Where is this all heading?

[CommonTater]

First I'd like to say "Thanks" to CLR for posting the link to the Win8 preview... If nothing else it got me thinking about the future and where I want to go with it. 
 
I've always tried to be somewhere near the "leading edge", usually just far enough behing the "cutting edge" that I don't get any on me  ... I've also been at this a long time, my first computer was an Altair (Just like the one in the kid's bedroom in "War Games") I built from a kit back in the 1970s .. I've also been across several programming languages... first hexcode (all those switches!) then ASM then MBasic (on a pre-IBM PC) then Pascal where I stayed for nearly 20 years until switching to Pelles C in late 2004.
 
In the beginning, there was a lot of rapid change and interesting innovation.  Not surprisingly most of it came out of people's basements and garages.  There's hundreds of patents out there with people's given names and home addresses on them (including a couple of my own).  The pace was amazing, often things would change so much in a single day it was hard to keep up...
 
But "the good old days" are gone now.  Now it's all robotic assembly, computer design and under the control of a relative few large corporations... the little guy is less and less in the picture as time goes on.  I think we all have some sense of how things have changed in the last couple of years... with some very real challenges to any private developer or inventor.
 
Finally, with my little dip into Windows 8 came two realizations:
1) Windows 8 is SO different and SO bad that I want no part of it.
2) Everything has become so corporatized that pre-packaged meat looks original beside it.
 
So, I'm a little curious to hear other people's opinions about where this is all going...
 
What is the future here?  Do we even have one?
 
 

[Monolith]

I would say that we have about 7-8 years before everything has to be plugged into the internet to do anything, even to log on to your OS. Ads will be everywhere. Thanks to people like Pelle, GCC developers, Free Pascal developers...,  developing for the little guys, the individuals, should keep going. I suppose that at a certain point, native code development will lose hold and most everything will have to be developed with GUI kits and high level languages like Python. Hopefully not any time soon though.

[CommonTater]

I would say that we have about 7-8 years before everything has to be plugged into the internet to do anything, even to log on to your OS. Ads will be everywhere. Thanks to people like Pelle, GCC developers, Free Pascal developers...,  developing for the little guys, the individuals, should keep going. I suppose that at a certain point, native code development will lose hold and most everything will have to be developed with GUI kits and high level languages like Python. Hopefully not any time soon though.

Excellent thoughts... thanks for posting them.

On the topic of languages...

We've had half a century to work out programming languages and paradigms.  If you ask me Pascal came the closest... a nice straight forward procedural language with real strings, sized arrays and garbage collectors.  It was easy and fun to work with but had "bare metal" abilities of C, when needed.  Of course Borland got into it with OOP, sent everyone off in the Delphi direction and, in the process, killed one of the best Languages ever devised (in my opinion, of course). 

Since then it's been rather primative... even the mighty third letters -- C and C++ -- are lacking core level support for crucial features such as strings.  It's like we're half a century into this and already becoming stagnant and stodgy.  One might reasonably have expected C to get a native string type on it's first revision back in the 1980s ... it is afterall the most used variable type in modern computing.  But instead we are still working with pointers to arrays and arrays of unknown dimensions and size.  Heck, even lowly MSBasic had strings! 

It seems almost as though we're trying to write the future from inside a time warp...

I also agree that advertising is getting way too high handed... Not long ago there was a group in Toronto who wanted to put advertising on the FLOOR of the subway stations...  Enough is enough guys... The last thing I want is advertising on my desktop. 

[Monolith]

I do agree with you that Pascal is one of the nicer languages. I appreciate the case insensitive coding. It allows for much cleaner code. In C/C++ you have to type the code according to other people's naming conventions (Windows - Hungarian, C - lowercase, C++ - lowercase with underscore...).

Since then it's been rather primative... even the mighty third letters -- C and C++ -- are lacking core level support for crucial features such as strings.

Speaking of strings, at some point the C/C++ standard should call for making all literal strings inherently Unicode. Sometimes it is necessary to cut strings with old standards.

[CommonTater]

Speaking of strings, at some point the C/C++ standard should call for making all literal strings inherently Unicode. Sometimes it is necessary to cut strings with old standards.

At least have a pragma or compiler switch...

