield... for
Windows. But I suppose you will be using a computer with Windows on it
to download it from the Internet (although I used to use a shell
account I could access from a DOS-only machine... BTW, some of the
lenses have long file names).
Although I make these minor criticisms, I think that Mr. Krastev has
still done a great service to those interested in optical design. I
hope he does well, selling copies of his program... but he will have
to make it a bit easier to find out how to purchase a copy!
I also hope that Jim Klein does well. I don't know if any code from
SIGMA did make its way into Zemax, but I do see that Zemax is billed
as having non-sequential capabilities (in the EE version), and global
optimization, and that Zemax EE is $1,000 more than Zemax XE, which
used to be the higher-end version, with Zemax SE being the basic one.
I certainly am saddened that one major optical design program has
disappeared, particularly since it was well-loved by its users. And I
too am puzzled how a $2,000 program could become popular.
On the other hand, paying every year for support is better than paying
every year for the program, since when you stop paying, at least you
still have the program.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
West Coast Engineering
>On Tue, 23 Mar 2004 16:57:51 GMT, "James R \(Jim\) Lynch III"
><jrlyn...@comcast.net> wrote, in part:
>
>>For teaching yourself ZEMAX, I don't think you can do better than to get a
>>copy of: "Introduction to Lens Design (With Practical ZEMAX Examples)", by
>>Joseph M. Geary, Willmann-Bell, Inc.(www.wilbell.com), ISBN 0-943396-75-1.
>
>THERE is a reason for using Zemax if you don't need Code V.
>
>Not only do many other people use it (but nearly all other packages
>out there import Zemax lens designs), making them practically the
>Microsoft of lens design software, but because of its popularity,
>excellent textbooks in its use are out there.
>
That must make Ken Moore is our own Bill Gates.
Gives me goose bumps. :-)
West Coast Engineering
and my user support includes help and guidance rebuilding the program
and getting started modifying it into your own code. You get to rename
it, make it your own in-house code (be your own support for this
modified version) and even give it away or re-sell under its new name.
Compaq Visual Fortran and Winteracter are required. When I switch to
Intel Fortran 8 later this year, ports to LINUX will be trivial using
the LINUX version of the compiler and the LINUX version of
Winteracter.
Put that in your pipe and smoke it
WCE
John Savard
<bpa...@auvergne.metaphysique.org> wrote, in part:
> ACCOS V should in principle be still available from OPTIKOS, AFAIK.
>It's still listed on their website.
It is? A search turned up programs that import from ACCOS V, but at
www.optikos.com, many products and services are advertised, but the
company does not even appear to sell lens design software of any
description.
They seem to be now confining themselves to optical testing and
metrology. OpTest, VideoMTF, and EROS appear to be their current
optical software products.
They may have decided that sales volume didn't justify the effort of
continuing to market, enhance, and support the product.
Apparently, though, ACCOS V *was* one of the major optical design
programs, right up there beside CODE V, from all the references I am
seeing to it.
But when you say "in principle", perhaps you recognize this, and it is
mentioned somewhere else on the web site, such as in a description of
the history of the company.
There definitely are others I haven't mentioned.
dbOptic appears not to do automatic optimization, and it is $89.
OPTEC-IV is another one I overlooked. However, the web site says it
will _soon_ be available.
There's also an interesting free program called Snell Trace. It only
does _meridional_ rays, but it uses Snell's Law, not an approximation.
Thus, it may be actually _more_ useful than a full featured optical
design program for one specific application of some interest to me:
input a prescription for an existing lens from a patent or wherever,
and draw a cute-looking ray trace to stick on a web page.
http://www.velocity.net/%7Etrebor/List.html
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
Steve Eckhardt
jsa...@excxn.aNOSPAMb.cdn.invalid says...
>
>On Tue, 23 Mar 2004 22:58:52 -0500, "Laurent d'Mascarenhas"
><bpa...@auvergne.metaphysique.org> wrote, in part:
>
>> ACCOS V should in principle be still available from OPTIKOS, AFAIK.
>>It's still listed on their website.
>
>It is? A search turned up programs that import from ACCOS V, but at
>www.optikos.com, many products and services are advertised, but the
>company does not even appear to sell lens design software of any
>description.
Groot Gregory, who now works at Lambda Research, still supports ACCOSV for
Optikos in his spare time. You'd have to ask him how to get the program.
>There's also an interesting free program called Snell Trace. It only
>does _meridional_ rays, but it uses Snell's Law, not an approximation.
>Thus, it may be actually _more_ useful than a full featured optical
>design program for one specific application of some interest to me:
Huh? I think I'm missing something here. Every full featured optical design
program out there uses Snell's law for every ray it traces. It is, of course,
reformulated in terms of direction cosines for simplicity with skew rays and
computational speed, but it's still there.
>John Savard
>http://home.ecn.ab.ca/~jsavard/index.html
--
Best regards,
Steve Eckhardt
skeckhardt at mmm dot com
John Savard
wrote, in part:
>Huh? I think I'm missing something here. Every full featured optical design
>program out there uses Snell's law for every ray it traces. It is, of course,
>reformulated in terms of direction cosines for simplicity with skew rays and
>computational speed, but it's still there.
I did not mean to imply otherwise. Of course full-featured optical
design programs use Snell's Law, even if they may do _preliminary_
optimizations using Gaussian, Seidel, and fifth-order optics. Some
even do seventh-order optics - to do that, though, they have to
prgress from one surface to the next just as they would for full
Snell's Law raytracing.
When I said "more useful", I simply meant that since it only did _one
thing_, the interface might be simpler.
Actually, because of the way one has to define the rays desired, I
would have to do the kind of diagram I'm thinking of in two passes
with that program; and a full-featured optical design program makes it
easier to pick appropriate diameters for each element. Also, it
doesn't handle mirrors. So it doesn't seem to be as interesting as I
had hoped.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
wrote, in part:
>Groot Gregory, who now works at Lambda Research, still supports ACCOSV for
>Optikos in his spare time. You'd have to ask him how to get the program.
