Free Pascal Compiler 3.0.0 Is Out; Adds Support For 16-Bit MS-DOS, 64-Bit iOS (freepascal.org)

Halo1 writes: Twenty-three years ago, development started on the first version of the Turbo Pascal and later also Delphi-compatible Free Pascal Compiler, for OS/2 no less. Two decades and change later, the new Free Pascal Compiler 3.0.0 release still supports OS/2, along with a host of older and newer platforms ranging from MS-DOS on an 8086 to the latest Linux and iOS running on AArch64. On the language front, the new features include support for type helpers, codepage-aware strings and a utility to automatically generate JNI bridges for Pascal code. In the mean time, development on the next versions continues, with support for generic functions, an optional LLVM code generator backend and full support for ISO and Extended Pascal progressing well.

Comments

Short FPC history and goals overview

[Halo1]

I thought I added the link to my original summary, but it seems to have gotten lost when I submitted it. In any case, Sourceforge's Project of the Month April 2014 interview with the founder of the Free Pascal Compiler, Florian Klaempfl, contains a good overview of the project's history, goals and development methodology.

Re:Short FPC history and goals overview

[Anonymous Coward]

23 years ago? That makes it 1992? I was using Turbo Pascal in 1989!

Re:Short FPC history and goals overview

[Halo1]

23 years ago? That makes it 1992? I was using Turbo Pascal in 1989!

Yes, Florian started the project when it became clear Borland was not going to create a 32 bit version of TP/BP.

Re:Short FPC history and goals overview

[Anonymous Coward]

Yeah but too bad with all the time it took, Pascal has become somewhat obsolete, supplanted within its core support community by Modula.

Re:Short FPC history and goals overview

[vux984]

23 years ago? That makes it 1992? I was using Turbo Pascal in 1989!

Re-read the summary.

It says 23 years ago development on the Turbo Pascal compatible project "Free Pascal" was started.

It does NOT say 23 years ago development on Turbo Pascal itself was started.

Re:Short FPC history and goals overview

[Anonymous Coward]

Just because pascal isn't trendy anymore doesn't make it less awesome. Free Pascal / Lazarus is an excellent environment for getting stuff done.

Re:Short FPC history and goals overview

[TapeCutter]

The summary says: "Twenty-three years ago, development started on the first version of the Turbo Pascal and later also Delphi-compatible Free Pascal Compiler" Perhaps you should re-read the summary, your paraphrase edits out the bit that makes all the difference???

Re:Short FPC history and goals overview

[Lunix Nutcase]

Which is still saying the same thing. The part GP omitted is simply saying that the Free Pascal Compiler was Turbo Pascal-compatible then also later Delphi-compatible.

Re:Short FPC history and goals overview

[gstoddart]

Twenty-three years ago, development started on the first version of the Turbo Pascal and later also Delphi-compatible Free Pascal Compiler

No, Turbo Pascal is not 23 years old ... the grammar suggests that, but reality doesn't.

I know this, becaise 23 years ago I had a second hand 286 PC with Turbo Pascal on it. And it wasn't exactly new even then.

Turbo Pascal has been around since 1984 .. that would be 31 years ago.

So, you can argue the sentence should have read as "Twenty-three years ago, development started on the first version of the Turbo Pascal (and later also Delphi)-compatible Free Pascal Compiler".

But what you can't do is argue that Turbo Pascal is 23 years old. Because that's utterly incorrect.

Re:Short FPC history and goals overview

[BetterThanCaesar]

Twenty-three years ago, development started on the first version of the Turbo Pascal and later also Delphi-compatible Free Pascal Compiler

Parsed that for you.

Re:Short FPC history and goals overview

[Anonymous Coward]

> The summary says: "Twenty-three years ago, development started on the first version of the (Turbo Pascal and later also Delphi)-compatible Free Pascal Compiler" Perhaps you should re-read the summary, your paraphrase edits out the bit that makes all the difference???

