What is going on with Qt these days...

[A.K.G.]

We have been in contact with KDAB and they wanted to charge me 17 000 euros for 1 week training. I have this is writing if things are questioned. I find this a total outrage.

I am in Philippines, and KDAB also gave me a price for a 1 week seminar for 150 interested people joining at the university. The price was over 5, 119, 950 pesos. Over 5 million pesos. After traveling expenses to get here and hotels, they have 5 million pesos in profit.

To put things in perspective - for this ONE week that someone from KDAB will come here, you can buy almost 4 brand new cars in cash from the dealer down the street from where they are request to hold the seminar.

One Toyota Hilux is 1,235,000 million pesos. So they could buy 3.8 cars of this type to be exact.

If Toyota makes a profit after manifacturing an entire Hilux, then shipping it here and selling it, it becomes nutts that KDAB neads 3.8 times more money to send a person here to talk for a few days.

KDAB is associated wit Qt, if not synonymous - and this destroyed all trust.

The next issue I want to address is the documentation in Qt3D.
The reason why we asked for the training and seminar is that the documentation is very bad, and very much lacking compared to the rest of Qt that we are so used to.

KDAB agreed that the documentation is not good, but could not offer any documentation by email, or refer to any examples or give us any other material at all. He only offered training and seminars. That resembles a high school project to be totally frank. Certainly not something worth 5 mil pesos for a week.

KDAB is totally trashing the Qt standard, bot with lack of documentation, and with prices that seem ... i have no words ...

There is more... what is up with the QML ???

The majority of all you get in to when dealing with KDAB is QML, but we want Qt to remain a C++ library, so what is going on here ??

In my circles people are turing away from Qt right and left because of KDAB.

We are used to proper documentation, we are used to sensible pricing, we are not interested in having everything explains to us in QML... or not explained at all (lack of documentation) we are c++ people that expect the Qt standard to be maintained and improved.

I am really worried about the future of Qt. We know that Qt will survive and be great, but it might turn more and more over on a different path. ... if this kind of stuff is allowed to be incorporated in to the product on a regular basis.

I feel Qt is better off without Qt3D and KDAB, because they have over time been a force that is dragging Qt of course.

I am really disappointed actually, that my company have invested large amounts of money in Qt, to be in a situation where our foundation is shaking because of a KDAB. Sure the Qt3D might be great, but what C++ developer wants to stop using C++ ?? None !! So why include all this QML ?? There should be a seperate QML library if some people love QML so much. Keep it out of the main Qt.

What person wants to pay 5 million pesos for 1 weeks trainig?? None !!

And what person wants to invest in something that has no literature/documentation ?? None!! Yeah, there is documentation but it is like this. For example: count() It said it holds the count. That is great to know, but the count of what ??

Let me add that KDAB did not even reply when I commented on their prices.

Sure, KDAB has the right to charge what ever they want, and if they think they are worth that price, good for them. But this is a Qt Partner, and that makes things different. This is effecting all of Qt, and therefor also our situation.

Anybody else have any thoughts on this matter ???

[SGaist]

Hi,

Disclaimer I do not work for KDAB.

First the technical stuff:

• QML is only one module in Qt, C++ is not going away anytime soon.
• The maintenance of the widgets module has seen an update (since you seem to be more interested in that)
• The Qt3D module provides both C++ and QML interfaces and AFAIK, it's a 1:1 scenario (which is not the case for the rest of the QtQuick related modules).
• The lack of documentation is a known problem and AFAIK, the folks behind Qt3D are striving to improve that as fast as they can and also provide higher end example for the modules.

As for concerns about the Qt reputation and future, you should rather contact the Qt company directly. This is a community forum so you are not likely to reach the people you would like to.

[A.K.G.]

@SGaist Than you for your time and your reply, I really appreciate it.

I already contacted the Qt company directly.

It is not hard to improve the documentation. These people just do not have it as a priority. I have been a developer for over 20 years, and the Qt3D documentation could be dramatically improved by just a few days work by the correct people.

This is a clear priority decision, and there is no excuse for releasing a "high-school student" standard documentation in Qt.

With such a lacking documentation, it is close to impossible to use the 3D library for people - unless they are very experienced with such library already. Anybody can see that for themselves in every release of Qt that ships with Qt3D.

I recruit people to use Qt on a regular basis, and the outstanding documentation for widgets, makes it so easy to convince people to change over. Now If they see the Qt3D library documentation, it is all over and they just look at me like I'm stupid for suggesting using Qt3D... so what ever anybody have to say on the matter... they must acknowledge that these facts tell a story of how negative it is to release something prematurely (or with deliberate lacking documentation).

I simply can not agree that they are improving the documentation as fast as they can. I have followed the Qt3D library from the beginning, and there I NOTHING IMPROVED over YEARS !! This is an attitude/priority problem in KDAB !!