But _they_ sell OSLO. Wouldn't that be...
Oh, no. First Zemax took SIGMA out behind the barn... and now, OSLO
has done the same to ACCOS V? Although, since Optikos is still in
business, I don't see how that could have worked that way.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
<westcoaste...@westcoastengineering.com> wrote, in part:
Since you have, with some justification, interpreted my query as a
veiled criticism of some of your previous comments, although I was
also thinking of CODE V there, I suppose I should say something here.
Frankly, I don't feel like criticizing you at all. I think you deserve
praise for developing an optical design program which you are making
available, with only slight limitations, widely available free of
charge.
Also, I confess to having misunderstood your pricing policy, based on
my interpretation of one of your recent posts. Looking at your web
site, I see that there isn't an up-front charge of $2,600 for the
upgrade plus the yearly charge; since the yearly charge is comparable
to what Zemax charges for support only, I presume it is also very
considerably lower than whatever CODE V leases for.
If it isn't a violation of the license terms for someone who has
purchased a year of support for the source code license at $260 to
continue using a rebuilt version at 500 surfaces after the first year
without support, that's great. Or is it? I suppose you're being
protected against being ripped off in this way by the cost of Compaq
Visual Fortran and Wintertacter.
I don't want to search for some loophole in your license terms by
means of which I can rip you off. I want you to succeed and make lots
of money.
As a private individual, I can speak my mind freely on USENET, which I
do on controversial political topics elsewhere.
Corporations, in their official statements, tend to be circumspect.
For one thing, they are worried about getting sued by their vicious
and underhanded competitors, who would jump at the chance to drive the
competition out of business.
You can see where this is leading.
Potential customers for Enterprise are quaking in their boots,
thinking that instead of you *merely* getting hit by a bus, the good
people at Zemax will sue you for millions of dollars you don't have,
one day, and your only escape will be to sign a consent decree in
which you will burn all the copies of your source code.
At least you have escaped the danger of a cease-and-desist letter from
Warner Brothers. Unless you already got one.
Then, there's the small matter of any nondisclosure agreements you
might have signed while working at Bausch and Lomb... since you
recently noted in a post here that your product is a HEXAGON clone.
Unless I misunderstood that post, too.
I am not personally offended by any frank comments you may have made,
but since you fail to see why they may be terrifying some people, I
have to be a bit worried. But then, maybe I'm just failing to
recognize what ought to be obvious from the beginning, that you're
"not in it for the money".
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
DonJan
ACCOS is used by a limited number of aerospace companies such as
Lockheed Martin of Orlando, Raytheon ... . They have custom macros for
this code that are usefull for missle design, trackers ... Optikos
probably gets funds from them for fixing various bugs, changing
computers.... I don't think that it has been in the same league as
Code V for many years.
jsa...@excxn.aNOSPAMb.cdn.invalid (John Savard) wrote in message news:<4061a1a...@news.ecn.ab.ca>...
Brian B
news:107cd10.04032...@posting.google.com...
> DonJan
>
> ACCOS is used by a limited number of aerospace companies such as
> Lockheed Martin of Orlando, Raytheon ... . They have custom macros for
> this code that are usefull for missle design, trackers ...
ACCOS V has an interesting history. It was written by James Spencer while he
was at IBM Research Labs, and he left there and set up Scientific
Calculations Inc. It was well ahead of its time. Even though
it was written in FORTRAN, the input on the teletype was completely free
format. CSS Inc offered it
as on online service to customers in Europe through their bureau in London,
linked to an IBM 360 in Norwalk Conn. Given that the access we had was 10
cps (8 hole paper tape for storage - we are talking 1973 here), and plotters
were a thing of the future, it was a huge leap forward then for us to have
what we now all take for granted - accurate diffraction-calculated MTF
calculations of systems with any number of tilts and aspherics. What made
this all far less excruciating was an excellent macro language - again way
ahead of its time.
Eventually plotters came in and of course it eventually graduated to Unix
workstations, but by this time the cost of supporting it probably exceeded
the income from its users. It is still used in the UK, mainly by defence
companies. But its influence was wider than one might think. The terminology
and sign conventions it established way back then for some of the basic
commands (RTG, PIKUP, SLV, TILT, ALPHA, AD etc) are still used in OSLO,
which saved us all having to re-learn new commands. Even more so, I believe,
in the case of WCE Enterprise software - the command syntax is totally
compatible with ACCOS V.
Can anyone date the first issue of ACCOS V (or IV, III, II or I ;)? And does
that make it the grand-daddy of all software still in use today? And what
became of James Spencer?
Brian
James R (Jim) Lynch III
Dr. Gordon Spencer wrote ACCOS and founded Scientific Calculations, Inc.
West Coast Engineering
No violation at all, but as far as I know, no one has ever done it.
> I suppose you're being
>protected against being ripped off in this way by the cost of Compaq
>Visual Fortran and Wintertacter.
It is not really that bad. Combined, the cost of CVF and Winteracter
is less than the cheapest ZEMAX. CVF is a bullet proof compiler and
Lawson Wakefield's support for Winteracter is nothing less than
perfect.
Fortran is as simple as BASIC. My 80 year old father-in-law, who
learned Basic at 77, converted his financial annalysis code to Fortran
in 1990. He took classes at Radio Shack. He knew nothing of computers
or programming before that. I once thought Fortran was so easy, I
could teach it to my German Shepard, but he said he didn't have the
time.
>
>I don't want to search for some loophole in your license terms by
>means of which I can rip you off. I want you to succeed and make lots
>of money.
I had those dreams too. How could anyone pass on a full up optical
design program, based on ACCOS-V and HEXAGON which included source
code for $260.00 and which allowed the user to modify and expand the
code to his or her delight and then turn around and use it either as
an in-house code so as to get the edge on the competition or to sell
it for profit? Apparently I was really stupid.
