Object Pascal decline?

[Paul Breneman]

Exactly...
I've heard that Pascal procedural runs as fast as a program written in C.
I do not know if this is true, but I would not be surprised if it were true.

See the section near the bottom of this page (where you can compare the speed of different languages):
http://www.turbocontrol.com/embeddedfreepascal.htm

[Shpend]

@Paul Breneman
I think this "guy" is either a bot or some shill. All his links are scam websites, @staff consider banning this guy. I'm not kidding.

Avoid visiting his website below.. It is probably big scam shit.

[marcov]

No, the sites were real, but seem to have expired recently.

Wayback has site snapshots for it as recent as last summer:

https://web.archive.org/web/20200710163045/http://www.turbocontrol.com/embeddedfreepascal.htm

[Shpend]

@marcov:
what is with his other site: "controlPanel" you get a super low lvl webpage from 1990 - 2000, filled with only Chinese casino content and other colorful epileptical blinking pages, idk if that has to do with expiration.

[Fred vS]

Paul Breneman is a man with highest skill in computers.

He has nothing to do with what appends to his sites, this is hackers work.

We all wait for news from him, I hope all is ok.

[ccrause]

what is with his other site: "controlPanel"...

I assume you are referring to the link in his signature.  That used to be a real site too: https://web.archive.org/web/20191122001741/http://www.controlpascal.com/

Paul is a known legitimate user on this forum (check his profile), seems like he just stopped maintaining his web sites.

Edit: Or got hacked as suggested by Fred vS.

[440bx]

I doubt the decline of Pascal can be stopped but I'd love to be wrong.

I believe an important reason for that decline is that, as a language, Pascal does a number of things reasonably well but, it really doesn't excel in any specific area and, in most areas, maybe even all areas, it has significant shortcomings that are not addressed. Aside from that, its purpose was/is a way of teaching good programming but, what its purpose/goal is as a development tool remains unclear to this day.

C as a programming language is a disgrace (and that's an understatement) but, when a programmer wants code to be tight and efficient (and doesn't mind a maintenance disaster), it is still (rather unfortunately and deficiently) the "best" choice.  From the moment it was created, its goal/purpose was crystal clear : a high level assembler (with a syntax that makes assembler feel like a 4GL in comparison.) <chuckle>

Currently Pascal is a reasonably good general purpose programming language.  A jack of all trades, master of none (sadly.)

[Peter H]

The website brenemanlabs.com has vanished and most google answers say "Paul Breneman" passed away.
So he might have gone, silently, hopefully in peace.
I can say this freely, I am 67 now and know even my life has an end, sooner or (hopefully) later.

This said, behind other languages there are big companys and big money, gcc was even supported by US military, because these guys need a certified compiler if they write certified software. US military is going from ADA to C++, because integration to other systems is easier and C++ 20 now got a module concept similar to MODULA II. Other extensions -including simplifying and clarifying syntax and error messages, especially for templates- are in the pipeline for C++23.

Pascal is a beautiful language, but it suffers from some dogmas that are historically justified but hinder its evolution, and currently only emba and the freepascal team are behind it.
BTW, C was never designed to evolve. It had no design and the syntax was pragmatically, however, therefore it was able to evolve.

Finally: Pascal was not designed to evolve, however in software industry you need forward compatible evolution, not revolution or stagnation, this is the credo of Bjarne Stroustrup and his C++, and this explains the success.
BTW, C was not designed to evolve. The syntax was pragmatical and has evolved from "no function prototypes and integer can be used as pointer and vice versa" to today.
Another point is, C was distributed with Unix without additional cost.
Therefore many Universities and finally Microsoft and Apple used it.
I can remember when it was reported, C is faster on a Unix computer than a machine natively designed for executing Pascal P-Code and LISP on a Unix Computer (Sun-SPARK) was faster than LISP on a dedicated LISP machine and cheaper.

[Mr.Madguy]

Dunno, why Pascal declines. I guess, that's because Delphi is too expensive to be popular (not sure, how it still lives) and FPC/Lazarus isn't treated as professional tool. For example I'm in difficult situation now. Delphi is too expensive and I can't update to latest version. My old version has too many restrictions and even bugs. I've got tired of that "out of memory" problem. Community edition doesn't suit me, cuz I can't use it on my work. And after 12 years FPC still isn't Delphi 2009 compatible.

Biggest problem with understanding such problems - is that complex tools aren't needed, if your project isn't complex. For example when you need to duplicate some methods from class implementation - then you can just copy-paste them. But it isn't that easy, when you have hundreds and thousands of methods. Things like copy-pasting start to be clunky and crutchy. Because it's always better, when things are consolidated in one place. It's easier to track changes in this case. And therefore better syntax/language instruments are needed to make code reusable.

