Articles such "Pride Of Being Among The Greatest" are valid and truthfully, but are not enough. Amiga has a wonderful Operating System (OS), but there are few people effectively notice what is a truly multi-tasking environment, of how easy is to add a new device, without the need of pactches, drivers, and so on.
Amiga is still a step further and everyone has to contribute to assure this advantage should not decay, but grow and consolidate strongly.
Discussion we are doing now, does not point its attention to Amiga's hardware, but upon software programming style.
There are features that ALL new programs should have :
New programs does not have to try to still be OS2.0 compatible, and should forget completely the old Kickstart V1.3, at least until this compatibility would not very easy to obtain.
OS3.0 is a stone, well written. If you only think it is dated 1993 and that it is still better than various Windows 95, OS2 and so on, you should understand how much good work lies behind it. We cannot say the same thing for the OS2.0, which sadly introduces some very big loss, which are work-aroundable, but which weight on software development.
For my own part, I am writing a commercial program OS2.0+ compatible and personally it has cost me a lot in performance terms : I had to sacrifice some OS3.0 features for a "compatibility" which I do not think will worth the work.
Starting from the next version, anyway, I will require OS3.0+ to work properly.
This is one of the hardest thing to obtain : a robust code. When you are writing veryt complex programs, it is already a good deal not to make any mistake : lets work to obtain at least a code Enforcer's Hits free, since inside a multitasking environment such Amiga is, programs generating Enforcer's Hits surely will not work properly and my cause problems, menacing system integrity. So use Enforcer during all developing phases and be sure that final release version no errors will occur.
Take the OS to the maximum : do not limit yourself for no reason. All programs should support, for example, datatypes for image loading and BOOPSI gadgets for interface creation. And, please, do not say datatypes are slow : ther are not, and ZGIF is an emblematic example : if you find a slow datatype, then rewrite it in a faster way. I think graphic interfaces do not need MUI to work on all screens, MUI just uses some (few) features of BOOPSI : instead heavy gui-engines such as MUI, use ClassAct or just write BOOPSI gadgets : all Amiga users will greet well written gadgets and "ready to use".
Do not give limits to your programs, be OS compliant, reach for the stars, all you write must be the top of the standard. Use IFF format files where possible, parse the command line with ReadArgs(), read/write preferences in ENVARC :, use clean calls to graphic routines. In a word : do all you can to assure a program written today would be used (and perfectly working) next five, ten years.
Try to give all you can to the Amiga community. Write unique programs, more flexible, manageable and working than same tools for other platforms. Do not query if now AGA chipset is old : new machines will appear soon even more powerful and more futuristic. And AmigaOS will be always there to manage it all. Personally, I strongly desire my programs will work also on new machines, that a future user can use them, and it is the direction I am moving to. And this is the direction all of us should move to.
Do not ask "Where you want to go today ?", because present is not forever and technology is evolving quickly. The right question is "Where you want to be tomorrow ?"
Written By: Fabio Rotondo e-mail: fsoft@intercom.it
C.so Vercelli 9
28100 Novara
ITALY tel: (ITA) - (0)321 459676