Added parentheses 'cause of defect in user's parser operator precedence.

Re:Short FPC history and goals overview

[silentcoder]

I was a long time contributor and user of FPC and Lazarus but I moved on to other work a few years ago and haven't been near the Pascal scene since. What's the status of Lazarus ? Still under active development ?

Re:Short FPC history and goals overview

[Halo1]

What's the status of Lazarus ? Still under active development ?

Absolutely: http://www.lazarus-ide.org/

Re:Short FPC history and goals overview

[jandersen]

Good, old Pascal. What a horrible language, but I liked it :-)

Re:Short FPC history and goals overview

[OrangeTide]

conjunction junction, what's your function?

Re:Short FPC history and goals overview

[Bitbeisser]

Yeah but too bad with all the time it took, Pascal has become somewhat obsolete, supplanted within its core support community by Modula.

Sorry, but Modula has gone nowhere. There isn't just one Modula compiler project out there (which are few and far between anway) that comes even close to FreePascal. And with the Lazarus IDE, you get a very easy to use RAD development environment for all the major GUI based OS, allowing for easy cross-platform development. And just because of all the buzz of the "new kids on the block", Pascal hasn't become "obsolete" by a long shot. Pretty much all the people that I have met that made that claim have never written a single program in Pascal, they just parrot what they heard elsewhere...

Re:Short FPC history and goals overview

[illtud]

Isn't that what he said? (using parentheses rather then bold)?

Re:Short FPC history and goals overview

[illtud]

no it isn't, ignore me.

I remember

[Anonymous Coward]

I remember having to save my allowance to buy Borland Turbo Pascal for around $60. It came in a yellow box. A couple of other things I remember:

1) I am getting old
2) Those days sucked

Thanks to GNU and gcc we will never have to put up with that crap again.

Re:I remember

[Anonymous Coward]

I remember when I first moved from Turbo C to djcpp; from "tiny", "small", "irritating", "large", "flatulent" and "huge" memory model to ... flat. It was as wonderful as not being repeatedly kicked in the balls.

Re:I remember

[Anonymous Coward]

You missed out on the horror days when borland finally completely lost the plot and the cheapest you could get anything from them was the beginner version of delphi for well north of a $1K. And they wondered why dot net left them in the dust. Which is a shame. Delphi was actually a pretty great system, in its day.

Re:I remember

[Anonymous Coward]

All those goofy memory models were of course courtesy of Intel, who foolishly stuck with the segmented memory model and 16-bit offsets for years. The 286 was a stupid design in that it expected you to break up any data structure larger than 64K bytes, and you had to fucking reset the CPU (after setting a flag in CMOS for BIOS to know what to do) to get out of its protected mode, back into the mode that all existing software ran in. IMHO it set the industry back by almost a decade having to futz with that shit.

Meanwhile, I was happily hacking the Macintosh with its linear addressing space. Not that I didn't see 8086 addressing happen; MS Word 1.0 ran a bytecode VM that used segmented addressing and couldn't run higher than 1 meg in memory. Switcher/Multifinder had to give it special (as in education) handling by allocating other apps from the top of memory down.

Re:I remember

[Anonymous Coward]

Most people buying Turbo Pascal weren't "IT people". They were hobbyists (especially back in the 80s). I bought Turbo Pascal (as well as MS C and MASM) long before I started making money as a software developer. Without "cheap" tools, I never would have been able to start my career.

Re:I remember

[Gramie2]

They could afford if it they already knew and had confidence in it. How do people get started when the basic configuration is $1000+, and that's without client/server DB access ($300+ more)? And now you need to pay another $300+/year for a subscription to get bug fixes.

There was a time when Borland had a no-nonsense license (i.e. install wherever you want, as long as only one copy is used at a time), and they didn't have DRM that made the boot-up time almost double. They used to put customers first. Now they try to slip in restrictions in new EULAs, and back off if they get caught.

