I'm developing a statistical software and I need a mathematical (algebra) unit for FreePascal. Right now I'm using Jedi_Math with Delphi but I don't want the VLC/CLX dependency and I hate using proprietary software (Delphi or Kylix). If the unit is object-oriented, even better.

Thanks to the last release of the 2.0 beta I was able to compile all of the non-visual classes of the program!

Jedi math hardly has a math dependancy, at least the last time I looked, it worked with FPC 1.9.x in a few minutes. (on FreeBSD)

I even got the _visual_ part to work with Lazarus, but that was more an empty form.

Thanks to the last release of the 2.0 beta I was able to compile all of the non-visual classes of the program!

Other possibilities:
- there is numlib in the packages somewhere. crude, but if you need a few routines, snatching it from there would work. It's more linear algegra tho.
- I once wrote a unit symbolic that can derive an expression (it can also evaluate expressions. Jedi Math's parser borrows from it)~
- TP style tehre always is:
http://www.unilim.fr/pages_perso/jean.debord/tpmath/tpmath.htm They also have a FPC/Delphi/GPC version
- Looking here might also bring you something
http://www.efg2.com/Lab/Library/Delphi/MathInfo/Resources.htm

There is the delphi-compatible math unit in the RTL, but I suppose this is not enough. I have the 'Numerical Recipes in Pascal' CD-ROM, but I suspect there might be copyright issues involved when distributing these algorithms :/

Do you want symbolical or numerical algorithms ?

Michael.

Only numerical. In fact, given that I need to be 100% confident on the results I'll have to write all the numerical functions myself. The only thing I need is a flexible and easy to use matrix/vector library.

Although I had a lot of success working with FreePascal, I can't say the same of Lazarus. In fact, the ONLY reason for what I'm using Delphi (and Windows) is the current unstable state of Lazarus and the abandonment of Kylix by Borland.

Other possibilities:

I'll look into them. Thanks.

Here is another link you might find useful:
http://www-rab.larc.nasa.gov/nmp/fNMPhome.htm

__________________________________
Do you Yahoo!?
Yahoo! Mail - You care about security. So do we.
http://promotions.yahoo.com/new_mail

A version of the NR code is freely available on the net as a zip file complete with the code and the examples. The code matches the book copyrighted 1986/1989, I don't know if there is a later edition.. I think the readme in the zip file calls this an older version released as shareware....

The code is not particulary free but it seems a bit hazy. The colection is not free. But, for example, I'm working on a problem using sigular value decomposition right now and the SVDCMP.PAS routine in NR is apparently a translation of free-er EISPACK routines.

I would be cautious of GPL/LGPL units that that use NR code. It's not clear that that's legal...

-Alan

Yes.
I found it somewhere in finland. The book is from 89, the readme from 91.

Ah, the (in)famous (EIS/LA)PACK. How many times I had to install it :)

I would be cautious of GPL/LGPL units that that use NR code. It's not clear that that's legal...

As far as I can see from the README It would not be illegal for us to distribute it, if we keep the whole package together, and we don't base any of our code on it (not likely for a compiler ;)).

It seems to me that it is then up to the user to decide what to do with it once he decides to use it. The FPC units are under LGPL so they don't require that you make your projects open source.

In each case: in my search for numerical routines, I found many links to units and suites.

Michael.

----- Original Message -----

1) Using proprietary software to build scientific applications (like in my case) is just wrong. You never know what there might be happening "under the hood". And you are bound to licenses and EULAS that are designed to avoid any kind of responsability of the software maker.

2) There is a Delphi personal edition. It's being sold for usd 100. I'm not going to pay usd 100 for a software license that tells me what can o can't do with it. I rather donate that money to an free software project or a local charity.

3) Everytime I have to boot on Windows I feel like a little part of me dies. Everytime time Windows hags that little part goes to Hell and says "hi" to Bill Gates.

Hi,

I have a program which use keypressed and readkey (crt unit) in linux. When I press CTRL-C
readkey return #3. It's okey and working but I like to disable this option in keyboard subsystem
because I want to use my only CTRL-C function.

I have implemented my function for signal SIGINT but I think keyboard subsystem is caching this
signal and it doesn't reach to the my function.

best regards

----- Original Message -----

Ok, try to compile a _working_ fpc with delphi :>

----- Original Message -----

Harald Houppermans wrote

I doubt you know what happens "under the hood" of free pascal.

That's not the point.

Also I believe the free pascal source code is fully available so you might look at specific parts ;)

That's the point :) No, seriously, I have some years of experience working with Java and C, C++ and I'm almost a CS graduate, so I guess if I had the time I could see "what's under the hood" and understand it.

I believe there is a free delphi version somewhere ;)

Borland removed it from the web and ftp site. Anyway I bought an old magazine that had Delphi 6 PE, and apparently it's still legal to use it. If you think about it it's pretty weird given that they have similar versions for C++Builder and JBuilder.

Open source usually have licenses as well which tell you what you can and can't do.