[czerny]

I've always tried to be somewhere near the "leading edge"

As a young man I thought, that all what not can be calculated by mental arithmetic is not worth to be calculated at all. Then, I don't know why, I bought an atari 800xl and startet programming in 6502 Assembly. I coded in Turbo Pascal for very long time, too.
But today the circle slowly is closing. I can not found anything realy worth to be calculated. Maybe in two or three yeares I will put my computers in mothballs and consecrate myself to the essential things of live.

czerny

[CommonTater]

As a young man a thought, that all what not can be calculated by mental arithmetic is not worth to be calculated at all.

LOL... with today's generation that would end someplace around 12 x12 .... I don't know about where you are but in Ontario Canada, most teenagers and young adults don't even know how to do long division and can't handle multiplication or addition --even easy stuff-- without a calculator. They learn that getting the right answer is more important than understanding the process which has the unexpected result that many of these same people now treat technology as magical without even caring to understand how it works... those who do exhibit curiosity about substance and process are generally labeled "geeks" and ridiculed.  And we have the nerve to call ourselves a progressive nation...

[Bitbeisser]

We've had half a century to work out programming languages and paradigms.  If you ask me Pascal came the closest... a nice straight forward procedural language with real strings, sized arrays and garbage collectors.  It was easy and fun to work with but had "bare metal" abilities of C, when needed.  Of course Borland got into it with OOP, sent everyone off in the Delphi direction and, in the process, killed one of the best Languages ever devised (in my opinion, of course).

The fact that (and how) Borland added OOP to Pascal isn't by itself anything that killed Pascal as a language. After all, it is still Pascal and you are not forced for the main part to use OOP for your own programming logic. It's the switch to Windows (or any other GUI based OS) that makes you use certain amounts of OOP programming to interface with the GUI and a lot of OS API functions.

I am still using Delphi (and FreePascal for adaptation to Mac OS X and Linux) as my main programming environment, currently working on a fairly large multi-server/client application. I barely use OOP, rather stay with a more "old school" approach, using OOP only where I am more or less forced to. Works just fine, and is IMHO much more readable and most of all maintainable as it is clearer what a function/procedure does than stuff that is hidden deep in some class/object dependencies where it is less than obvious what possible side effects are...

I started programming in Pascal 36 years ago (well, in July, with a basic Wirth compiler on a PDP-8), added 6502 assembler and Tiny BASIC on my first own computer (KIM-1) and then got into Z80 and various other assemblers as well as various BASIC dialects, as well as early C, Fortran and COBOL compilers even before there was the first IBM PC.
Since then, I have at least tried every programming language that came along and "en vogue", though very few could convince me to even start and do anything serious (among them FORTH and Perl). And today, I think there isn't anything that I can't do with either Pascal, C and assembler. (<=that's a period!)

It looks to me that nowadays not only the computer users have been stupefied but a lot (most?) of programmers (if you still can call them that) as well.
Rather than to think about what one is doing to make a program more efficient, well "let's throw some more RAM at it"! More speed efficient algorithms? Hell no, let's just get a faster CPU!

And in turn, that leads to the abyss we now see with Windows, Mac OS and as well with all the recent Linux distros. Bloat where ever you look, fancy new user interfaces that someone thinks is "cool" and doesn't care if anyone else is able/willing to use it...
People (like the clients for my before mentioned project) nowadays just lough or don't even start to understand when I tell them that there were times were the same basic functionality of an application was achieved on a multi-user system (called "mini-computers back then, the size of an estate fridge) with 256KB (yes, "Kilobytes", that 1024 single bytes) of RAM all together, serving up to 80 text mode serial terminals, with a restriction of 12KB for both code and data per active user!

Ralf

[Stefan Pendl]

It looks to me that nowadays not only the computer users have been stupefied but a lot (most?) of programmers (if you still can call them that) as well.
Rather than to think about what one is doing to make a program more efficient, well "let's throw some more RAM at it"! More speed efficient algorithms? Hell no, let's just get a faster CPU!

More RAM and faster CPUs are mostly the solution for gamers.
A utility can always benefit from better algorithms and there is always room for improvement

[DMac]

In the late 70's my Dad brought home an AppleII from work.  At that time the state of Minnesota was beginning to implement computers and educational software in the schools and he was on the "cutting edge" of the educational computer revolution.  This marvelous machine was carted around to various educational conferences that Dad attended and sometimes found its way into our basement.  I was fortunate to have access to such a device at that time and a better boy would have taken full advantage of it.

