AK-Datatypes In Development

[kolla]

What do you mean with 'again'?

convinced, again.

Anyways, just an idea - AmigaKit/AEON could possibly have their own operating system by buying pOS from ProDAD (if possible) and build their business on that.

[rxxic]

What does datatypes v46 offer over the official datatypes then? And how is compatibility with future “official” OS3 releases, or is this the dawn of yet another OS3 “spin”?

Our developers are working hard to make sure that it is compatible with legacy versions.   However with almost zero documentation, there is a lot of trial and error.  Additionally the guidelines that Commodore published back in 1993 have not been adhered to by third-party datatype developers since and this means that our new library needs to be forgiving enough with the many datatypes out there on Aminet.

For the future, we have some interesting plans.  If you recall the sound datatype project replacement introduced streaming for the first time to the Amiga.  The new project of the datatype library is a foundation for many progressive developments.  No one has worked on the datatype system in any meaningful way since the late 90's.  It is exciting to plan for the possibilities and follow the path that Commodore would have gone down with the system if they had survived.  Watch this space!

Will the akJPEG Datatype also read the EXIF metadata? Now that would be useful!

[rxxic]

What do you mean with 'again'?

convinced, again.

Anyways, just an idea - AmigaKit/AEON could possibly have their own operating system by buying pOS from ProDAD (if possible) and build their business on that.

If I was them, I'd buy the sources for Windows 3.0. It already runs on my Amiga2500 with Bridgeboard card. It could be ported to the 68000, imagine the potential! 

[Pgovotsos]

Will the new datatype.library still be compatible with the warp datatypes?

[bubbob42]

Dear Matthew,

2.  Our developer is a professional coder and doesn't invest months of hard work in Amiga software projects for free.  Are any of the OS developers paid for their work? 

while it’s absolutely not mandatory, especially since we are talking about a system with an old tradition of self-taught coding enthusiasts, most developers who invested years of hard work in OS3 have been indeed professional coders with degrees for decades or at least were for extended periods of their career. And not taking money was our own choice, as you know very well.

So what point exactly are you trying to make with that statement? Please note I’m not criticising your programmers for taking money, that’s absolutely legitimate.

Don’t get upset by Kolla, embrace the bear instead.
Best regards,

Marcus

[amigakit]

@bubub42

I have greatest respect for the good quality work you do, giving up your valuable free time in the process.  As professional coders by day, I trust you make this choice with your eyes wide open.   I hope that those that collect the significant proceeds of that work with ease take a moment to recognise the immense contribution that volunteers give.

Sadly we have witnessed commercial exploitation of goodwill far too often in the past in this community, unfortunately having witnessed it first hand.  We have always endeavoured to re-invest our business income back into the Amiga eco-system, including funding developers.

[Gulliver]

@amigakit

Saying sorry when you wronged someone is much better than a lengthy excuse that says nothing.

[Orphan264]

No, it is a replacement and enhancement for the system datatypes library.

 

I have now lost interest in this project. Good luck to you!

[wiser3]

...and with this project AmigaOS continues to splinter. Good intentions, bad result.

[amigakit]

Getting closer to the pre-release now.

A few problems showed up in testing using the v46.1 datatypes library:  in some datatype descriptors - the built-in function was indicating it could handle the data, yet the pattern was failing.  A solution was found by our developer after many days of debugging.  We are now working towards v46.2.

[kolla]

in some datatype descriptors - the built-in function was indicating it could handle the data, yet the pattern was failing.

Aha, it was probably interference from the tripolymer ionic polarity units that caused the pattern to fail. Classic!

[rxxic]

pffffuahahahahahaha!

[remotenemesis]

@amigakit will the new datatype library be made available for OS4 at some point?

[amigakit]

@remotenemesis

That is a distinct possibility so there will be a unified library for Classic and the X5000, X1000 and A1222 Plus platforms.

[amigakit]

The developers have been steadily working on the V46 datatype system files to bring them closer to release standard.   They have made some great progress recently.

An update to the datatypes.library to v46.2 now fixes some unwanted Enforcer hits and tests so far are positive.

Datatypes library has CPU versions for 68000, 020, 030, 040 and 060.

The AddDatatypes command also gets a version bump to v46.1 with some additional improvements added.

[Pgovotsos]

I asked before, but it must have gotten lost in the clutter.

Is it expected that this new system will be compatible with all classic datatypes, like the 3rd party warp datatypes, or will it require previous datatypes be updated to conform with it? I would assume so, and hope so, but it is never safe to assume.