Once you start to modify the code, you are responsible for user
support of that code. If you are still under yearly $260.00 user
support for ENTERPRISE, I answer all questions, including questions
concerning my source code.
>
>As a private individual, I can speak my mind freely on USENET, which I
>do on controversial political topics elsewhere.
>
>Corporations, in their official statements, tend to be circumspect.
I'm a one man show too and I'm sure I have alienated more than a few
potential customers. Then again, I don't think you will find anyone
who bought ENTERPRISE and who took advantage of my user support,
complain that I didn't give them world class user support.
>
>For one thing, they are worried about getting sued by their vicious
>and underhanded competitors, who would jump at the chance to drive the
>competition out of business.
I have no competition, except for CODE-V.
>
>You can see where this is leading.
>
>Potential customers for Enterprise are quaking in their boots,
>thinking that instead of you *merely* getting hit by a bus, the good
>people at Zemax will sue you for millions of dollars you don't have,
>one day, and your only escape will be to sign a consent decree in
>which you will burn all the copies of your source code.
1. There are no good people at ZEMAX. I personally believe they are
all evil.
2. I don't own a house. I'm worth about 100K. If someone messed with
me the way you describe I'd tell them to "kiss it" and then the source
would go puplic.
3. I will with my own two hands, burn the source, 10 days after I'm
dead. :-) My wife has instructions to distribute it for free before
that. By then, however, it will be a little late to ask me to explain
the operation of a function or subroutine. :-)
>
>At least you have escaped the danger of a cease-and-desist letter from
>Warner Brothers. Unless you already got one.
4. Time-Warner couldn't find it in the dark with a flashlight and
instructions. I did consider using the name KODE-10. Klein's Optical
Designer Evaluator 10 with the sales pitch "What comes after CODE-V"
but I like those guys too much.:-)
>
>Then, there's the small matter of any nondisclosure agreements you
>might have signed while working at Bausch and Lomb... since you
>recently noted in a post here that your product is a HEXAGON clone.
>Unless I misunderstood that post, too.
5. In 1987, I told my X-Hughes buddies what I was working on. They
said, we don't care. No one cares about anything (Philbin quote in
Phatom of the Paradise)
>
>I am not personally offended by any frank comments you may have made,
>but since you fail to see why they may be terrifying some people, I
>have to be a bit worried. But then, maybe I'm just failing to
>recognize what ought to be obvious from the beginning, that you're
>"not in it for the money".
Of course I'm in it for the money. I'm just no good at selling warm
clothes to freezing naked people at the South Pole. :-)
WCE
>
>John Savard
>http://home.ecn.ab.ca/~jsavard/index.html
West Coast Engineering
>"DonJan" <don.ja...@itt.com> wrote in message
>news:107cd10.04032...@posting.google.com...
>> DonJan
>>
>> ACCOS is used by a limited number of aerospace companies such as
>> Lockheed Martin of Orlando, Raytheon ... . They have custom macros for
>> this code that are usefull for missle design, trackers ...
>
>ACCOS V has an interesting history. It was written by James Spencer
Gordon Spencer
Gordon Spencer is alive but not doing optical design.
>
>Brian
>
Skywise
<Snipola>
> Potential customers for Enterprise are quaking in their boots,
> thinking that instead of you *merely* getting hit by a bus, the good
I seem to recall some many moons ago something being mentioned
by those at WCE (although a different name at the time) that in
the event of the demise of the software's creator, that arrangements
have been made to make the source code 100% public domain.
> people at Zemax will sue you for millions of dollars you don't have,
> one day, and your only escape will be to sign a consent decree in
> which you will burn all the copies of your source code.
I have a suspicion that if such an event were to take place, that
somehow the complete source code would somehow appear posted on
many different websites and newsgroups by mulitple anonymous
parties thereby negating the effort of any such legal action.
This is what happened when RSA tried to sue the creator of PGP.
<Snipola>
<rant>
Some of the best software I have used (quality, features, stability,
ease of use) has been by individuals or small collaborative groups
and is almost always open source and free or very cheap, or both.
There's a couple of programs I use for the security of my system
that I can trust for the simple fact that the full source code is
available on their websites.
With all the _billions_ of dollars at microshits disposal why are
they incapable of producing one piece of software that actually
works right. Can you imagine the backlash against M$ if people
were to see their source code?
Since the OS is integral and fundamental to the security of the
computer, how can one trust the software when the source code
is guarded like it were the Presidents Football?
BUT, since 95%+ of the people I encounter are gullible shipdits(!)
that won't lift a finger to solve the wrongs of the world for
fear of offending someone, the likes of M$ will thrive.
I guess my point is that even though I do not use optical design
software I really appreciate the way WCE conducts their business
and I wish more companies worked like that.
</rant>
Brian
--
http://home.earthlink.net/~skywise711/LasersOptics/laser.html
"Great heavens! That's a laser!"
"Yes, Dr. Scott. A laser capable of emitting a beam of pure antimatter."
John Savard
wrote, in part:
>And what
>became of James Spencer?
He seems to have changed his name to Gordon H. Spencer.
But you still knew more about this than I did. After a brief web
search, other than that correction, I find that Gordon H. Spencer also
made a very fundamental contribution to optical design software,
inventing the Lagrange multiplier method, which is one of the two
fundamental approaches to optimization used in optical design programs
- the other being damped least-squares, introduced to optical design
by C. G. Wynne.
Also, I see that I did misunderstand Mr. Klein; apparently,
Enterprise, like Hexagon and Optima, is based on ACCOS, so it is not
really the case that Enterprise is based on this secret internal
program Hexagon.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
$199.95 and which, while lacking exotic features like non-sequential
ray tracing, is fairly complete. It does lack Buchdal (fifth-order)
aberrations, apparently (I could have misunderstood the feature list),
but everything else seems to be in order.
http://www.diginaut.com/shareware/ados/
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
West Coast Engineering
>jsa...@excxn.aNOSPAMb.cdn.invalid (John Savard) wrote in
>news:4061bc32...@news.ecn.ab.ca:
>
><Snipola>
>> Potential customers for Enterprise are quaking in their boots,
>> thinking that instead of you *merely* getting hit by a bus, the good
>
>I seem to recall some many moons ago something being mentioned
>by those at WCE (although a different name at the time) that in
>the event of the demise of the software's creator, that arrangements
>have been made to make the source code 100% public domain.