As it was all I learned was "Load" "List" and "Run".  The only program that was of interest to me was an apple basic demo titled "Amazing"  that I would use to make the old Teletype model 43 Dad was using for a printer, print reams of greenbar with mazes that I would carry off to a table and solve by pencil.  The cool thing about that Teletype model 43 was that it functioned as a terminal to the Apple II.  I could type from its keyboard and the Apple II would load and run Amazing just as easy as if I had used the AppleIIs keyboard.   What about writing something, perhaps a game in Apple basic?  Boring!  I couldn't be bothered with such stuff.

By the time I was in high school Dad was bringing home a Commodore 64 and none too soon either because the Commodore had some kind of word processing program on it with *Spell Check*.  It was great, no more chunking away at the old Royal typewriter for an hour only to find that I had made a mistake in the first paragraph that I couldn't easily remedy.

Now that Commodore was a great piece of hardware.  It had a good solid (probably MilSpec) keyboard.  It was a pleasure to tickle Qwerty and listen to the soft click of the keys.  It also had a high resolution color monitor (Still works 30 years later and makes a great mini TV with the video input) and an industrial grade, 9 pin dot matrix printer.

My nephew visited us one summer and played with the cool tech stuff Dad had lying around, got interested and went on to pursue a successful career as a talented software engineer.  Me -- no interest beyond getting a few papers off in my senior year of high school.

Fast forward to the early 90's, I was in grad school and bought a used Tandy Tl 2 1000 from a friend with Norton Commander installed and a copy of WordPerfect for DOS 5.1.  It was *da bomb* there wasn't anything I couldn't write using that system.  The hardware was built for a tank and I remember one summer doing a practicum in Berea Kentucky.  The dormitory lacked air conditioning and I was on the sunny side of the building.  Must have been 90+ degrees in the room and I'm typing away in nothing more than a t-shirt.  I finished off the course paper and retired to a more comfortable location where I learned that several of my fellows had not fared so well.  Their boxes had overheated and were useless!

My first foray into programming anything was using WP DOS's prn.exe to write a driver for my printer.  I spent quite a few weekends perfecting that driver and was quite pleased with it in the end.

Fast forward to 1999, I was newly married and in California after several years abroad.  I happened to be working in an engineering capacity at a start-up by some miracle because my background was not engineering.  We needed to document stuff and I invited the AutoCad jocky, who did drawings for us on the side, to dinner several times and got him to show me how to work AutoCad in the evenings after supper.  Eventually I was doing a lot of CAD and even went on to pursue a short career in Electronics design and drafting.  While working in that capacity I discovered AutoLisp.  With numerous examples and trusty notepad I was able to figure out how to write an actual program that did something useful!  I wrote, among other things, a routine that modeled twisted pairs of wire along a 3 dimensional path.

After just one layoff that career came to a screeching halt!

I got a job as a bench tech, for a year I was rubbing shoulders with a LOT of test equipment.  Took a class at the university extension to learn VB.NET (I could only afford one) and began writing Test Automation applications.  I wasn't satisfied with VB.NET and wanted to see what was happening under the hood.  With the help of BCX I taught myself C and wrote applications in that, since it was more fun than VB.NET.

Another layoff and for 3 months I really hustled to find a job.  At the time I cast out three lines, applying for software engineer, draftsman, or tech and waited to see on which one I'd get a bite.  Landed Sr Tech and turned it into a software engineer in the first two weeks.

I had read somewhere that the C# developer made more money than the VB.NET developer for writing the same stuff simply due to the fact that he was *perceived* to be more intelligent and professional than the former.  Since I was in the position to call the shots and was familiar with C syntax, I wrote everything in C#.  Several years later I find myself the chief of my own little fiefdom and even have some applications *in the wild* woo hoo!

Where do we go from here?

During each of these chapters of life, I notice that my relationship with computing extended along the lines of need.  First along the lines of my own need as a user and then along the lines of the needs of others as a programmer.  Wherever Microsoft takes it's OS, it will only achieve success in the marketplace if it fulfills the needs of the businesses and individuals who rely on its products.  The new trends will certainly generate new needs.  As the hassle factor increases so too will the tools designed to streamline and bypass the hassles.  After all it wasn't that long ago that developers were hassled with having to develop and debug complex code without the aid of an IDE.

For the time being Windows 7 will likely enjoy a long run and join the ranks of Win98, 2k, and Xp.  Over the next few years they'll work the bugs out of it and then folks will be ready for Windows 9.

[CommonTater]

The fact that (and how) Borland added OOP to Pascal isn't by itself anything that killed Pascal as a language. After all, it is still Pascal and you are not forced for the main part to use OOP for your own programming logic. It's the switch to Windows (or any other GUI based OS) that makes you use certain amounts of OOP programming to interface with the GUI and a lot of OS API functions.