Re:I remember

[Bitbeisser]

I remember having to save my allowance to buy Borland Turbo Pascal for around $60. It came in a yellow box. A couple of other things I remember:

1) I am getting old 2) Those days sucked

Thanks to GNU and gcc we will never have to put up with that crap again.

My sincere condolences that you still have to put up with gcc, I know, it scars you for the rest of your life... >:)

Free Pascal is awesome.

[shihonage]

I wrote a DOS game in 1996 in Turbo Pascal which used $B800:0000 textmode space to display the action. Thanks to Free Pascal I successfully ported it to Windows... of course minus the literal memory addressing and such.
Free Pascal is amazing at how it "just works" with legacy Turbo Pascal syntax where Delphi would present more trouble. Lazarus, the Free Pascal IDE, is also very resemblant of Turbo Pascal IDE, with some modern touches.
Pascal is an underrated language. It may have been designed for education, but it has many advanced features, the executables are nearly as fast as C++ ones, it compiles fast, and the runtime diagnostics are detailed and specific. It "just works".

Re:Free Pascal is awesome.

[MightyMartian]

I loved the Pascal family of languages. I did a lot of work in TurboPascal and Basic-09. The latter was basically Pascal with a bit more BASIC-like syntax. I still prefer Pascal style variable declarations.

Re:Free Pascal is awesome.

[jonwil]

I got my start in programming with Turbo Pascal and still have some old Pascal code from back in the day hanging around.

It might compile with Free Pascal and run on modern Windows PCs although that would depend on whether Free Pascal supports the Borland BGI graphics library on modern Windows machines somehow and whether it supports the mouse code I used back in the day.

Re:Free Pascal is awesome.

[Halo1]

FPC comes with a (mostly) TP-compatible graph unit for Windows (32 and 64 bit). It doesn't use BGI drivers, because it doesn't need to. Your mouse code won't work though, since you can't access the mouse driver under Windows using DOS interrupts. We do have our own cross-platform mouse unit you may want to use instead.

Re:Free Pascal is awesome.

[DrXym]

C became predominant because code ran faster, was generally more portable (more commonality, less proprietary extensions between competing compilers), more terse / expressive, was better able to talk with the operating system (strings, structs etc.), was the defacto language on some operating systems and it was the first system programming language many people encountered.

Pascal eventually overcame many of its shortcomings but by then it was too late. C (and the emerging C++) had already gained dominance and mindshare.

Re:Free Pascal is awesome.

[silentcoder]

Worth mentioning that FPC will also let him port his program to Linux with minimal effort as the units are almost entirely identical.

Re:Free Pascal is awesome.

[silentcoder]

Correction: Delphi is an IDE - the language Delphi compiled was called Object Pascal. That hackish horrible thing you refer to never had a name of it's own it was just "object oriented programming support added in Turbo Pascal 5"

Re:Free Pascal is awesome.

[rochrist]

I also loved Pascal. Getting the UCSD Pascal system for my first Apple II was a life changing event!

Re:Free Pascal is awesome.

[rochrist]

Lighten up, Francis.

Re:Free Pascal is awesome.

[DrXym]

You're right. It wasn't. Nutcase.

Re:Free Pascal is awesome.

[StormReaver]

I did a lot of work in TurboPascal and Basic-09.

In the mid 1980's, I was writing a lot of BASIC-09 on OS/9 Level 2. At the time, I was the librarian for the local CoCo club, and had access to the club's Pascal compiler and C compiler. I decided to try Pascal, since a high school classmate raved about it, and installed the Pascal compiler. Or rather, I tried to install the Pascal compiler without success. It installed, but it wouldn't run.

So I then tried installing the C compiler, and succeeded. I then decided that I would learn C rather then Pascal. The rest of my career was then driven by C-like languages. Looking back, I chuckle at how such a profound life decision was decided by a random act of convenience.