100% correct
>
>
>> people at Zemax will sue you for millions of dollars you don't have,
>> one day, and your only escape will be to sign a consent decree in
>> which you will burn all the copies of your source code.
>
>I have a suspicion that if such an event were to take place, that
>somehow the complete source code would somehow appear posted on
>many different websites and newsgroups by mulitple anonymous
>parties thereby negating the effort of any such legal action.
100% correct
>
>This is what happened when RSA tried to sue the creator of PGP.
>
><Snipola>
><rant>
>Some of the best software I have used (quality, features, stability,
>ease of use) has been by individuals or small collaborative groups
>and is almost always open source and free or very cheap, or both.
>
>There's a couple of programs I use for the security of my system
>that I can trust for the simple fact that the full source code is
>available on their websites.
>
>With all the _billions_ of dollars at microshits disposal why are
>they incapable of producing one piece of software that actually
>works right. Can you imagine the backlash against M$ if people
>were to see their source code?
>
:-)
>Since the OS is integral and fundamental to the security of the
>computer, how can one trust the software when the source code
>is guarded like it were the Presidents Football?
Right on!
>
>BUT, since 95%+ of the people I encounter are gullible shipdits(!)
>that won't lift a finger to solve the wrongs of the world for
>fear of offending someone, the likes of M$ will thrive.
Hammer on the head of the nail.
>
>I guess my point is that even though I do not use optical design
>software I really appreciate the way WCE conducts their business
>and I wish more companies worked like that.
I do too.
WCE
></rant>
>
>Brian
West Coast Engineering
out there.
WCE
Laurent d'Mascarenhas
"John Savard" <jsa...@excxn.aNOSPAMb.cdn.invalid> wrote in message
news:4061a1a...@news.ecn.ab.ca...
> On Tue, 23 Mar 2004 22:58:52 -0500, "Laurent d'Mascarenhas"
> <bpa...@auvergne.metaphysique.org> wrote, in part:
>
> > ACCOS V should in principle be still available from OPTIKOS, AFAIK.
> >It's still listed on their website.
>
> It is? A search turned up programs that import from ACCOS V, but at
> www.optikos.com, many products and services are advertised, but the
> company does not even appear to sell lens design software of any
> description.
>
> They seem to be now confining themselves to optical testing and
> metrology. OpTest, VideoMTF, and EROS appear to be their current
> optical software products.
>
> They may have decided that sales volume didn't justify the effort of
> continuing to market, enhance, and support the product.
>
> Apparently, though, ACCOS V *was* one of the major optical design
> programs, right up there beside CODE V, from all the references I am
> seeing to it.
>
> But when you say "in principle", perhaps you recognize this, and it is
> mentioned somewhere else on the web site, such as in a description of
> the history of the company.
>
I also think that it's on their backburner but it's mentioned on their
website in the category products.
I also remember an ad a while back (OPN, OE, Photonics Spectra???) which
said: "Home of ACCOS V".
Or something to that effect.
There is also a pdf brochure available on the site.
Try to download the brochure.
For some reason I can't seem to complete it.
Maybe my connection or more likely Bill Gates' XP.
Thanks for starting an interesting thread.
Ld'M
Laurent d'Mascarenhas
Oops, I meant being a major contributor ;-)
> Ld'M
West Coast Engineering
ACCOS-V is still supported but not sold new.
HEXAGON and OPTIMA (HUGHES and LOCKHEED)
are super sets of ACCOS-V built on ACCOS source code.
My program, ENTERPRISE, is a superset of all of these. It is all new
source code (reverse engineered from a HEXAGON manual) but is
functionaly a replacement for all of them. Hughes/Raytheon and
Lockheed have been offered free versions of ENTERPRISE but there has
been no response. Duh!
ENTERPRISE is under constant development. It is well supported. It is
supplied with 100% full source code including build scripts. Including
the price of the compilers, it is cheaper than the cheapest version of
ZEMAX.
You can modify it. You can expand it. You can make it your in-house
code. You can rename it, convert it and even sell it with no strings
attached.
Do I need to translate this into FRENCH or are you just too stupid to
understand?
How dumb is the optical design community? I offer a full optical
design code. It comes with full source code and user support. It is
dirt cheap. It comes with source code. And no one is interested?
And there is a free version, only restricted by surface #s and macro
length, for those interested in trying it out first.
I answer all email, even from those using the free version.
On top of this, everyone buys ZEMIN and defends it in the face of
bugs, an unresponsive development staff (one guy) and user support
only equaled by the user support of QUARK EXPRESS.
Oh, and no hope of ever seeing one line of ZEMAX user support.
You guys are MS brain washed.
Do I need to offer it to SPACE ALIENS to get an interested group to go
for it or what?
DUUUUH!
WCE
Juan Rayces
> that make it the grand-daddy of all software still in use today? And what
> became of James Spencer?
>
> Brian
If it helps, ALEC was the name of the software before it was changed
to ACCOS and it was dated about 1963, the time when Spencer published
"A Flexible Automatic Lens Design Procedure" in Applied Optics, Vol.
2, No. 12.
If you want you can rightly call Spencer (Gordon, not James) the
grand-daddy of all software that uses the constrained damped least
squares method of optimization. But the true ancestor of most present
day lens design software, including Spencer's, was Levenberg (K.
