a case for architectural change in oxygen?

[Charles Pegge]

OxygenBasic itself is only a programming language, and this will continue to be my main focus. But here, Kent is referring to a major work of literature, so to speak.

We have created quite a comprehensive collection of examples already, encompassing maths, data-processing, multi-threading, audio and graphics, which I hope will encourage further development, by extending and combining these elements.

With some of the features, I've included in Oxygen, it should be possible to push through the scale-barrier, often encountered with other Basics.

Charles

PS: I'm in favour of re-inventing the wheel
Some of us have seen this video before:

http://www.youtube.com/watch?v=JVB1ayT6Fdc

[Aurel]

Kent...
With full respect to you i think that you are not sure what you wish and what
kind of shape this your game engine will look like.
I am also not sure how i would like to see GUI part of my coding in oxygen.
Some points i have found in AutoIt code and some in Gui4Cli because sources for
both are open.
Level of prototyping in oxygen is excellent if you ask me and i don't know what would be
next level....

[kryton9]

No doubt OxygenBasic is very exciting and it has come a long way. But as John said, it would be neat to have some sort of roadmap.

@Charles, yes I am also for reinventing the wheel and my post was wrong in using that phrasing. Maybe it should be re-gathering the same parts and putting the same thing together over and over again.

@Aurel, you are also right. It is very hard to plan for something so complex while still developing the other parts.

As I said, I will eventually come back to oxygen once I figure my engine out and get it working and have some code that can be studied and ported over. Things are changing in many fronts.
SDL 2.0 Beta which I think is SDLHG now, supports multiple monitors and windows now. There is Nokia's QT which is quite interesting as it compiles to Native Windows, Linux and Mac. Oxygen written in QT would be interesting as it has a great framework from which to launch from and QT does not have a BASIC version, so that is a good fit. I don't know how all that works, my guess is Oxygen would tap into QT's parser and add BASIC syntax?

Of course there has always been free pascal and a new compiler called clang. 2 to 3 times faster in compiling than gcc and captures many errors that other compilers miss. In fact, many major projects are advertising passed through clang. For example, in one of the videos I watched, they said code that came out squeaky clean no errors in gcc, had around 150 errors in clang identified. clang is now becoming the Mac compiler of choice. clang is also connected to LLVM in ways I don't understand at the moment.

Anyways just some thoughts while I search for my engine development path.

[efgee]

There is Nokia's QT which is quite interesting as it compiles to Native Windows, Linux and Mac. Oxygen written in QT would be interesting as it has a great framework from which to launch from and QT does not have a BASIC version, so that is a good fit.

There is the KBasic compiler that uses QT (from the same developer that started work on Objective-Basic for iMacs).

... and a new compiler called clang. 2 to 3 times faster in compiling than gcc and captures many errors that other compilers miss. In fact, many major projects are advertising passed through clang. For example, in one of the videos I watched, they said code that came out squeaky clean no errors in gcc, had around 150 errors in clang identified. clang is now becoming the Mac compiler of choice. clang is also connected to LLVM in ways I don't understand at the moment.

With LLVM came new/very good competition for GCC and the GCC folks finally woke-up and started adding better optimization to GCC - which is a good thing IMHO.

LLVM is the core and back-end of a compiler infrastructure. Apple helped/pushed LLVM because they didn't want to rely solely on GCC.
Clang is the C front-end of LLVM and there are other front-ends for C++ and Object-C.

Actually LLVM is the core of Apples XCode: https://developer.apple.com/technologies/tools/

LLVM is available for many different operating systems.

[efgee]

Where does Oxygen fit in all of this?

Don't know.

My safest bet is that there is only one steering wheel and Charles has it in his hands.
Wherever he wants to go Oxygen will go.


