posted January 04, 2007 09:48 PM            

The thing that has really driven the whole PC revolution has been the applications. VisiCalc was a 'must-have'. WordStar changed the concept of the typewriter, and the list goes on.

People generally don't care what gets them there. If you have an application that runs only on a MAC and that application is the best (perhaps only) one, people switch to Macs.

But where to people learn about these applications and find out what they can do for them? A good portion of the time its people responsible for such things.

What I see happening over the next 8-10 years is the current group of jr/sr high kids who are very Linux literate, will start to enter the job market and show the "old-timers" things that can be done on Linux cheaper and with the same precision as their Windows counter parts. Additionally, there will be a huge increase in the number of experienced Linux technical people to support the platform driving the support costs down too. This increased awareness will generate far more interest by developers to create "must have" applications that run on Linux systems. In fact, people will probably develop more for Linux because the size of the market will be growing and there will (initially) be less competition.

I understand what you mean by the open-source pundits refusing to accept the idea of commercial software, but that is really a non-issue in the next revolution. Business people understand that there are costs associated with technology but when you remove the OS license costs, you put a huge dent in the competition. Remove the $100k just for seat licenses, and you can easily justify an moderate increase in application spending, perhaps to have the software customized rather than being forced into the software's mold. This creates more developer demand which creates more interest in supporting those systems.

Its not an anti-MS issue at all, but an economic one. Even/especially the home consumer will eventually choose the "free" Linux OS over Windows as soon as they are comfortable that support is available (ie: all the kids understand it) and all issues with compatibility (my Linux-based Word processor must read MSWord docs and visa-versa) are ironed out, its a natural. Hardware vendors would love to cut $100 off the top without losing a penny in profits so you know they are going to push this.

Its all a simple matter of time and I firmly believe its already a done deal, we just don't see it yet.

posted January 04, 2007 10:44 PM            

That's part of the inherent inertia, however. Right now, it's more cost effective for them to keep Windows as the only platform. And changing that is going to more than a group of upcoming programmers and system administrtators who are familiar with Linux. It's not like Linux was just created out of whole cloth and it's a radically new paradigm. Under the covers, it's still UNIX.

Companies that think they're going to be saving money by switching to Linux on the desktop (as it is now) are setting themselves up for a rather rude awakening. Something tells me that you're vastly overestimating the technical capabilities of the average office worker or home user. When you start factoring in costs of actually maintaining open source software solutions, the differences really start to diminish.

Let's say that your theoretical company saves that $100,000 on operating system license fees. But then they find that they have to turn around and hire another system administrator to maintain those desktops and provide support. So, yeah, you're not paying Microsoft that $100,000. You're paying a Linux admin $70K/year. With benefits and taxes, you're paying more for another warm body, and the privilege of jumping through hoops to make sure that the stuff that "just worked" with your Windows environment is equally functional under the new Linux setup.

If what you're saying comes true, I guess it's good news for Linux admins, but not necessarily the businesses who currently have a huge investment in the Windows infrastructure. Telling shareholders that the benefits will manifest themselves 10 years down the road, but in the short term they're going to take it in the shorts, doesn't usually make Wall Street happy where they live and die on 6 mo. cycles.

I think things are a whole more complex than "Linux kids are gonna change the world". It's not like it was back in the 1970s and 1980s when the industry was still very immature. At a certain point, when systems become well-established, they start to inherently resist radical change in favor of incremental change. Even though technology moves along at a much faster pace, it still fundamentally follows that rule. Think over the past 20 years. It's no conincidence that computers, software and the Internet have become increasingly commercial, "revolution" has been increasingly replaced by "evolution". There's been no truly radical, revolutionary changes to the industry for years.

In any case, in 2017 we'll have this conversation again and see where things are.

posted January 05, 2007 03:35 AM            

quote:

---

Originally posted by Mike Stefanik:
Resisting change because it represents "irrecoverable costs" is like saying that you want to resist the rotation of our planet because nighttime increases the amount of electricity that is used for lighting.

---

Er, no. Controlling costs is s pretty widespread business practice. Adapting is what I do, conforming to an irrelevant norm isn't!

posted January 05, 2007 05:31 AM            

quote:

---

Originally posted by Nick Luick:
Does anyone remember Big Blue !!

---

IBM revenue (2005): $91.1 billion
MS revenue (2006): $44.2 billion

If that's the result of being forgotten, please forget me immediatly

Knuth

posted January 05, 2007 07:18 AM            

nobody ever got fired for recommending IBM!

posted January 05, 2007 10:06 AM            

quote:

---

The thing that has really driven the whole PC revolution has been the applications

---

quote:

---

..over the next 8-10 years[] the current group of jr/sr high kids will ..show the "old-timers" things that can be done on Linux cheaper ..

---

.. In the same post?

Either it's the applications or the operating system... and for my $0.02 it's NEVER the operating system.

And let's not forget the development tools, either: Methinks one can do a user screeen faster and cheaper (labor costs) using Microsoft Visual Basic than with PowerBASIC. Matter of fact, I think the PowerBASIC people believe that, too... which is why they added 'DDT' tools and now the XPRINT/GRAPHICS stuff to their development tool offering.