Levenberg, Quart. Appl. Math., 2, 258, 1944) who introduced a damping
device into the method of least squares. There are, though, other
design methods that do not use least squares; the family tree of lens
design programs is big indeed, and very interesting. J. Rayces
matt
"West Coast Engineering" <westcoaste...@westcoastengineering.com>
wrote in message news:krm4605a2j0nbtj40...@4ax.com...
if you really wanted to sell your software, you'd hire a salesman .
Managers and purchasing departments understand salespeople a lot better than
they understand cynical engineers . And stop acting like a rebellious
teenager . Say please and thank you even if there's nothing to thank for .
Take a Dale Carnegie course . Smile and wear a suit and tie. Write a
business plan and find some investors . Of course, while doing all these
things, stop writing software . Hire some warm body to answer the phone and
read from the provided scripts, user manual or other documentation rather
than spend valuable engineering time on customer support ;-) you know how it
goes...
best regards,
matt tudor
John Savard
in part:
>If you want you can rightly call Spencer (Gordon, not James) the
>grand-daddy of all software that uses the constrained damped least
>squares method of optimization. But the true ancestor of most present
>day lens design software, including Spencer's, was Levenberg (K.
>Levenberg, Quart. Appl. Math., 2, 258, 1944) who introduced a damping
>device into the method of least squares.
Yes, Levenberg invented damped least-squares.
It's Charles Gorrie Wynne who was the grand-daddy of all software
using the damped least squares method of optimization.
Gordon H. Spencer is the grand-daddy of the *other* important
method... Lagrange multipliers, which is an _alternative_ to damped
least squares. Or at least that's what I've understood from what I've
been seeing.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
West Coast Engineering
>
>if you really wanted to sell your software, you'd hire a salesman .
>Managers and purchasing departments understand salespeople a lot better than
>they understand cynical engineers . And stop acting like a rebellious
>teenager . Say please and thank you even if there's nothing to thank for .
>Take a Dale Carnegie course . Smile and wear a suit and tie. Write a
>business plan and find some investors . Of course, while doing all these
>things, stop writing software . Hire some warm body to answer the phone and
>read from the provided scripts, user manual or other documentation rather
>than spend valuable engineering time on customer support ;-) you know how it
>goes...
>
>best regards,
>matt tudor
Hi Matt,
Managers and purchasing agents only come in after the engineer asks to
buy something and all my prices are in the petty cash range and NO I
won't raise my prices.
I like writing code and helping people use it. That is the whole point
for me in creating ENTERPRISE. Except where ZEMIN comes in to play, I
always say PLEASE and THANKS.
I stopped wearing suites 8 years ago. I found that the tie cut off the
blood flow to my brain and I became as intelligent as a manager. It
was blissful but my code production dropped way off.
Carnegie! Didn't he treat all the people that worked for him like
slaves? Oh, maybe that was Andrew.
I did consider having a single drop dead gorgeous blond female Ph D
optical designer (seven of nine type) in a marginal tight fitting
miniskirt attend SPIE conferences to sell ENTERPRISE but then found
out that single, drop dead gorgeous blond females (at least among
human beings) and a Ph D in Optics were mutually exclusive.
It is the Big B, little b inverse relationship or something like that
which Dave Akins explained to me once.
Well, there goes the other 51% of my potential market. :-)
Someday, someone will be successful at using SEX to sell optical
design software. I hope I'm not too old to be aroused when they
succeed.
Yes, I'll rename my program VIAGRA, the sonic molecularly charged
optical design program. I can hide lude graphics in it which pop up
whenever the merit function gets really close to zero.:-)
Kind of like the naked zebra lady in Mac Write circa 1989.
WCE
>
Jamie Carter
Have you tried the Watcom FORTRAN compiler that is available now, as
Open Watcom, for free at http://www.openwatcom.org/? It supports many
platforms but I don't think that it supports Linux (yet). If
Winteracter was compatible with this very powerful tool (I use the
C/C++ for a variety of lab applications and find it much better than
the Visual C++ from you know who), I would happily buy Winteracter and
dive into Enterprise customization.
Regarding the Digital/Compaq Visual FORTRAN, do you use the IMSL
libraries, thus requiring the "Pro" version ($720)? I guess this
question is answered implicitly in your post about moving to Intel's
FORTRAN as version 8 is available now but version 8 "Pro" with the
IMSL and other features from the Compaq Visual FORTAN is not available
until later this year...
It seems to me that the infratstucture for "rolling" our own
customized Enterprise editions is easily over $1000 not to mention the
time required to "dig" into the code. Maybe that's why no one has had
the time or money to do so. If they're anything like me, my free time
is scheduled sometime in late 2008.
How hard do you think it would be to port the IMSL calls to Numerical
Recipe (http://www.library.cornell.edu/nr/bookcpdf.html) algorithms
(these, I got in C and FORTRAN)?
I don't think I am alone when I express a sincere desire to see
Enterprise developed in tools and libraries that are open and
available without expense that greatly exceeds the cost of your
license. And think of the money that you would get to keep as well.
Just my 2cents...
Jamie
West Coast Engineering <westcoaste...@westcoastengineering.com> wrote in message news:<4v7360plrol6n5gfm...@4ax.com>...
West Coast Engineering
>WCE,
>
>Have you tried the Watcom FORTRAN compiler that is available now, as
>Open Watcom, for free at http://www.openwatcom.org/? It supports many
>platforms but I don't think that it supports Linux (yet). If
>Winteracter was compatible with this very powerful tool (I use the
>C/C++ for a variety of lab applications and find it much better than
>the Visual C++ from you know who), I would happily buy Winteracter and
>dive into Enterprise customization.
Watcom is only Fortran 77 and I have been using 90 features for
several years so I'mm going to need to stay with a 90 compiler.
>
>Regarding the Digital/Compaq Visual FORTRAN, do you use the IMSL
>libraries, thus requiring the "Pro" version ($720)?
No, I do not use IMSL so you don't need the PRO version
>I guess this
>question is answered implicitly in your post about moving to Intel's
>FORTRAN as version 8 is available now but version 8 "Pro" with the
>IMSL and other features from the Compaq Visual FORTAN is not available
>until later this year...