You can freely quote me on that to anybody. I seldom voice negative opinion about anybody or anything, but this time I feel that the need for making money is degrading a product that is central in my own business. They are to focused on offering training, and hinder people from figuring things out from the documentation. That goes against the spirit of Qt. Again, look at the widget documentation. Very few people needs any training after reading that documentation. Qt3D will never get such documentation because KDAB would not have many customers for training if they did.

QML i only one module in Qt - yes, a module that should not be there !! It should all be in a seperate library.

You can google Qt3D and see how much documentation, videos, tutorials you can find... you can not find any at all !!! With all the rest of Qt that KDAB does not controll, you can find a ton of stuff everywhere...

There are just to many things at the same time, for me to ignore how deliberate this all feels after several years waiting for that documentation...

[tekojo]

Hi,

KDAB is a consulting company and they have their reasons to price trainings at they see fit. I do not see how this affects Qt overall.

If you don't like Qt3D or don't want to use it, don't. It is not a mandatory part. Widgets and QML will continue to work happily if you never include Qt3D anywhere.
The same obviously goes with QML, if you do not like it, do not include it in your projects.

Comparing widget documentation that has 20 years of work in it to 3D documentation which has a couple years of work in it, is a bit absurd.

Also KDAB as a consulting company puts consulting first, as that is what pays the salaries of the developers there. They are very pro-open source, doing as much as possible to improve Qt and any other open source software they use.

[A.K.G.]

@tekojo said in What is going on with Qt these days...:

Hi,

KDAB is a consulting company and they have their reasons to price trainings at they see fit. I do not see how this affects Qt overall.

I answered this in my post in the other thread.

If you don't like Qt3D or don't want to use it, don't. It is not a mandatory part. Widgets and QML will continue to work happily if you never include Qt3D anywhere.
The same obviously goes with QML, if you do not like it, do not include it in your projects.

I never said I do not like Qt3D, and I never said I dont like QML. Please do not put words in to my mouth.

Comparing widget documentation that has 20 years of work in it to 3D documentation which has a couple years of work in it, is a bit absurd.

I am a software developer, I have been doing this kinds of stuff for more than 20 years. I have been a leader of software companies that both write code and document that code. The quality level of KDAB's documentation is very poor. Writing documentation is something you do alongside the code. It takes second to put in a line that tells you what the cont() is for. When KDAB writes that it holds the could, well then they have said nothing. It should not take 20 years to write what the heck they are counting... seriously, I understand that you find my post a bit overboard like you have said, and I am going to try to be more sensitive towards your opposite view on the matter, but this is simply a question of very bad documentation. It is not appropriate to put that in to Qt, because Qt is not KDABS product. They do not have the right to degrade Qt with this kind of documentation.

Also KDAB as a consulting company puts consulting first, as that is what pays the salaries of the developers there. They are very pro-open source, doing as much as possible to improve Qt and any other open source software they use.

I see your point, and I appreciate all that KDAB has done, especially in the Qt3D, but no matter how much I like what they have done in the code, I can not ignore the fact that there are negative sides as well. If I come to a business and demand 3 and a half brand new pickup trucks from Toyota as payment for a spending 5 day with them... there is not one person in the world that would find that to be an unacceptable price. You seem to be so very happy with that price, but would you pay it yourself ?? You are talking as it is all good to charge that much, but I bet a lot of money - that you would never pay this amount yourself.

So why would anybody pay that much ?? Only if there is no other alternative. And to push a product in such a way that you create these dependencies where there is not other alternative, that is a very negative marketing for that product. I have like I said, done this for over 20 years. And I know you would not be willing to pay this price yourself, so its kind of strange to me that you find it ok to accept something for others that you never would have accepted for yourself.

KDAB is not a bunch of movie stars. I once had 3 people from Microsoft spend a week with our company, so I know what the big league is charging.

[tekojo]

@A-K-G

I do see the point of Qt3D not being on the level where it should be. Both in terms of documentation, and also on the abstraction level of the API. It is still too close to OpenGL, that is something that development is working to fix.

What I recommend, is to go on the Development mailing list (http://lists.qt-project.org/mailman/listinfo/development), and voice your concern in a sensible fact driven way. Shouting out prices for a training isn't really productive. The core problem you have, is that you need(?) Qt3D somewhere and with the current documentation, it is too much guesswork to make things work.

[A.K.G.]

@tekojo I appreciate your view on this - but I have done both. I am in dialog with JC Kim, the Senior Channel Manager APAC for The Qt Company.

Thank you for the suggestions to bring this to the Development mailing list by the way, I really appreciat your help and input.

My dialog here is to hear what others think about this. I am not here to resolve any issues, there are as you said other channels for that. This is The Lounge, where one can discuss any topic.

[tekojo]

@A-K-G
Yes, certainly good to discuss here too, but to solve the problem the Development list is the place.

Qt3D isn't complete, and as you say the documentation isn't there yet unfortunately. And as said, I believe the bigger target is to have a higher abstraction level api to 3D. Right now it is too close to OpenGL, requiring some understanding of that too.