Larry Wall's comment on python...

Carl Banks imbosol at
Fri Sep 6 18:58:12 EDT 2002

mmaddox at wrote:
> I would argue that the "older" block-structured languages were considered
> a breakaway from the column-based languages - RPG2 and Fortran, for
> instance. Although I grant the Python structure and the column-based
> structures are not completely analogous, they BOTH burden the programmer
> with arbitrary rules to govern the physical layout of the code.

Python's forced indentation certainly doesn't burden me at all.  I
don't know about you, but I indent whether I have a block delimiters
or not.  Python doesn't force me to do anything I wouldn't do anyways.

What *is* burdensome is having to use block delimiters, AND to make
sure they are properly sychronized to the indentation.

> Although
> this layout may be considered more elegant from a readability standpoint,
> it's the enforcement of the layout that is irritating to the more
> Libertarian of us. It's a mindset difference similar to that between
> Pascal and C++.

Well, if you're complaining about your lack of freedom, then you're
complaining about your freedom to write unreadable and misleading
code.  It's a freedom I don't care for (usually).

> Having read Wall's comments on Slashdot, I think he is really just mildly
> complaining about ridiculously long page-widths in his editor, not really
> slashing Python seriously. He's just giving some props to his own baby at
> the expense of another. (He had to say something, right?) You must agree
> that Python can lead you to horizontal scrolling in your editor - the bane
> of user interface design.

No, not any more than C would.  Do you not indent in C?  If your C
code runs against the margin, is your solution to simply stop
indenting, and run all the blocks together?  (If it is, then you have
worse problems than being forced to indent.)

And, as a matter of fact, I never exceed 79 columns in any of my
Python code.  And I use eight-space tab stops.  And my editor doesn't
have horizontal scrolling; it wraps.


More information about the Python-list mailing list