Great work!

[SirJorgelOfBorgel]

I've used Turbo Pascal and Delphi for years, later followed by FPC and Lazarus. It was an enjoyable time. While I was no stranger to C++ even in those times, I simply preferred Object Pascal (so sue me). We even used it on Windows Mobile up to 6.x.

Unfortunately, when the time came (years ago now) to focus on Android, FPC's RTL had fairly serious issues, that I personally did not have to expertise to fix, and it wasn't a priority for anybody else who did.

While I have no plans to return to FPC, I'm still a fan, and I love to see it progress.

Pascal Changed my life

[wassomeyob]

I could show this article to my wife and truthfully say "we would not have met and had offspring if it weren't for this computer programming language you've never heard of". My first real job that eventually led to a career and a move to this city was because I had done work in school on a parser for Pascal, ended up programming on an HP1000 mini (RTE-XL), meeting her, getting married etc etc. Then I think, 'heh, never mind'. Very nostalic; glad to see it's out there. Thanks, Pascal.

Re:Pascal Changed my life

[Anonymous Coward]

MIne, too.

I remember despising C for its absurd syntax ("==", "!=" etc.).

I still do.

If I have a chance, I plan on doing some programming again with FreePascal/Lazarus.

Re:WHY?

[Halo1]

Why the fuck would anyone bother supporting 16-bit MSDOS? Really?

Simply because someone cares. Someone else is also reviving Amiga 68k support, because he cares. Our compiler is generally modular and generic enough so that such support does not result in too much interference with other functionality, or luggage that makes things unmaintainable over time.

Where's the CP/M support for Z80 then?

It'll get added as soon as you provide good quality patches for it :)

Re:WHY?

[Nyder]

Why the fuck would anyone bother supporting 16-bit MSDOS? Really?

Where's the CP/M support for Z80 then?

Does anyone use Pascal any more?

http://www.z80.eu/pas-compiler...

There are some pascal compilers for cp/m. Enjoy.

Re:WHY?

[HotNeedleOfInquiry]

Because there's a buttload of legacy embedded apps running it? I know of at least 20 I've had my hands near, including one running on the International Space Station.

Tandy 1500?

[AndyKron]

Will it work on a Tandy 1500 laptop?

Re:WHY?

[MightyMartian]

Believe it or not, there is some embedded hardware out there that still uses DOS variants. The messaging module of our phone system uses an embedded version of DOS.

Put it in the library

[tepples]

NO language needs a garbage collector, though an option to use one selectively would be nice.

Let's split the difference and say a language needs a garbage collector in its standard library that a programmer can choose to enable. C++ calls its reference-counting garbage collector std::shared_ptr.

i was about to try it....

[Anonymous Coward]

But then I saw the systemd requirement, and decided to pass.

OpenWatcom

[Wootery]

Speaking of ye olde compilers, OpenWatcom seems to have ground to a halt in 2010. Can't tell if I think that's a shame, or if its time has come, or both.

Re:OpenWatcom

[Reziac]

There may not be much left to change or update. But the project itself is still alive, far as I can tell.

Re:OpenWatcom

[Wootery]

There may not be much left to change or update.

How's that? It would be great if it could be brought up to speed with compatibility with modern C++, and modern optimisations. No small task, of course.

C vs Pascal == Perl vs Python

[DrYak]

I remember despising C for its absurd syntax ("==", "!=" etc.).
I still do.

And I was the opposite, I despised the vebosity of pascal (begin/end/etc.) and it's tendency to try to hide some low level details on the grounds of making it easier to learn.
To each his own preferences.

That's a definitive proof that the Perl vs Python debate didn't actually need theese language and the whole concept dates back much further in computing history.

Re:C vs Pascal == Perl vs Python

[DoctorBonzo]

at some point I completely gave up on Pascal because there wasn't any 'return' statement available in functions thus forcing ridiculous amounts of indentation.

or maybe I was just ignorant?

is there one now?