I've done development on IBM and Burroughs/Unisys mainframes, *nix boxes from only the Almighty knows how many different hardware and software vendors and PCs flavored MS-DOS, Win/16 and Win/32.. and you know what? Parts is parts! It's just a box!

posted January 05, 2007 10:52 AM            

I don't want to start a religious war over whether that's necessarily a good thing, and I know all of the arguments that Visual Studio turns programmers into "component assemblers" who don't really understand how to program well. But the reality is, that is the future of software development, like it or not. And that is how future applications, on the desktop and on the web, will be built.

One of the reasons that Windows really took off was the release of Visual Basic and the veritable flood of applications that followed (both good and bad). Unless Linux can achieve that kind of critical mass with the development community, I really don't see it happening.

posted January 05, 2007 10:56 AM            

The reason I can say both in the same post is by keeping them in context.

Applications have always driven this business, but when you reduce application costs, you fuel application development and investment. Business 101.

posted January 05, 2007 11:04 AM            

 IF (Application Development platform and tools ) > (Application Itself) THEN
   GOTO Bankruptcy Court
 END IF

posted January 05, 2007 11:57 AM            

Added: Not the only one, then.

------------------
kgpsoftware.com - Downloads
kgpsoftware.com - Development and Consulting

posted January 05, 2007 01:13 PM            

My predessor where I work wrote a huge app our agency used for
I'm guessing 20 or more years. It was a mainframe app written
in Algol. It was a labor of love for him and it was superb.
When he left we had to change because Burroughs >> Uinsys was
dropping support for the mainframe, and we spent millions redoing
the app using SAS technologies. That ran about 5 or 6 years
and then my PowerBASIC system replaced that. But it seems the
apps are lasting shorter and shorter periods of time before they
are replaced.

To some extent that figured into my choice to use PowerBASIC. I
looked at the huge .NET documentation (the entire .NET 2.0 SDK
is I believe 230 odd megabytes which unzips and installs to
several gigabytes), and I knew I always come down on the side
of the issue where you write superb systems with unbelievable
features that just WOWs people, but you have to become truely
expert in the development platform to do that, and I didn't think
I had enough life left in me at age 52 to become expert in a system
that comprised several gigabytes of documentation. The other
thing I knew with absolute certainty was that long before I could
achieve a high level of mastery of the .NET framework, Microsoft
would be anouncing its new replacement with unbelievable new
features that everybody had to adopt immediately (at that point
the .NET framework would be 'deprecated', and everybody would be
urged to move to the latest and greatest as soon as pssible). I
really suspect people at MS are working on it already.

But the thing is, even with the little .NET programming I know
I can throw together simple apps that work somewhat. Maybe even
good enough. They won't be fancy or great, (actually, they'd be
awkward as ****) but why bother with that when they'll be replaced
in a year or two at most?

So, I'm wondering if perhaps this should be the plan. Every new
development system or upgrade MS comes up with, get it as soon as
possible (get it in early beta if you can). Then get the easiest
how to book for beginners you can find (one with lots of pictures
and good explanations of how to use the wizards), and use that to
quickly drag and drop an app together. Here and there you may
have to write a line of code or two to tie it all tigether, but try
to keep that to a minimum, cuz it'll be lost time when you upgrade
in a year or two (probably a year only) and won't work with the new
version.

When people use the app they'll double click on it and the hourglass
we'll appear and the hard drive we'll churn for five seconds or so
while the .NET framework starts to get its immense bulk moving
(users now days don't even know anymore where an app is running from,
being as network use is so pervasive, and the time it takes to get the
app moving we'll appear to be network related).

And I believe most of the grief I cause myself in programming is due to
my resistance to that whole model. I tend to want to make the apps
really good and lasting rather than just barely workable. But I have to
agree with Mike, I think that is the future. Very high level graphical
development with objects.

posted January 05, 2007 01:28 PM            

And, as we've discussed, if you want some flavor of Linux to eclipse Windows, you're going to have to convince the Windows developer community that it's worth the effort. And that community dwarfs the number of Linux developers by several orders of magnitude. The last numbers I've read were that there were over 6 million professional VB and VB.NET programmers, not counting all of the other BASIC languages like PowerBASIC, C++, C#, FoxPro, etc. There are more lines of Visual Basic code written than any other language (including legacy code in languages like COBOL). Those are the people they're going to need to convince that Windows is no longer the development platform for them.

To whoever that is, I wish them the best of luck. I won't be holding my breath, though.

posted January 05, 2007 05:00 PM            

If you are keen on minimalist systems this will give you the
choice of writing console style or windows style, cross-platform with
the same compiler.

I hope PB will rise to the challenge. It is not too late since
Freebasic is still very much in beta phase.

posted January 05, 2007 05:41 PM            

quote:

---

Originally posted by Charles Pegge:
Power Basic has a younger cousin, and potentially serious rival,
Freebasic

---

Diolch yn fawr Pegge da.