oyster@uwmacc.UUCP (Vicarious Oyster) (12/31/86)
In article <1881@batcomputer.tn.cornell.edu> braner writes: >In response to postings about C compilers and about ACTION: > >Programs in ROM are not any faster than programs in RAM. They just >don't need to be loaded from disk. To reiterate: >> The main advantage of Action! is the fact that it's hardware. The editor >>and compiler (as well as "library" routines) are in the cartridge, so >>the compile/edit cycle is sped up considerably. Now, who said anything about faster programs? At any rate, you would answer that by: >In the case of a program you use >repeatedly in one session (e.g. editor, compiler, linker) you only >need to load it ONCE, into RAMDISK. There is a very limited number >of ROMs you can plug into your ST at once, but it is easy to boot >various disks. Which is fine, if you have that opinion. However, in a subsequent article, you complain about RAMdisk copy startup time, as well as trying to keep disk access down to a minimum, and the tedium of swapping disks in and out. Maybe cartridges aren't so bad after all, eh? And this doesn't even take into account RAM space, which at 1MB is still precious for some applications. I need a 512K ramdisk to hold my compiler, linker, editor, include files, etc. GEM and TOS take up more space (~120K?), which leaves less than 400K for my application (which is easy to eat up if you're setting up screen images for animation, for example). An extra 512K would be marvelous; just pop in a cartridge, boot the machine, and presto! your very own development system-- no waiting for ramdisk copies, no extra wear- and-tear on your heads, and 512K bytes extra RAM to work with. Sure, it'd have to be a fairly robust system, since a software update is a bit easier to distribute than another cartridge, but that, to me, is the only down side. >Their current use is for a sceme of copy- >protection (ugh: "we trust you SO LITTLE that we won't even SELL you >a disk"),... What's so bad about it? A ROM device is not nearly as likely to get trashed as is a diskette, so the need to make legitimate backup copies is negligible (at least for those who don't need to run from a hard disk). The bit about trust is merely paranoid cynicism. >...and also for turnkey diskless systems (e.g. an ST set up >as a dumb terminal). That's a good use, although every turnkey system I've seen has been quite a bit more than just "a dumb terminal." To summarize, cartridges are not the universal cure, but they're not as bad as you make them sound; after all, if somebody at Atari didn't think they were important, why go to the expense and bother of sticking a cartridge slot on each ST? (A comment on that from an Atari employee might be illuminating...) -- - Joel Plutchak uucp: {allegra,ihnp4,seismo}!uwvax!uwmacc!oyster ARPA: oyster@unix.macc.wisc.edu BITNET: plutchak@wiscmacc
braner@batcomputer.tn.cornell.edu (braner) (01/06/87)
[] There ARE some things I would like to have in ROM. But I only need to load the compiler, etc to the RAMdisk ONCE, then the compile cycle is fast for however many times I need to go through it. As for running out of 1 Meg: If the 128K of ROM is what you lack, you're using the wrong compiler (Megamax compiler+linker+libs, + microemacs, is about 250K). And soon we'll have 2 Meg and more! Atari has had a long tradition of ROM cartridge ports, they do it for nostalgia. But it's obsolete. Lets have a real bus expansion connector (a la Amiga) instead! Whoever sells software in ROM rather than on disk, as a form of "copy protection", is the one guilty of "paranoid cynicism". - Moshe Braner