Re:C vs Pascal == Perl vs Python

[dwywit]

You're not ignorant. Pascal sucks.

Re:C vs Pascal == Perl vs Python

[Toshito]

I see that you forgot to take your meds today!

What happened to Pascal, anyway?

[swb]

I remember in the 1980s it seemed like kind of a big deal, an "advanced" programming language that required a compiler and a more real computer than an Apple ][ (although, yes, there was a Pascal system for the ][, IIRC it was worthless without two disk drives and really not an ideal platform). I knew people writing commercial software in Pascal. They taught it when I was in college. I think "Inside Macintosh" Vols. 1-3 that documented the Macintosh used Pascal.

It was kind of everywhere, and then it wasn't. What happened to it? Was it not really meant to be a "practical" language and meant to be kind of an advanced educational language? Did the growth of Unix-like systems on x86 push everyone into C? Did stuff like the availability of maybe Visual Basic or something grab the users who would have used Pascal?

Circa 1986 or so, you wouldn't have thought "kind of a dead language, nobody uses it for anything anymore" and you wouldn't have thought it would get that way any time soon.

Re:What happened to Pascal, anyway?

[metamatic]

The Pascal community fragmented. The 8-bit systems carried on using ISO Pascal or UCSD Pascal, but Wirth and other key Pascal experts went off and created Modula-2, which was much more practical for real world programming. (I used Modula-2 on the Atari ST, it was a much nicer experience than trying to program GEM in C.)

But instead of Pascal or Modula-2, Borland went off and did their own thing, producing a proprietary "Pascal" that wasn't compatible with anyone else.

Then the Modula-2 community split into the Oberon (Wirth) and Modula-3 (everyone else) communities to add OO, and Borland again did their own thing and ignored everyone else.

Now we have Go, which takes C and adds in ideas from Modula and Oberon. And Free Pascal still isn't even compatible with 1982's standard ISO Pascal.

Re:What happened to Pascal, anyway?

[Anonymous Coward]

And Free Pascal still isn't even compatible with 1982's standard ISO Pascal.

Why is that a problem in particular? Things have moved on a lot in the last 33 years. Even the first C standard isn't that old.

Re:What happened to Pascal, anyway?

[metamatic]

As I said, fragmentation is what killed the Pascal community. Or at least, that was my view as a participant. The fact that we still don't have a common Pascal standard today means it's not going to come back from the dead.

Re:What happened to Pascal, anyway?

[aberglas]

But it don't support pointer arithmetic. Every slash dot reader knows that you need pointer arithmentic (char *x; x++) in order to be efficient.

Seriously, the early Pascals were rather fascists about type safety. So the whole world jumped to that miserable excuse of a language called C which let them do what they thought they wanted to do.

(I grew up on the original Wirth Pascal on a Cyber with 60 bit words. That is what all that packed character stuff was about.)

Re:What happened to Pascal, anyway?

[MarioRayMahardhika]

And Free Pascal still isn't even compatible with 1982's standard ISO Pascal.

Correction: FPC has ISO compatibility mode that gets mature, at this time of writing, as Jonas said in the mailing list, only 1 little bug that didn't get into 3.0.0 and already fixed in trunk. I personally have compiled P5/P6 compiler provided by standardpascal.org and it compiles flawlessly with FPC. I'd like to run it against Pascal Test Suite that should declare whether a compiler is standard compliant or not, but I just told standardpascal.org people and hope that they will do it. They should have the complete package which I can't find anywhere.

Re:What happened to Pascal, anyway?

[MarioRayMahardhika]

But it don't support pointer arithmetic

Seriously, you need to RTFM before stating such a thing: http://www.freepascal.org/docs...

Re:What happened to Pascal, anyway?

[metamatic]

The two major Pascal implementations (Free Pascal and Delphi) are fairly compatible with each other so it's not as fragmented as you think.

