eWEEK is currently featuring an article on the UserLinux and KDE debacle. Indeed, exciting times are ahead as we forge on with our plans for both KDE::Enterprise in general and the KDE/Debian project. You've heard of cool hacks to make KIO slaves accessible on a system-wide basis, you've heard of GTK/KDE integration (don't be surprised if KDE Gimp, a.k.a. Kimp, makes an effortless return), you've heard of the new Debian tools in development, you've heard of a KDE/Debian live CD. All of this and more is coming to fruitation. See you next year!
Dot Categories:
Comments
"You CAN NOT run a QT purchase past these "Decisionmakers" as long as it's not corporate policy to use it."
Well, that rules out GTK then.
For any company, in house development is a very big issue. The safest route is with the biggest standard - which is MS Visual C++ or Visual Basic. But with this approach comes the whole licensing issue, cost per desktop, one platform, one vendor lock-in etc. Which of course is why many of us have chosen to develop and operate on Linux. So what are the Linux/Win alternatives?
I chose Qt and it is an excellent C++ toolkit. It is well written and stable and thus, allows somewhat rapid development, though not like VB. The development tools (Designer, KDevelop, etc) are now mediocre (KDevelop is still fairly buggy, no GUI integration). The code is somewhat non-standard with moc, .uih, QValueVectors etc. It is very expensive ($3000 per dual-license plus $1000/year (per programmer - not even per seat) and has very strict commercial licensing. As my company expands, so does my overhead. So in retrospect, it costs me more money over MS tools, keeps me in the whole licensing mess, and is somewhat Vendor specific, requiring a programmer to be educated in the Qt way. It is very cross-platform, but is that enough to offset the other disadvantages?
For non-gui intensive applications, there is java (which is free), an entire development package (Ant, Eclipse, Junit, NetBeans, etc) which are also free, and most CS professionals have some java experience. Maybe SWT will overcome the gui issues with java since the gui coding, performance, and visual issues are very real (been there, done that).
For fast, pure C/C++, gui applications there is Gtk/Gtkmm which is freely distributable, free in cost, and is the standard that Sun, RedHat, and Novell have chosen. Is it as pretty as Qt or MSVC/VB? No, but good enough. Is it less productive than Qt? Probably, but the how much is debatable. Does it have the developer tools as MS or Java? No. Are the developer tools much worse than Qt? Not really. If I choose Gtk (or better yet, Gtkmm), I might lose some productivity of my employees (certainly a lot from VB), but gain the advantage of cross-platform, free cost, and no licensing hassles. Is it worth it over MS VC/VB? Maybe. Is it worth it over Qt? Comparing the facts, probably. At least productivity can be addressed with templates, etc.
Most companies are slashing overhead where ever they can. And all software development productivity, no matter what package is used, is viewed as lousy. Thus, when you compare approaches, Qt costs $3000, MSVC/VB costs $1000 (the safe approach), and Gtk costs $0, Qt is going to be a very tough sell. A niche product. And do I want to base my company software on a unique product that is a tough sell?
So I hope this explains why LGPL and low-cost(if not free) development tools are necessary to provide the incentive for companies to develop on Linux instead of Visual C/VB.
Where are the tools and support for GTK? Qt is also a heck of a lot more than graphical toolkit.
For any company, in house development is a very big issue.
[...]
You have no clue what you are talking about. You can do as much "in-house" development with KDE/Qt as you whish. GPL is all about _distributing_!! As long as is stays in-house, you don't need LGPL. Since you don't know this, I assume atht you didn't do the sligtest research. since you didn't do the slightest research, I assume you didn't actually spend $3k for commercial Qt or you have so much $$ that you don't care. Since the former is probable, I conclude that you are a liar and have no clue about any big picture.
Cheers
Hello-
I recently read both sides of this UserLinux (KDE vs. GNOME) debate on the www.userlinux.com site, and I must say, I'm a little perplexed by something and I was hoping that someone could explain it to me. Let me start though by saying that I am a much bigger Mac OS X user then a day-to-day Linux user and that my development group actually uses professional Qt for all of our products (which incidently run on Mac OS X, Linux, IRIX, and Windows) - so you know where I am comming from.
Anyway, Bruce makes these statements:
"But all of the efforts to unify these two desktops do not change the fact that there are two entirely different GUI SDKs. The two competing GUIs are each of a complexity equal to or greater than that of the Linux kernel. For developers and support staff, maintaining expertise in both of two GUIs is an expensive proposition. Many IT shops, when faced with such choices, have decided to consolidate to fewer options in order to reduce expense."
"UserLinux is intended to be a system for business people. Central to its design is a network of competing for-profit service providers, who perform engineering and support services for the system. Because these service providers are basing their business upon a commodity product, there are already economic limits upon how profitable they can be. The difference between one and two GUIs may spell profitability or bankruptcy for some of our service providers. In a similar vein, internal support and engineering staff at businesses that employ UserLinux would like to have only one GUI SDK to develop for and maintain. This is not to say that choice is bad. Rather, it's bad when people aren't allowed to choose."
I certainly can't disagree with much of this statement. It is certainly true the toolkits are complex and there may be greater cost associated with doing it this way. However, I do take issue with one very important point: Apple also maintains separate, competing toolkits - very successfully I might add. You can completely develop an application from the ground up on any one of these toolkits without touching any of the other frameworks. These include:
Carbon - a C/C++ based framework that is more closely related to the "old" Mac OS 9 (and lower) "Toolbox".
Cocoa - an ObjectiveC-based framework that is more closely related to the Next/OpenStep.
Java - well we all know what Java is.
It seems that the real question is whether you want to provide the greatest breadth of applications with a unified look-and-feel for users, along with the greatest number of choices for developers so that there can continue to be the greatest breadth of applications. I think we can all agree that the current level of unified look-and-feel for these applications is certainly not there and that much of this is probably more or less due to the greatest number of choices for developers (perhaps more should be spent on this KDE-GNOME unification front). However, it seems that as UserLinux summarily, and may I say illogically, throws out a toolkit just because they don't want to worry about support, that the bigger problem will be the lack of applications. He makes a very good point that you can freely develop proprietary software using gtk and that is something that you do not get with Qt. But, one could also argue that a very important reason why Qt is better then gtk is so many ways is that it does have a budget for QA/QC. What is really the additional cost associated with providing Qt? Again, analogously, Apple too provides several toolkits, yet they encourage developers to develop using Cocoa. Can't UserLinux do the same thing?
Oh well. Admittedly I really don't know Bruce, but it seems that his logic is flawed. I can't see how this version of Linux will be any different from any other version of Linux - except perhaps that it is less useful since it lacks most of the applications that many people use (without going through an install process).
-Lance
____________________
Lance M. Westerhoff
Chief Software Engineer
QuantumBio Inc.
- "A quantum improvement in drug discovery."
things come to fruition, it's true, and this year seems like a bumper crop!
It' kind of deja vù.
After Gnome 1 -> Qt goes to an OSI aproved licence (QPL was that ?)
After Gnome 2, Sun & Gnome Foundation -> QT goes GPL and KDE Free QT Foundation.
Now with only one of these happening...
- IBM Desktop initiative goes GNOME.
- Novell & SuSE goes GNOME (hey Sun says GNOME supports java apps ;-)
- UserLinux & Debian taking some ground on corporate use
... then Qt will go LGPL, and THIS IS GOOD, we will have two free, enterprise ready desktop alternatives.
All I can say is how sorry I'm going to feel for those UserLinux users who won't be able to use the most beautiful CD/DVD Burning software in the entire world... K3B !
I wouldn't want to use UserLinux purely for that one reason.