home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!gatech!nscf!lakes!rock
- From: rock@lakes.trenton.sc.us (Rockerboy)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: Going to the metal
- Message-ID: <BLc7wB3w165w@lakes.trenton.sc.us>
- Date: Sun, 10 Jan 93 13:44:46 EST
- References: <1993Jan8.203137.13728@freenet.carleton.ca>
- Reply-To: rock@lakes.trenton.sc.us (Rockerboy)
- Organization: Lakes Public Access
- Lines: 59
-
- aa302@Freenet.carleton.ca (Russell McOrmond) writes:
-
- > You missed the point - Library IMPLEMENTATION diversity is a good thing,
- > but interfacing diversity is destructive. The requestor libraries
- > are a good example. There are usefull libraries which patch in and
- > replace asl.library, and then there are IMHO useless libraries that
- > require that the application use a different interface in order to get this
- > different environment.
-
- What is your point? Of course there will be uselss versions, just as
- there are countless other uselss things. That still doesn't make the
- useful versions unworkable.
-
- > If you can't see the obvious magnitude in difference between
- > functionality diversity and interfacing diversity, go back to grade 2
- > programming.
-
- How arrogant of you to assume that you have any business telling me what
- is and is not right? Who died and made you programming guru? I'd
- suggest you go back to kindergarten, or first grade, or wherevere, in
- your country, they teach children a _truly_ important concept: manners.
- The 'magnitude of difference' is indeed obvious. You are, hoever,
- conveniently ignoring the fact that there would be useful versions as
- well as useless ones.
-
- > The code that these compilers generate all run on the same computer -
- > if you are proposing different implementations with a standard
- > interface, then we agree - Otherwise what you are suggesting is nothing
- > bug destructive.
-
- That is _precisely_ what I am proposing, you silly goose!
- Hardware.libraries to allow pseudo hardware registers, with machine
- specific versions to interpret standard commands and tailor it to
- specific hardware! It might not be substantially faster than using the
- OS, but there would be a pronounced flattening of the learning curve for
- it. Now where, pray tell, is the harm in that?
-
- > The point of a .library is to provide a common interface to
- > multiple implementations of some code. When there is a
- > 1-1 ratio of application and library compatability, the whole purpose
- > of writing the library is nullified and the .library code should just
- > have been put into the application itself.
-
- You cannot prevent people from implementing their custom version of a
- standard library, but, as I have tried to point out, such implementations
- would have little impact. If they were different, people would likely
- react negatively. We're not discussing something like a requester, where
- you say, 'Oh, that's neat', but rather a library that will have no visual
- factor. It would simply be a simplified OS, on designed in a chip
- oriented manner, and with certain commands such as SuspendOS() and
- RestoreOS(). Ideally, said library would posess duplicated of several
- other library functions, so that, by using one, simple library, game
- designers can power down the system, access disks, allocate memory, and
- then bring the system back up when he's done. Coupled with routines to
- access hardwrae registers oin a low a manner possible, this would be a
- most effective compromise from game designers. Why is it that some
- people feel that need to kick and scream so hard against it? Perhaps, as
- I suspect, you simply feel that games and demos have no value, and should
- not be allowed to exist?
-