Re: [9fans] once is enough: VOID

[Brantley Coile <brantley@coraid.com>]

[-- Attachment #1: Type: text/plain, Size: 568 bytes --]

My favorite error message of all time was `unintelligent control
statment was output with diagnostic.' That was on the batch job
control on an CDC 18/20 ITOS system.  It was years later that I
realized that my program had stopped without reading the last record
in the sysin and a data card was the next thing read by the batch
system.  I had put two sentinel numbers as input instead of one.  It
just couldn't figure out what to do with 99999 as a job control card.

The message should have been, `Hey stupid!  What's 99999 supposed to
mean?'

  Brantley

[-- Attachment #2: Type: message/rfc822, Size: 3936 bytes --]

From: Bruce Ellis <bruce.ellis@gmail.com>
To: Fans of the OS Plan 9 from Bell Labs <9fans@cse.psu.edu>
Subject: Re: [9fans] once is enough: VOID
Date: Sun, 15 May 2005 04:21:29 +1000
Message-ID: <775b8d1905051411214f40e25c@mail.gmail.com>

One day I'll fix cc so it's diagnostics are more like those of cyntax.
A diagnostic should indicate the culprit, not where the parser
was when the semantic analysis was done.

Then again one day dumb-ass airline will show up with my luggage.

Consider yourself lucky.  I spent half a day trying to work out what
a gcc diagnostic meant and eventually cracked open a six pack
and wrote the entire file again from scratch - discarding the first
attempt.  Never did find out what the message meant.

brucee

On 5/14/05, Brantley Coile <brantley@coraid.com> wrote:
> I just spent 20 minutes figuring out what the above warning meant.
> The line number given in the warning is incorrect, so it took the 20
> minutes before I noticed that I had used the keyword `void' twice for
> a function type a few lines below the last line of the previous
> function, which was the line in the warning message.
> 
> Hope this saves someone else their 20 minutes.
> 
>  Brantley