>
>It seems to me that the infratstucture for "rolling" our own
>customized Enterprise editions is easily over $1000 not to mention the
>time required to "dig" into the code. Maybe that's why no one has had
>the time or money to do so. If they're anything like me, my free time
>is scheduled sometime in late 2008.
Still cheaper than ZEMIN.
>
>How hard do you think it would be to port the IMSL calls to Numerical
>Recipe (http://www.library.cornell.edu/nr/bookcpdf.html) algorithms
>(these, I got in C and FORTRAN)?
? I never used IMSL for anything.
>
>I don't think I am alone when I express a sincere desire to see
>Enterprise developed in tools and libraries that are open and
>available without expense that greatly exceeds the cost of your
>license. And think of the money that you would get to keep as well.
All people need to do is do it. Compared to all other optical design
solutions, it is cheap and the compiler is a pretty much one time
thing and updates to Winteracter are about $250 or and are available
every year or so.
1. Install CVF
2. Install Winteracter
3. Buy and install the source for ENTERPRISE
4. Go to the main directory and type MAKEENT.
All source is in the SRC directory. Editing the resource for the GUI
is done with WIDE (Winteracter Dev Env).
The whole process takes maybe an hour the first time and way less
after that. On a P4 machine, the program links in a few seconds. A
full recompile takes a few min and the AUTOMAKE only recompiles what
was changed.
There are no catches. I am really offering the full source code to a
professional level design code for $260.00.
WCE
John Savard
wrote, in part:
>I don't think I am alone when I express a sincere desire to see
>Enterprise developed in tools and libraries that are open and
>available without expense that greatly exceeds the cost of your
>license. And think of the money that you would get to keep as well.
I have to admit, though, that given the terms he says he licenses it
under, this could increase his exposure to software piracy.
Of course, for people who are waiting with bated breath for LensMaster
3000 (choose between optimization algorithms named after famous lens
designers which are weighted to imitate their design styles!), other
applications are doubtless more interesting than Enterprise.
As he has himself admitted, noting how well Zemax, his bete noire,
lends itself to people shooting themselves in the foot.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
West Coast Engineering
>On 25 Mar 2004 06:35:33 -0800, jaca...@onebox.com (Jamie Carter)
>wrote, in part:
>
>>I don't think I am alone when I express a sincere desire to see
>>Enterprise developed in tools and libraries that are open and
>>available without expense that greatly exceeds the cost of your
>>license. And think of the money that you would get to keep as well.
>
>I have to admit, though, that given the terms he says he licenses it
>under, this could increase his exposure to software piracy.
Gee, really. I never thought of that.
Seriously, when someone sells source code and places no restrictions
on what may be done with it, you would think that the issue of piracy
was a non-issue to the person selling it. Then again, maybe when a=b
and a=c, it is not obvious that b=c.
Steve Eckhardt
jsa...@excxn.aNOSPAMb.cdn.invalid says...
No, first OSLO took GENII behind the barn in Nov '93... It was years later
when Zemax attempted to take over support for SIGMA and found that most of the
users just wanted a cheap way to switch to Zemax. This was not due to
any deficiency in SIGMA, rather it was to obtain a program for which they
could expect to find continued support. (Most of us, with the exception of
Mr. WCE, do get good support for Zemax.)
The situation with ACCOSV is different. It's dying a natural death due to
starvation, as did SODA, Willey's program and a host of others. Groot was
just supporting the last remaining users.
West Coast Engineering
>could expect to find continued support. (Most of us, with the exception of
>Mr. WCE, do get good support for Zemax.)
>
If so, good.
WCE
John Savard
<westcoaste...@westcoastengineering.com> wrote, in part:
>Seriously, when someone sells source code and places no restrictions
>on what may be done with it, you would think that the issue of piracy
>was a non-issue to the person selling it. Then again, maybe when a=b
>and a=c, it is not obvious that b=c.
Actually, very little is obvious.
For example, with release 7.54, monthly support contracts were
abolished; with release 7.55, they're back. Still, this is a price
reduction, which deserves praise; I see, according to your web page,
as compared to a recent post, that you've added phone support back in
to what people get for $100 a month.
Even the most basic facts about your product, therefore, are a little
confusing. Of course, Ivan Krastev, for example, appears to have done
worse in that department.
Having researched what's out there in optical software, I can
understand why you are quite proud of Enterprise. It really does
include features you won't see anywhere else this side of CODE V.
And on top of that, you've made it available free of charge to anyone
in a version with a very generous 25 surface restriction.
It would be nice, for example, if the menu item "IMPORT a ZEMAX Lens
as the Current Lens" pulled up a standard Windows dialogue box,
letting you pull that lens from any directory on any drive, but
prettying-up the interface is the kind of work that people do tend to
expect to be paid for, and I attach no blame to that. (There is a
LENSDIR command mentioned in the manual, but it may not be applicable,
as this command pulls files from the main program directory, not the
LENSES directory.)
There are a lot of far more limited lens design programs out there,
which have "free-of-charge" versions with much more extreme
limitations.
Thus, you're justified in saying that there's (almost) no reason for
anyone to bother with anything else.
There definitely are people who have very modest design ambitions, and
who won't have the time for anything without a slick Windows
interface. Even with such an interface, of course they may find trying
to use an even halfway-serious optical design program is no picnic.
And there are people who really are looking for a *free* optical
design program; for example, because they lost their copy of Buchdahl,
and want to find the fifth-order aberration equations from the source.
Thus, if a copy of ORDEALS was sitting somewhere on the web to
download, they would be interested... perhaps as something to give
them a leg up on writing their own.
It does seem to me that Zemax SE is a tad overpriced, but on the other
hand, Zemax EE might be a bargain. They got where they were by being
first on the market, and so they can afford a large programming staff
to provide a fancy user interface, and they have a user community that
benefits them. I can understand that $2,000 might not seem like much
to a company that has expensive lens grinding equipment, for example;
I presume most amateur telescope makers, for example, DID research the
market, and wound up using something else.
