We move to c++ :(

[DarkDragon]

Are SDL and OGRE not good?

SDL is good but OGRE or the OGRE wrap PureBasic use along physics systems is not good as is for my needs. It's unbelievable that we can't access some important data about Entity such as transformation matrix used to relocate them each frame. For example just to align a probe box for Steering Behavior, using the current 3D system, you will have to reprogram the entire Matrix calculation system and recompute a transformation matrix before to align the Probe Box. PureBasic for mathematics calculation is very fast but computing the same transformation matrix twice at each frame slow down the entire game loop whatever as fast as your code can be. Moreover, we have to reprogram the entire collision system because once again the data can't be accessed in real time.

This is due only to the fact that libraries in PureBasic as to be simple. OK I agree, but in my point of view the simplicity level as to be raised a little bit to be usable.

You may be surprised but some physic engines don't even allow direct transformation matrix manipulations (or at least not the full 4x4 matrix). They often split it into a 3x3 matrix (but you still can't scale, as it messes up the physical calculations - or they use a quaternion) and a positional vector. That's maybe why PureBasic doesn't allow direct access to the transformation matrix.

[Guimauve]

The Cyclone Physic engine allow that and this the 3D physic engine I'm currently re-write in PureBasic for my 3D game project.
It use Quaternion to retain the orientation but generate a 3X4 Transformation Matrix and a 4X4 OpenGL compliant Transformation Matrix.

Just my 2 cents, If a 3D Game engine not allow to access or generate a 3D Transformation Matrix it should not be used, period.

Best regards
Guimauve

[auser]

I am willing to wait a few months to see if the Linux issues will get solved.

And what would you do if they are still there afterwards? Hope a bit longer? The "bugtracker" here is crap. You even don't know if somebody noticed your problem and thinks that it's worth to do something against or if you are the only guy in the hole world that uses that stuff and you get left alone. For example I've reported regarding an issue with #PB_Any and CatchXML (and no - I don't bump indirect here I've workarounded already anyway). How many guys are using that in the hole world (on linux64) except me? One? Or how is it possible that I'm the first that reported that (compared to the time how long it is already implemented somehow into PB)? My project already had many lines of code. I've even spent some time for a wrapper for GnuTLS (next chapter: In C++ I would not write a wrapper for lots of functions but just do "#include <header.h>" and if I get an array back I would just use it in a simple readable loop and not with some "*pointer + SizeOf(struct_foo)"). Anyway my project was too big to move fast to another language but I don't get hope that it would get fixed soon (or why should I have a reason to do so?) so I've made a workaround and fear that "#PB_Any" stuff for now (not only at CatchXML). Is that wrong? Or what should I do? Ignore that and use it as there was no bug and hope that it would sometimes work?

At first I liked PB very much (I think I've found it about 10 years ago). Today I don't hate PB but I sometimes got the feeling I should excuse myself if somebody is behind me and is looking over my shoulder and is asking what I'm using here. It makes live easier in some parts (creating some quick gui in windows - and use it only in windows) but even more complicated in lot others (bugs, incldues, ...) and sometimes I get stuck at basics where there should not be that limitation (even in terms of multiplatform). Using PB I sometimes feel like an alien compared to the rest of the programming-world.


Greetings,
auser

[Kuron]

And what would you do if they are still there afterwards?

That is something for each individual to decide for themselves. Fred has been back to work on PB only for almost two weeks. He seems to be plugging along.

Personally, I am hoping the final version of 4.61 for the Linux will solve the majority of the issues. If not, then it is time to activate plan B.

At first I liked PB very much (I think I've found it about 10 years ago). Today I don't hate PB but I sometimes got the feeling I should excuse myself if somebody is behind me and is looking over my shoulder and is asking what I'm using here. It makes live easier in some parts (creating some quick gui in windows - and use it only in windows) but even more complicated in lot others (bugs, incldues, ...) and sometimes I get stuck at basics where there should not be that limitation (even in terms of multiplatform). Using PB I sometimes feel like an alien compared to the rest of the programming-world.

I do not disagree. I am merely willing to give Fred a chance to fix things since he has recognized there are problems with PB and he has made a major change in his life so he can address those problems.

[dman]

I think people just need to relax about updates to PureBasic, there is only so much a person can do updating different platforms, just chill people, life is way too short

[USCode]

... SDL is good but OGRE or the OGRE wrap PureBasic use along physics systems is not good as is for my needs. It's unbelievable that we can't access some important data about Entity such as transformation matrix used to relocate them each frame. For example just to align a probe box for Steering Behavior, using the current 3D system, you will have to reprogram the entire Matrix calculation system and recompute a transformation matrix before to align the Probe Box. PureBasic for mathematics calculation is very fast but computing the same transformation matrix twice at each frame slow down the entire game loop whatever as fast as your code can be. Moreover, we have to reprogram the entire collision system because once again the data can't be accessed in real time.

This is due only to the fact that libraries in PureBasic as to be simple. OK I agree, but in my point of view the simplicity level as to be raised a little bit to be usable. ...

I don't write games so I don't pay too much attention to posts about the gaming libraries but when I do read them it doesn't seem like anyone is ever happy with their implementation, so why bother? Maybe it's time to deprecate those libraries. Personally I think Fred's time would be better spent working on other things like improving the gadget (more control over appearance), database (more supported DBs), mail (AUTH support), packer (ZIP support), etc. libraries instead. But I imagine folks that do write games would disagree...

[rrpl]

I don't write games so I don't pay too much attention to posts about the gaming libraries but when I do read them it doesn't seem like anyone is ever happy with their implementation, so why bother? Maybe it's time to deprecate those libraries. Personally I think Fred's time would be better spent working on other things like improving the gadget (more control over appearance), database (more supported DBs), mail (AUTH support), packer (ZIP support), etc. libraries instead. But I imagine folks that do write games would disagree...

+1

[Foz]

The Cyclone Physic engine allow that and this the 3D physic engine I'm currently re-write in PureBasic for my 3D game project.
It use Quaternion to retain the orientation but generate a 3X4 Transformation Matrix and a 4X4 OpenGL compliant Transformation Matrix.

Just my 2 cents, If a 3D Game engine not allow to access or generate a 3D Transformation Matrix it should not be used, period

Personally instead of implementing a 3D engine, I think I would prefer a built in wrapper - for DirectX or OpenGL (or both!), and then let the community write the required engines. People could port existing engines, others may prefer to create their own.

This way, a lack of feature X, Y or Z in <insert current 2d/3d engine here> would never simply arise. The only thing that would be raised would be calls that don't work or missing calls.

But what is done is done, and I highly suspect that their minds not only touched on this, but discarded the idea as being laughable that anyone would want to use it due to it being stupidly complex.

[Zach]

I would love to see steps taken that would more easily facilitate the community coming together, and skilled coders making an Engine we could all use, that takes a lot of the hard work / maths out of it.

Many of us just lack the brain power, time, skills, or all of the previously mentioned to even begin to understand how to come up with our own Engine.

[Brujah]

Wow. That were a lot of posts to read

We are still having a look at each new version of the compiler.
And I am very glad that development continues on the linux side of the compiler.

It seems utf-8 is finally working which will enable us to make a chinese version finally.
That is probably the most requested feature for our game.

mouse is working in fullscreen. and the keyboard seems to work at the same time. I am impressed!

So if the last few bugs get elimitated as well (its not many!) its still possible we continue the game with purebasic.
Its not that everything is just better with c++. We are still purebasic fans!