It's isn't fragmented now, because it's dead other than those two non-standard compilers, all the other implementations having vanished along with their communities...

Re:What happened to Pascal, anyway?

[Halo1]

For better or worse, what Delphi does has become a de facto or so-called industrial standard. Just like Apple with Objective-C and Swift. Official standards are indeed not a guarantee for success.

Re:What happened to Pascal, anyway?

[metaforest]

Wizardry: Proving Grounds of the Mad Overlord by Sir-Tech, was written in Apple UCSD PASCAL.

Re: What happened to Pascal, anyway?

[techcodie]

Ya think? Just wondering - When did programmers stop ignoring that, along with range checking, and when did buffer exploits start happening?

I remember it being about the same time.

Coincedence?

Re:What happened to Pascal, anyway?

[MarioRayMahardhika]

It's isn't fragmented now, because it's dead other than those two non-standard compilers, all the other implementations having vanished along with their communities...

Let's see... besides those two, there are:

1. Smart Mobile Studio
2. AVRco Pascal
3. MikroPascal
4. Vector Pascal
5. paxCompiler

The first one is I know for sure very much alive as the author is very active in the community and they really have customers. The next four might be less known, but they can be considered active looking at their latest release date.

Re:Pffft

[TapeCutter]

Most business' neither need nor use a real time O/S. All applications that allocate random blocks of memory must find a way to deal with garbage collection, built-in GC makes it easy for the coder. Manual GC just means the coder must manage memory himself, which is mandatory on a proper real time O/S.

Increment an pointer into an array

[aberglas]

Does the new Pascal let you increment a pointer through an array yet (char *x; x++)?

Every slash dot reader *knows* that that is an essential feature of any programming language in order for it to be efficient.

This simplifies porting from Delphi

[Barryke]

Let me just paste a code comment i wrote two months ago;

  * Written for use in Delphi 2010.
  * Seems to work (with minor tweaks) in Lazarus v1.4.4 with FPC 2.6.4, Windows.
  * Could work with fewer tweaks with FPC3.

Now i'm wondering if i can get this pet project to compile and run on Linux ..just for sports.

Re:WHY?

[OrangeTide]

Yes, there are still devices that run on pc/104 and other embedded x86 platforms. The older ones still running ROMable versions of DOS.

Z80 support would be nice for me, I'm still using SDCC so I'm still stuck in C land. Of course I'd also like 68hc11 support too, but I'm not going to hold my breath.

Sure, some people still use Pascal. Some people still use Java even though it's obvious that Oracle is eventually going to screw it up.

Re:WHY?

[Anonymous Coward]

Some of the Canon pro-sumer cameras are running DOS as well.

Re:Pascal in the real world?

[rochrist]

First two programming jobs I had we're Pascal. I'm sure it's used far less now, but /something/ is keeping Delphi alive and supporting those astronomical price tags.

Re:WHY?

[Bitbeisser]

Why the fuck would anyone bother supporting 16-bit MSDOS? Really?

Simply because someone cares. Someone else is also reviving Amiga 68k support, because he cares. Our compiler is generally modular and generic enough so that such support does not result in too much interference with other functionality, or luggage that makes things unmaintainable over time.

+1 (at least for the 16-bit MSDOS part ;-) )

Where's the CP/M support for Z80 then?

It'll get added as soon as you provide good quality patches for it :)

(Un)fortunately, I think we will have to wait a long time for that to happen... ;-) In general, it is sad to see how arrogant and ignorant the (programming) world has become...

Re:Pascal in the real world?

[MarioRayMahardhika]

but with the costs of Delphi and that its been around so long and still going means someone is using it.. but who ???

From Embarcadero: - http://www.embarcadero.com/pro... From Lazarus/FPC wiki: - http://wiki.lazarus.freepascal... - http://wiki.lazarus.freepascal... - http://wiki.lazarus.freepascal...