On Fri, 28 Mar 2025 15:02:23 -0400, Rick Troth  wrote:
>
>I have always taken robustness to mean, for example, always send
>ASCII+CR+LF as a line of internet plain text, but *require* only
>ASCII+LF. If the CR is present, fine. It's going to get stripped out
>anyhow. (By whatever interpreter is crunching it. I'm not talking about
>files, necessarily.)
>
No.  The RFC requires it, so it's fair for Windows which (used to?)
need it to depend on its presence.

>Stuff like that is NOT chaos.
> 
If you don't like the standards, appeal for a change.

>I found that XLC (for normal C, not for C++) rejected a lot of my code
>where I had gotten lazy and used // for comments instead of proper /(*
>... */.
>In hat case, I appreciate the compile time errors. They only made my
>code cleaner.
>
Yes.  "Toleration is the mother of mediocrity."  Russian(?) proerb.

>But I have been bitten by IBM implementations where Big Blue were such
>sticklers for the rules that WIDELY accepted exceptions just did not
>fly. What a pain!
>
Blame the coders who mistake "WIDELY accepted" for a standard,
not Big Blue.

I have sometimes felt the best way to protest an unrealistically low
speed limit is to obey it at rush hour.

-- 
gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to