python was: transputers again was: The dissolution of Commodore 2915

Oh, good grief! Of course! I fought a battle one time about output of compilers and buttemblies not being bit-for-bit reproducible from the same sources. I won that one.

We weren't talking about pleasing formatting; we were talking about white space that generated code; this, like self-modifying code, should be declared as a deadly sin.

Sigh! Right and that is why. There are many, many authors and more coders that will do something a little bit different. This is because every body has their own style which works for him-her, and only him-her. Now when you get a conflicting style that has an effect on what code is generated, the computing biz has serious unforeseen, non-debuggable future messes.

And will be the whim of compiler coder, too. You, as a coder, should only be interested in what is getting output into the REL blocks.

That's not problem. The problem is that one compiler will generate one kind of instruction and another compiler would generate another kind of instruction has has the opposite effect. This can change from one version of a comiler to the next. What you need to be able to see is what it generated not how pretty the ASCII shows up on your screen. Granted this aspect is important, too.

We had one or two people who had their own really oddball ideas about how to format ASCII and would not change. If you had that kind of brain damage working on 1000 lines of code of a product that was a couple million, and this brain damage generated short instructions rather than the long, how long do you think it's going to take you to trace it back to the idiot who had a non-conformists' style of source formatting.

That's why no white space should ever generate code.

Dang! I locked my soapboxes up in a room and they appear to have slipped out.

BAH

BAH