I won't argue the point... I still have a copy of Pascal that I rooted out from under Borlands VCL and I use it occasionally to maintain some of the earlier versions of my software.  But for all intents and purposes Pascal is a dead language, used only by a very few; not even enough to make a blip on the Language Popularity graphs you see on some sites. 

Like you, I try to avoid OOP except when necessary preferring to use classes and objects only when the OS interface demands it.  The rest is written in trusty old procedural style. 

I really miss Pascal... Hardly a day goes by when I don't find myself wishing there were an up to date 32/64 bit compiler and library I could plug into my system and get re-acquainted.  I've tried Free Pascal and have no problem using it for console apps, but the windows units are a massive dog's breakfast and whoever created them should be shot.  What is needed is a set of windows units (32 and 64 bit) with a one to one relationship between Pascal units and Windows headers... If it's out there, I haven't found it. 

[CommonTater]

More RAM and faster CPUs are mostly the solution for gamers.
A utility can always benefit from better algorithms and there is always room for improvement

Absolutely!

[CommonTater]

For the time being Windows 7 will likely enjoy a long run and join the ranks of Win98, 2k, and Xp.  Over the next few years they'll work the bugs out of it and then folks will be ready for Windows 9.

I spoke to a programmer friend of mine and he suggested pretty much the same thing.  He figures windows 8 will be adopted mostly on tablets and win7 will remain the desktop OS of choice.  Seems kind of silly though because there do appear to be some incompatibilities which may cause problems for developers....

Thanks for an excellent post... 

[Bitbeisser]

But for all intents and purposes Pascal is a dead language, used only by a very few;

Well, that's only because people keep prematurely talking it dead...

not even enough to make a blip on the Language Popularity graphs you see on some sites. 

Sorry, but that doesn't mean squat. It rather proves my point that nobody is really "programming" anymore these days.
Those popularity lists are not a true picture of what is really used, for real application purposes...
The most common reference these days is is poplang.com and in their list of the top 20, they have both Pascal and Delphi listed, which for all practical purposes, is the same.
If you add those two up and if you then take out those languages that aren't "general purpose", but web specific for example, (like PHP, Javascript, Actionscript), or entries like "SQL" or "Shell", Pascal/Delphi is all the sudden right there in the top 5-6, together with C, C++, and Java.
And a lot of those languages that show up in the "normalized" data, are only there because of all the hype around them, but nobody is actually doing anything really usable/usefull/generally available with them, things like Haskell, Erlang, Lua or D...
Otherwise, they would barely be a blip on the radar/charts.

In general, there is no better way to lie and produce perception than with statistics...

In real world life, there is far more real, everyday code written still in Fortran, COBOL and Ada then in all those exotics combined. But that's something that doesn't show up in those "popularity" lists...

I really miss Pascal... Hardly a day goes by when I don't find myself wishing there were an up to date 32/64 bit compiler and library I could plug into my system and get re-acquainted.  I've tried Free Pascal and have no problem using it for console apps, but the windows units are a massive dog's breakfast and whoever created them should be shot.  What is needed is a set of windows units (32 and 64 bit) with a one to one relationship between Pascal units and Windows headers... If it's out there, I haven't found it.

Sorry, I absolutely don't see this. And I would not even remotely bother with any "Windows by hand" programming using any Windows units directly, one of the reasons why I dont' bother much to do non-console stuff in Pelle's C.
 I rather create my forms for Windows GUI programs in the Delphi forms editor or for Windows/Mac OS X/Linux in Lazarus for FreePascal instead. So far, I barely had a need to deal with anything of the Windows API for example directly, only for a FTP client software that I started, but then also only as this is a part of this project that is decisively different in those three OS anyway.
The forms source can then be (for the most part) transferred from one OS to the other and might need some slight adjustments to fit the common style of the target system, but how this is actually handled "under the hood", I could give a rodent's posterior... 

Ralf

[CommonTater]

The forms source can then be (for the most part) transferred from one OS to the other and might need some slight adjustments to fit the common style of the target system, but how this is actually handled "under the hood", I could give a rodent's posterior... 

Hola Ralf!

I guess this is where you and I are different programmers... I like the nuts and bolts approach at the windows API level (I don't do cross-platform stuff)  and don't find it in the least uncomfortable.  In fact, I prefer to know what's going on "under the hood" because most often it's me who gets stuck fixing it.  

So what's your take on our future?