TrollTech to IPO?

burninginside writes "Yahoo is reporting that Trolltech, the world's biggest producer of Linux software for mobile devices, may be heading public. 'Sources close to the company' said the move may come as soon as 12 months but the official word is still that it 'is not in our immediate plans.'"

Comments

Will their tools stay free?

[Neo-Rio-101]

As long as Trolltech continues to make a developer kit so that us less financially inclined can just complile and test software on their platform for free, I will be happy.

I'm having fun with my Zaurus, thanks. Keep the free tools coming!

Hopefully a public float won't mean that the shareholders get greedy and cut off the free stuff in order to profit more. Hopefully these new ex-Microsoft execs quit Microsoft to get away from Microsoft's corporate practices.

Re:Will their tools stay free?

[stilborne]

actually, most of the core KDE developers don't work for Trolltech. many work for SUSE and other similar companies, are self-employed or are otherwise occupied with something else.

some of us are funded by TT, but in my case that simply means i get to do what i always did save that i don't have to have a day job.

your take on it mischaracterizes the Free Qt contract and the TrolltechKDE relationship pretty harshly. you're likely either a troll or someone who is very unhappy about KDE's success.

Re:Buy stock?

[Ctrl Alt De1337]

It depends. If you don't mind the prospect of some current or future management of Trolltech possibly ruining the company and losing all the money you invest, then go right ahead.

It's one thing to advocate open source via word of mouth, Internet message boards, etc. It's another thing to throw your money into an open source company without checking it out first. If/when the IPO draws near, do some research, dig into the financial reports, and find out if this is a financially sound company.

Would you buy stock in any random company out of the business section of the newspaper without researching it first? I would hope not. While it's noble to suggest such things as this, let's face it: Trolltech is a for-profit company. It's not running a charity. Do you own Red Hat stock simply because it deals in open source software? If not, then why would you blindly buy Trolltech stock instead?

Re:How complicated could these programs really be?

[Dot.Com.CEO]

Do you have any idea what you're talking about?

Re:Let's don't get ahead of ourselves

[Phosphan]

> C++ and Qt C++ is comparable to java and dotnet in terms of bloat and raw speed.

That's not even comparing apples and oranges - that's comparing cars, toasters and computers.

Let's get this straight. C++ is a programming language (and it's standard library). Java is a programming language, a standard library and a virtual machine. .net is a framework that can be used by several very different languages.

> You can just feel the object oriented speed penalty in both kde and trolltech windows, compared gtk or win32 api c.

Sure. What about comparing programs that actually do the same instead of some that just somehow look similar? There is no such thing as an "object oriented speed penalty". OOP is a way of doing things that's often used even in languages that don't support it. OOP languages just provide some help and eye candy for doing things the OO way. Your beloved gtk API is definitely object oriented. If you don't believe me (and can't see it from the API), just check http://developer.gnome.org/doc/API/2.0/gtk/gtk.htm l and read: "GTK+ has a C-based object-oriented architecture that allows for maximum flexibility.".

This "oop is bad" example is just ridiculous and has nothing to do with the actual experience in any OO language. The opposite is true. Normally you have something like:

something = SomeAllocFunction(something_specification)
someth ingElse = SomeAllocFunction(specification_for_something_else )
result = add_somethings(something, somethingElse)
FreeSomething(something)
FreeSome thing(somethingElse)

... and still don't have any reasonable error handling, which makes the code much worse readable.

In an OO language with operator overloading you would get: result = Tsomething(specification) + TSomething(otherSpecification)

About your "suggestion" to use C as an intermediate language - good morning, there are things that can't (or can only with a some performance hit) be translated into C.

Praising BASIC really is an evil idea. You should't expose novice programmers to this pest when there are so much cleaner languages around which make it harder to write spaghetti code and shoot yourself in the foot than BASIC. For example, you already mentioned Python.

