Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site well.UUCP Path: utzoo!watmath!clyde!burl!ulysses!ucbvax!hplabs!well!ewhac From: ewhac@well.UUCP (Leo L. Schwab) Newsgroups: net.micro.amiga Subject: Look what I found! Message-ID: <759@well.UUCP> Date: Wed, 12-Mar-86 02:12:19 EST Article-I.D.: well.759 Posted: Wed Mar 12 02:12:19 1986 Date-Received: Fri, 14-Mar-86 05:39:58 EST Reply-To: ewhac@well.UUCP (Leo L. Schwab) Distribution: net Organization: Whole Earth Lectronic Link, Sausalito CA Lines: 53 [ Deletion of this line by whatever means violates US Copyright laws :-] Guess what I stumbled onto! I was compiling my project, and had gotten to the linking stage, and had accidentally mistyped one of the parameters to Deep Thought (read Alink). Naturally, it screamed at me for being a twat, and showed me a correct argument list. Even though I knew what I did wrong, I looked at the diagnostic anyway, and noticed an argument I'd never seen before. It was called "faster". "Nah," I said to myself. "Can't be." So I tried it, just to prove it to myself. I typed out the command, giving "faster" as the last argument (something like "alink with link to ram:i faster"). Off it went, grinding away. Soon it reached the point where it lapses into catatonia. Except it didn't! It kept on grinding away, and soon thereafter, I got the "Linking complete - maximum code size" message. Deep Thought also usually thinks a bit more after that, but it didn't. It gave me a prompt back right away. "There's gotta be a catch," I thought. So I hacked up the WITH file to not include a code module, and ran it again, thinking that it was skimping on external reference checking. It wasn't. It reported the unresolved external references like it should.. And it didn't think about it. "Hmmm. Maybe it doesn't clean up after itself?" I thought. So I ran 'avail' around it. It was reclaiming all its memory. "I know, the final output file is bigger," I said, so I ran Deep Thought once with the "faster" argument, and once without, and compared file sizes. They were identical. "Good Grief!" I said. "It really works!" I now use the "faster" argument all the time. The thing I want to know is, why weren't we told about this new feature in the 1.1 release packet? And why isn't the "faster" feature on by default? Is there a bug in it I haven't stumbled across yet? If there is, is it a really insidious one? Both "faster" and "slower" versions of the linked program worked fine. What's the deal? Note: this pertains to Alink version 3.20. Another note: "faster" must be supplied in the command line; it can't be in a WITH file, or Alink will complain. Anyway, I thought you might be interested in this one..... Leo L. Schwab dual!unicom!schwab well!ewhac "I really got to think of a witty signature file...."