home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.amiga.programmer
- Path: cix.compulink.co.uk!usenet
- From: jralph@cix.compulink.co.uk ("Jolyon Ralph")
- Subject: Re: 680X0 -> PPC translator?
- Message-ID: <DnwA2w.835@cix.compulink.co.uk>
- Organization: Compulink Information eXchange
- References: <19960306.4543C8.11FB8@an218.du.pipex.com>
- Date: Thu, 7 Mar 1996 11:18:32 GMT
- X-News-Software: Ameol32
-
- > Saying "That's tough" isn't much help when only a fraction (I have seen
- > estimates of 10-20%) of applications are going to work on a system
-
- Data in code sections isn't going to be as tough to sort out as code in
- data sections. It depends on the methodology of your static translator.
- Something that works through the hunk byte for byte will get fooled by
- data. Something that understands 680x0 code and AmigaDOS executable file
- formats and runs through the code in execution order (recursively) should
- be able to ignore data in code hunks (not that I'd like to write it, of
- course! I'm just saying that theoretically it can be done).
-
- > Have you examined your Photogenics code carefully to see just how >
- accurately the hunks have been organised? - not an accusation, just
- > a question.
-
- Yes, and you're right, there's not a great deal we could do to improve
- it. Hopefully we'll be porting Pgs to native PPC code eventually anyway.
-
- > A more flexible dynamic emulation system with stored translations is a
- > much better solution, and will get a hell of a lot more applications >
- > working.
-
- The way to get applications working is to make sure developers have PPC
- development systems on time so they can port REAL PPC apps.
-
- Jolyon
-