By the way, good morning, Java bytecode does not have to run interpreted. There are both available: Just in time compilers and ordinary compilers.

Perhaps you should also notice that C++ is not an object oriented language. Java is. C++ just allows you to do OOP. Or ordinary imperative programming. Or generic programming. Or mix them all.

Reading your comment once again I really get the impression I am just feeding a troll. Perhaps you are not, but I just can't let it uncommented as "insightful" when you obviously don't know what you are talking about.

Re:So what happens to Qt

[10Ghz]

The problem with doing an IPO is that you lose a lot of control of your company. You might end up being taken over by someone else without your original vision.

Not necessarily. The old owners (in this case: mostly TT-employees) could still hold majority of the shares. You do not have to sell all your shares, you could just sell some shares.

What happens to KDE if the major shareholders decides to stop developing Linux/Qt and discontinue it, while keeping Windows /Qt and Mac/Qt.

Previous version of Qt would get re-licensed under BSD-license, as per agreement between KDE and TT. Also, KDE_folks (and anyone else interested) would pick up the lates free version of Qt, and start working on it. So there would be to Qt's: the commercial Qt provided by TT available on Mac/Windows only, and free-software Qt, developed by KDE and others, available for Linux, Mac and Windows.

Yes, but where does that leave proprietary applications?

If TT did what you fear, those proprietary applications would do just fine. In fact, they would flourish, since Qt would be under the BSD-license. That said, I don't care about proprietary apps. I find it strange that people are pushing free software. But when Qt and TrollTech is concered, those same people get their panties in a bunch because you can't write proprietary apps with it for free. Tough luck I say. Why don't you also whine because you can't take Linux-kernel and turn it in to something proprietary?

This is why having ONE company have complete control over the desktop APIs is a bad idea.

Huh? The toolkit is GPL'ed! I really can't believe when I see people whine because some piece of software is licensed under the GPL! I can see why someone would complain when some sofdtware is proprietary and you are dependant on it, but this is GPL'ed software! But if you don't like it, go use GTK+ and Gnome. No-one is stopping you.

Since when did free software and Linux turn from being about *gasp* free software, in to "we must satisfy the whims of companies who want to write proprietary software!". With attitude like that, why don't you people just stick to Windows?

Re:Sam Kinison once said

[GISGEOLOGYGEEK]

yes, you're right,

Theres absolutely no chance that any company could ever face rough times and demanding investors and survive.

Just look at what rough times and demanding investors did to poor Amazon.

err.

ya.

well then again, its only a 500% return if you invested 4 years ago during the punishment.

Imagine if they had actually made a profit during more than 3 quarters of the 40+ quarters the company has been in business!

lets see how they handle things THEN

Re:Let's don't get ahead of ourselves

[Anonymous Coward]

What would it be in BASIC, or C?

most likely it'd be something like -

complex_type_add(a, b);
a_string = complex_type_to_string(a);
printf("%s\n", a_string);
free(a_string);

Far worse than that; for example, you forgot to allocate the memory for the string. And you may not know how much is going to be required, so you can't safely use a fixed buffer...

A more realistic C example would be

int print_sum (foo *a, foo *b)
{
  foo ab_sum;
  char *string;
  int err, length;
  ab_sum.size = sizeof(foo); // I'm not joking.
  err = add_two_foos(a, b, &ab_sum);
  if (err) {
    return ERROR;
  }
  length = get_string_of_foo(&ab_sum, NULL);
  if (!length) {
    return ERROR;
  }
  string = (char*) malloc(length + 1);
  if (!string) {
    return ERROR;
  }
  length = get_string_of_foo(&ab_sum, string);
  if (!length) {
    free(string);
    return ERROR;
  }
  printf("%s\n", string);
  free(string);
  return SUCCESS;
}

No, that is not at all contrived: programming in C really is often like that. Of course, that was assuming a Win32-style API. Maybe there are saner APIs out there, but Win32 really does make you jump through that sort of hoop all the time.