A lot of you stick to Win32 and reinvent the wheel, but I stopped working on the GUI library I was working on (in Oxygen) because Win32 is dieing.
Working on GUI stuff for a new compiler utilizing old Win32 interfaces with all it's quirks/unpleasant surprises is not something I can afford doing in my spare time.
If I would be retired it would be different (more time...) but I have still many years to work, so time is precious.

Rather spend my time on things I can envision they are still around in 10 years or so and is not a dieing mule like Win32.
And no I will not start doing Metro programming because I think it's the wrong way to go for PC-workstations.
I'm not interested in gaming and for pure consumption (not creation) I have an Android tablet.

From a programming perspective it will be OSX or Linux for me - didn't make up my mind yet.
(still looking at available programming tools and technology)

The only place I will touch Windows in the future is my HTPC (because of Blu-Ray) and at work - where we still use WinXP  

Edit: typos...

[Peter]

Hi efgee,

because Win32 is dying.

do not fall into panic now.  

[efgee]

I'm not panicking, I'm just saying how it is.
Win32 is a dieing mule.

If you want to use your time to work on Win32 stuff - it's fine - have fun.
There are still people that program with QuickBasic on FreeDos on a regular basis and are happy.
Null problemo.

I just said that in my case I will not use my time for a dieing mule.

[Peter]

I think that you are right.
The whole windows is dieing, starts with Window8.  I say RIP and drive bicycle.

[Aurel]

Then... efgee Oxygen basic is not for you.
Windows dying ...yes because MacOS & Linux use only cca 5% people on planet

[efgee]

Then... efgee Oxygen basic is not for you.

Well don't say that - it hurts my feelings 

No seriously:
If Charles creates a C emitter OxygenBasic could be used for other OS as well.

Windows dying ...yes because MacOS & Linux use only cca 5% people on planet

Aurel,
sorry mate but you misunderstood:

I said Win32 is a dieing mule not Windows.
Windows and Win32 are not the same!

Win32 is an API - Windows is a OS.

Windows will be pretty much alive even after Windows 8.
(WindowsME anyone?)

But as I said before I'm not interested in METRO the new Windows 8 "holy grail".


You are right about Linux and OSX in that they have only a small userbase, but I don't care about that.
From a programming point of view I look at technology and programming tools.

[Charles Pegge]

A C emitter is certainly the next milestone for Oxygen, but in terms of library development it's more of an organic process.

I am not overly concerned about repetition. If a repeating code pattern becomes evident, it will be assimilated into a library at some point.

I think Win32 will always be present, even if most programs do not use it directly - much of Win32 forms the base of the pyramid.

Charles

[Aurel]

efgee
In your observation there are to many contra-points and sounds like i don't
know about what i talk...man...
maybe you think that i'm stupid or something...
your feelings are not my theritory
and from your post i see that you don't know many things
Metro style is HTML5 style of programming if you don't know that...
And you say win32 api is not of my interest ...man ...gee
do you know where you living...

[Peter]

do you know where you living...

LOL 

[efgee]

efgee
In your observation there are to many contra-points and sounds like i don't
know about what i talk...man...
maybe you think that i'm stupid or something...
your feelings are not my theritory
and from your post i see that you don't know many things
Metro style is HTML5 style of programming if you don't know that...
And you say win32 api is not of my interest ...man ...gee
do you know where you living...

@Aurel
I just tried to emphasize the fact that I never said Windows is dieing.
Sorry for not being clear enough and because of that you misunderstood many things I said.
Maybe the language barrier adds to the misunderstandings.

In any case: There is no reason for you to get personal.

[efgee]

A C emitter is certainly the next milestone for Oxygen, but in terms of library development it's more of an organic process.

I am not overly concerned about repetition. If a repeating code pattern becomes evident, it will be assimilated into a library at some point.

I think Win32 will always be present, even if most programs do not use it directly - much of Win32 forms the base of the pyramid.

Charles

You might be right in regards to Win32.
It is my understanding though that Win32 applications will be treated and look like an alien on Windows 8.

Take care