Yes, but you don't have to throw your money to a corporation that could leave you hanging. Think about what happened with Kylix. That project is obviously abandoned. Have you tried to make it work on any the newest linux distributions? Good luck. If it was a free software project someone (like myself) could take the source and make it work.

I think Delphi and Windows are well tested and a excellent choice for any type of application =D

I worked several years as a network administrator for a a school of engineering and I got a lot of feedback from the users about Windows. I seriously wouldn't use Windows for any production project. It it fundamentally flawed and a waste of money and resources. Linux could be flawed too, but my money goes to a local LUG that makes good use of it and not to a group of people that are rich beyond anyone's imagination.

I don't know about Delphi, I've been using it for a couple of years and it didn't hang even once. But, then again, maybe I was just lucky. I have nothing against Borland, I own several of their products, but IMHO free software is better than proprietary software.

I DO have a LOT agains MIcrosoft. Microsoft represents almost everything that's wrong with capitalism and does everything it can to get "better" at it every year. I live in a country that suffers from this kind of company's behavior. I refuse to be an accomplice.

Yes, he does.

I think the free pascal compiler was written in pascal ?

It is.

What pascal compiler is used to compile free pascal ? :)

Free Pascal :)

Jonas

The last time we tried to compile FPC with Delphi (and made the sources delphi compatible), Delphi didn't get more out than internal errors, so Delphi isn't always the choice :)

I think the free pascal compiler was written in pascal ?

Yes?

What pascal compiler is used to compile free pascal ? :)

FPC. Early versions were compiled with TP.

IMHO, Kylix was before its time a bit. I was at BORCON last year and there were quite a few people interested in it being ressurected. No, it has not been abandoned, luckily. Borland is just holding onto it for right now. At first I though this was a mistake, but I thought about it. Borland has to swim in the same economic waters as the rest of the world. They have a certain number of resources to stay afloat and make money. That being the case, it only makes sence to swim the MS way for right now. BUT, with Kylik, they have a small boat floating in a different stream, the Linux stream. If and when that one starts flowing, Borland can rightfully say that they were there first.

Of course that assumes FPC won't sink Kylix. Right now, my bet is 50/50.

I seriously wouldn't use Windows for any production project. It it fundamentally flawed and a waste of money and resources.

No offence but that is a bit short-sighted. I made a decent living for the last 10+ years writing Windows code and managing databases on MS software. In the grander scheme of things, the MS platform isn't that bad, as long as you know where the problems are.

I don't know about Delphi, I've been using it for a couple of years and it didn't hang even once. But, then again, maybe I was just lucky.

Perhaps 'luck' is right... Crashing Delphi isn't all that difficult.

Regardless of how you feel about capitalism and MS, I strongly suggest reading two books by Ayn Rand: _The Fountainhead_ and _Anthem_. If you are feeling adventurous, try reading _Atlas Shruged_.

Just my thoughts...
Ken -- KC7RAD

Agustin Barto wrote on 20/7/04 22:31:

... and the abandonment of Kylix by Borland.

Untrue.... Kylix is on hold. It has not been "abandoned"... That notion is complete FUD! I'm offline at the moment (typing this on my trusty palmpilot professional), but one of the Borland engineer's BLOGs mentioned the future of Kylix. Anders or Allen IIRC. There will be no release *this* year, but that does *not* signal the demise of the project!

Matt

That's very good to hear! Sadly, for the developers, there's no difference between an abandoned project and an "on hold" one. I hope they can bring it back to life.

No offence but that is a bit short-sighted. I made a decent living for the last 10+ years writing Windows code and managing databases on MS software. In the grander scheme of things, the MS platform isn't that bad, as long as you know where the problems are.

I worked with (and for) Microsoft software for quite some time before realizing that there was a better world out there. Sincerely, I don't mean to offend you in any way, but I don't miss those days one bit.

Regardless of how you feel about capitalism and MS, I strongly suggest reading two books by Ayn Rand: _The Fountainhead_ and _Anthem_. If you are feeling adventurous, try reading _Atlas Shruged_.

My feelings towards MS don't come from second hand experience. I could fill pages about the (illegal or unethical) things that MS does here. I could also do the same with IBM, but this is not the place. Anyway, I always appreciate a good book, so I'll try to get a hold on those you suggest.

Wasn't this an fpc mailing-list? :)

> What pascal compiler is used to compile free pascal ? :) FPC. Early versions were compiled with TP.

That's the most funniest thing in software...

Using previous software/version to create new versions... then using those again to make newer versions

and each new version has ofcourse new 'power' :) and with the new power... new power can be made :D

[ Charset ISO-8859-1 unsupported, converting... ]

Isn't that normal? Life comes from another life also.

The few cases that that doesn't happen, become, depinding on beliefs, religion,
really thick scientific books about "Origin of life", or alien intervention.

and each new version has ofcourse new 'power' :) and with the new power... new power can be made :D

A better explanation for the compiler case is that one team has control over both chicken and egg. This is in theory more work, but in practice saves work.

I quote from someone else:

"standard IEEE floating values, are highly inaccurate for scientific calculations."

I you sure you want to use floating point math for your library ? :)