Whether that something else was Enterprise SE for free, or one of
several programs available for prices considerably lower than that of
Zemax (such as VOB or Optics Lab) would be a matter of choice.
You invested your time in producing a powerful optical design program
similar to one you already had experience with, and I thank you for
sharing it with others. Serious professional users will need support,
and so naturally you will charge them for the use of your time. I wish
you success in your endeavours.
I presume that one of your goals is for Enterprise to be more widely
recognized for its merits and usefulness; and I think this a laudable
goal, but it has seemed to me, and to many others, that at times you
have let your frustration at the slow pace at which this goal is being
achieved, while other unworthy competitors are in the limelight, get
the better of you, and that with results that do not help that goal to
be achieved.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
Richard F.L.R. Snashall
John Savard wrote:
>On Thu, 25 Mar 2004 16:54:35 GMT, West Coast Engineering
><westcoaste...@westcoastengineering.com> wrote, in part:
>
>
>
>>Seriously, when someone sells source code and places no restrictions
>>on what may be done with it, you would think that the issue of piracy
>>was a non-issue to the person selling it. Then again, maybe when a=b
>>and a=c, it is not obvious that b=c.
>>
>>
>
>Actually, very little is obvious.
>
>
>
>It does seem to me that Zemax SE is a tad overpriced, but on the other
>hand, Zemax EE might be a bargain. They got where they were by being
>first on the market, and so they can afford a large programming staff
>to provide a fancy user interface, and they have a user community that
>benefits them. I can understand that $2,000 might not seem like much
>to a company that has expensive lens grinding equipment, for example;
>I presume most amateur telescope makers, for example, DID research the
>market, and wound up using something else.
>
But not all; at the time of my original (SE) purchase, the
ZEMAX user's manual was online; I took advantage of it.
>
>
>John Savard
>http://home.ecn.ab.ca/~jsavard/index.html
>
>
--
Rick S.
West Coast Engineering
No one ever asked for it before now. I'm pretty good buy I don't read
minds. I'll put it on my to do list.
>
>There are a lot of far more limited lens design programs out there,
>which have "free-of-charge" versions with much more extreme
>limitations.
>
>Thus, you're justified in saying that there's (almost) no reason for
>anyone to bother with anything else.
>
>There definitely are people who have very modest design ambitions, and
>who won't have the time for anything without a slick Windows
>interface. Even with such an interface, of course they may find trying
>to use an even halfway-serious optical design program is no picnic.
>
>And there are people who really are looking for a *free* optical
>design program; for example, because they lost their copy of Buchdahl,
>and want to find the fifth-order aberration equations from the source.
>Thus, if a copy of ORDEALS was sitting somewhere on the web to
>download, they would be interested... perhaps as something to give
>them a leg up on writing their own.
I was on punch cards only at U of R.
>
>It does seem to me that Zemax SE is a tad overpriced, but on the other
>hand, Zemax EE might be a bargain. They got where they were by being
>first on the market, and so they can afford a large programming staff
>to provide a fancy user interface, and they have a user community that
>benefits them. I can understand that $2,000 might not seem like much
>to a company that has expensive lens grinding equipment, for example;
>I presume most amateur telescope makers, for example, DID research the
>market, and wound up using something else.
>
>Whether that something else was Enterprise SE for free, or one of
>several programs available for prices considerably lower than that of
>Zemax (such as VOB or Optics Lab) would be a matter of choice.
>
>You invested your time in producing a powerful optical design program
>similar to one you already had experience with, and I thank you for
>sharing it with others. Serious professional users will need support,
>and so naturally you will charge them for the use of your time. I wish
>you success in your endeavours.
>
>I presume that one of your goals is for Enterprise to be more widely
>recognized for its merits and usefulness; and I think this a laudable
>goal, but it has seemed to me, and to many others, that at times you
>have let your frustration at the slow pace at which this goal is being
>achieved, while other unworthy competitors are in the limelight, get
>the better of you, and that with results that do not help that goal to
>be achieved.
Gee, I'm not frustrated. You need to make progress to be frustrated.
:-)
Thanks for your comments.
West Coast Engineering
You read the manual. No one reads the Manual. :-)
Skywise
<Snipola>
> I have to admit, though, that given the terms he says he licenses it
> under, this could increase his exposure to software piracy.
<Snipola>
Software is pirated because it costs too much. I'm not talking
about specialized niche markets where maybe a few thousand copies
a year are sold. I'm talking about big multi-tens of millions
of sales per year type stuff.
If a particular program sells 10 million copies a year and the
maker charges $10 a copy, the maker gets a cool 100 million
dollars a year. Now, even with all the wasteful overhead in
99% of business, do you honestly think it costs even a million
dollars to actually write the program??? Serious profit without
serious greed.
Now, I said $10 per copy from the maker. I realize that there's
production, distribution, and retail costs that will add to that
$10, but I see no reason why the final retail price of that
program should even approach $50, much less go over that. I have
worked in retail and I've seen how the mark ups work there.
I have also working in CD & DVD manufacturing and I know how
little it costs to mass produce a CD. (less than 50 cents per
disc)
But, once again, 95%+ of people are so gullibly stupid that
they willingly pay $100+ dollars for a mass produced piece
of crap. They'll complain about it and even agree something
should be done. But yet they still pay...
(This applies to just about any product, not just software)
John Savard
<westcoaste...@westcoastengineering.com> wrote, in part:
>Then again, maybe when a=b
>and a=c, it is not obvious that b=c.
My news server, unfortunately, misses some posts. So I have resorted
to Google to fill in the gaps. I was going to assure you that English,
not French, is my first language, but seeing your reply to one of my
earlier posts, which I also missed, I am confident that that post had
nothing to do with me.