All my unit tests use closures. It has taken months and years to develop this code. I just can't rewrite it without using closures. Describing them as objects would multiply amounts of code, I would need to write. My tasks also make me feel, that I need multiple inheritance to complete them properly. Because I need scary thing, called "multiple virtual polymorphism". I.e. I want to assemble one object from several others without using object composition or aggregation, because it's UGLY. For example it requires passing pointers to other objects to every object and also tons of copy-pasted stubs. That simply wastes memory, CPU cycles and lines of code. So, I really start thinking, that may be I shouldn't waste my time and should migrate to C++ immediately? Problem is - I just don't like C++ syntax. And I've got used to Pascal.

So, for now I decided to stay. I just need to overhaul my code a little bit to make things easier and therefore faster. But closures are still required. They prevent me from fully migrating from Delphi to FPC/Lazarus.

[jamie]

You do realize all that sugar Delphi gives you is just doing all of what you have described under the hood..

I have converted a few Delphi apps that use those syntaxes and I don't see any advantage as for speed and code space between them.

 I have studied the ASM code of Delphi in those regards just so I could get good understanding of their features and found that most of them are just sugar syntax to fool the eye.

 All this does really is maybe allow you to punch out a piece of code a little faster but not by much, the end results in speed and space I have not really seen any benefit and I like any thing that gives me an edge, be it a smaller file or extra speed.

[marcov]

I use Delphi at work for the more major apps, and Lazarus/FPC for inhouse work and the rest.

For the more major apps where I spend most of my time (90% of my pascal time), it is mainly the debugger that I need. Not that it is perfect, but still a level better than what lazarus provides.

Pricing is a concern, but more from a "what do we get for the money" than a "we can't afford it perspective", we currently use a pre mandatory subscription version (Seattle).  Before that we upgraded roughly once every 3-4 versions.

We also strain under the maximum install limits. Both that the old system with resetting via the website worked faster than the current "just call" scheme, and also because of threats of not honouring reinstalling of old versions any more, which is a business risk.

In short, you get the feeling from all sides that you are being held hostage, and even if it is partially true, it doesn't feel nice.

Language wise, except generics, most newer language feats are nice to haves at best, and I don't really understand how people say they can't live without it. Might just be a different bizz though.

I hate the fact that Delphi/x64 single math is slow (translates to double under the hood), that the debugger often fails to step out of the method properly (F7/F8 on the end of a method), and that the editor random hangs for a few tens of seconds, related to the highlighting/codetools parser. Also delphi seems to have problems with relative paths in the project file.

[Peter H]

A problem with emba is: big companies get discounts and individual developers pay much more.
However Delphi is perfect for individual developers and small companies and not so perfect for big companies.
I myself are hobbyist and get no money out of programming and disagree with their policy, forced upgrades each year with the community version and always one version behind, no updates known (heavy) bugs remain unfixed.
I also installed C++ Builder -and had to remove Delphi therefore- and found that Builder is unusable with C++ 17 (Clang) despite heavily advertised. The compiler works, but the IDE not, you cannot search for declarations, if you are new, you need this.
Therefore I deleted it and now I am at Lazarus and Code::Blocks + wxWidgets which also has a nice RAD GUI design Tool (wxSmith) and I have GNU C++20  .

[440bx]

I myself are hobbyist and get no money out of programming and disagree with their policy, forced upgrades each year with the community version and always one version behind.

That issue and related issues are mostly specific to Pascal/Delphi.

if the Borland descendants had not treated Delphi like a cash cow - ever increasing its price - they would quite likely have a much larger market share both in the corporate market and the hobbyist market than they do today.  Delphi would quite likely sell very well at $99.00 and, even at that low price, it is also likely Delphi would generate more revenue for the company than it does now.  They could charge extra for "premium support" to corporations (to provide guaranteed response times and other perks) which a good percentage would likely go for.

They (early Borland) had a winning formula and they (later Borland and descendants) learned how to mess it up and never looked back. Instead of Live and learn, they Lived and screwed up. Go figure!

[FPK]

I hate the fact that Delphi/x64 single math is slow (translates to double under the hood)

Did you try to play with the $EXCESSPRECISION directive?

[trev]

I use Delphi at work for the more major apps, and Lazarus/FPC for inhouse work and the rest.
[...]
and also because of threats of not honouring reinstalling of old versions any more, which is a business risk.

This!  It wasn't just threats. EMB refused to resolve a licence issue I had with one of their tools (name escapes me now) because I did not have a current maintenance subscription. So much for "perpetual licences".

As a result, I abandoned my then newly purchased and rather expensive investment in Delphi Pro (Berlin) as soon as I discovered that FPC + Lazarus had improved substantially since I'd last used it around 2009 and actually had a working x64 macOS compiler when Delphi still didn't. My money and effort now go into FPC + Lazarus.

I told EMB this when their Asia Pacific Marketing Manager emailed me to find out why I wasn't renewing my Delphi maintenance subscription. The fact that EMB bothered chasing a lowly Delphi Pro maintenance subscription renewal was interesting in itself.