I agree that FORTRAN is just about as simple as BASIC, and it has some
extra features that benefit people doing mathematical work.
And I also agree that lewd graphics popping up isn't the way to
promote serious software!
I certainly accept that you don't spend sleepless nights worrying
about software piracy! Instead, you chiefly value, and charge for,
your time in supporting the program.
But on the basis of a=b and a=c implying b=c, I concluded that you
still assign _some_ value to your program, per se.
Maybe it's because the rest of the world values programs instead of
support, and thus it's only for psychological reasons... but you did
include a slight differentiation between Enterprise SE and Enterprise
Pro, and the source code is only available with the more expensive
support contract.
Thus, if your licence agreement gives the people who have purchased
that latter contract complete freedom in using that source code -
perhaps they figure that is simply a measure of the trust you have
extended to them, and they have no intention of responding by
_abusing_ that trust.
(And, yes, the thought that the source might get anonymously posted to
alt.sources by many of these customers in response to certain news
reports, and thus that this was a sort of "backup system", did cross
my mind.)
I do think that you're one of the "good guys".
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
Detector195
> Thus, if your licence agreement gives the people who have purchased
> that latter contract complete freedom in using that source code -
> perhaps they figure that is simply a measure of the trust you have
> extended to them, and they have no intention of responding by
> _abusing_ that trust.
I have had some experience selling an electronics / software product
(nothing nearly as elaborate as Enterprise), with complete design
information and source code available on the Web. So far, nobody has
stolen my design. A few people have rolled their own, and I have
always given them helpful advice and encouragement.
Free source code is almost self-protecting. A person who would copy
the code and use it in a commercial product has to realize that
anybody else could do the same thing. You would really be competing
with free software. Also, it is very hard to make sense of a big code
in its entirety, even if it is very well written. Supporting someone
else's big code is a much bigger task than modifying it for a limited
purpose (with the author's help).
John Savard
(John Savard) wrote, in part:
>It's Charles Gorrie Wynne who was the grand-daddy of all software
>using the damped least squares method of optimization.
I see that the first person to apply damped least-squares to optics
was actually Baker - possibly the same Baker who invented the
Super-Schmidt.
>Gordon H. Spencer is the grand-daddy of the *other* important
>method... Lagrange multipliers, which is an _alternative_ to damped
>least squares. Or at least that's what I've understood from what I've
>been seeing.
And, seeing his paper, I see that this method also includes damped
least-squares, and Lagrange multipliers were important to allow a
program to smoothly take boundary conditions into account.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
wrote, in part:
>the Watcom FORTRAN compiler that is available now, as
>Open Watcom, for free at http://www.openwatcom.org/?
That site doesn't seem to function any longer. While I wouldn't mind
paying the nominal price for the CD, one needs a credit card to do
that.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
<westcoaste...@westcoastengineering.com> wrote, in part:
>Seriously, when someone sells source code and places no restrictions
>on what may be done with it, you would think that the issue of piracy
>was a non-issue to the person selling it. Then again, maybe when a=b
>and a=c, it is not obvious that b=c.
Further: I was not aware of an old post of yours that I turned up when
researching the history of ACCOS V today...
I did not know you had in fact done the "it would be nice, but it is
too much to hope or ask for" thing of offering your source code to the
opus effort. And no one took you up on the offer, which is sad.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
West Coast Engineering
Multiple times in fact. I was amazed, and then OPUS just died.
Today, I looked into the mirror and finally saw the, up to this point,
invisible warning lable written there. It said,
"I'm a total idiot, don't listen to a thing I say". :-)
West Coast Engineering
You need to be careful with lagrange multipliers. Even using CODE-V,
if you write constraints that are stupid or violate causality, you can
always and easily end up with a non-recoverable condition.
No code is as dumb as a human being in a hurry. :-)
WCE
John Savard
(John Savard) wrote, in part:
>Oh, no. First Zemax took SIGMA out behind the barn... and now, OSLO
>has done the same to ACCOS V? Although, since Optikos is still in
>business, I don't see how that could have worked that way.
Steve Eckhardt's reply to this didn't show up on my server; although
I'm sure it was done in the same tone of levity as this posting of
mine, I should note that the Cool Genii code _was_ integrated with
OSLO; as Mr. Eckhardt himself noted in a recent post, and a post by
Doug Sinclair himself named Dale Buralli as the programmer chiefly
responsible for that integration, as well as working on many other
parts of the program.
And Gordon H. Spencer, of ACCOS V fame, was part of the team as well.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
in part:
>If you want you can rightly call Spencer (Gordon, not James) the
>grand-daddy of all software that uses the constrained damped least
>squares method of optimization.
I see now that I missed *constrained*, which indeed was the
modification he introduced - by using Lagrange multipliers.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
in part:
>If it helps, ALEC was the name of the software before it was changed
>to ACCOS and it was dated about 1963, the time when Spencer published
>"A Flexible Automatic Lens Design Procedure" in Applied Optics, Vol.
>2, No. 12.
Would you happen to be the inventor of U. S. Patent 3,547,525, which I
ran across looking for patents on telescope coma correctors (although
that patent is for a solid mirror lens)?
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
John Savard
(John Savard) wrote, in part:
>That site doesn't seem to function any longer.
It was probably just a problem with my browsers. I was able to
overcome that problem with the help of Google.
However, it should be noted that the program comes only with a limited
SDK, because the commercial product from which it is descended
licensed one not available for free redistribution - and the current
license agreement with the package limits what you can do with your
compiled applications, although it is planned to change that in
future.
So I don't think that this compiler lends itself to consideration for
the next generation of Enterprise for _those_ reasons, unfortunately.
John Savard
http://home.ecn.ab.ca/~jsavard/index.html
chris.b...@gmail.com
Any opinions or resources would be much appreciated, and if anyone can provide some in-depth consulting in a phone call, I'd be happy to provide some compensation because this would be really helpful for my research.
Cheers,
Chris