Managing Google Groups headaches

rusi rustompmody at gmail.com
Sat Dec 7 08:43:05 CET 2013


On Saturday, December 7, 2013 7:54:50 AM UTC+5:30, Ned Batchelder wrote:
> On 12/6/13 8:03 AM, rusi wrote:

> > Leaving aside whose fault this is (very likely buggy google groups),
> > this mojibaking cannot happen if the assumption "All text is ASCII"
> > were to uniformly hold.
> > Of course with unicode also this can be made to not happen, but that
> > is fragile and error-prone.  And that is because ASCII (not extended)
> > is ONE thing in a way that unicode is hopelessly a motley inconsistent
> > variety.

> You seem to be suggesting that we should stick to ASCII.  There are of 
> course languages that need more than just the Latin alphabet.  How would 
> you suggest we support them?  Or maybe I don't understand?

Heh! Yes I guess that can be read into what I was saying.

Practically: I dont see that as an option or that the question of
going back to ASCII even arises.

I was talking more philosophically/historically.

Up until the time of Unix a file for example was a structured
heavy-duty concept motivated by entirely technological considerations:
http://en.wikipedia.org/wiki/Data_set_%28IBM_mainframe%29

By simplifying that into the modern concept of file -- just a stream
of bytes -- and allowing the puns:

  byte string
= char list
= text

some elegant systems could be made with people having 'beautiful thoughts:'

Everything that could be stored anywhere -- core or disk -- being bytes
one could go to the next stage and pass around these bytes between
processes. And so we get the elegant --  pipeline -- beauty of Unix
scripts.

Of course there was a catch (Isn't there always?):

Things that did not fit in with this philosophy -- eg clicks of a mouse,
bits on display -- were modelled badly or not at all.

Not-at-all: CLI
Badly: Monstrosity called X

And this explains some of the cultural kinks of our field:

Unix guys invariably think of CLIs as natural and obvious whereas GUIs
are just wasteful eye-candy.

[Yours truly is one of those old geezers who does not know how to
write a GUI to save his life. Almost normal in the Unix world except
that he's not proud of it]

Windows/Mac people do not suffer these delusions but then they dont think of 
programming as natural or obvious at all.

Ive often been amused at windows folk: They dont think of Word as a program.
Rather docs are things that magically open when clicked :-)

Brings me to the point I was trying to make (got side-tracked by
the failure of a character to roundtrip between me and Roy  -- Im none the 
wiser why)

The ASCII = Text = Unicode (non)equation is a relatively minor point.

The more central point is that humans use and need more than just
words to communicate.  By straitjacketing communication into the thin
channel of text we are severely impoverishing ourselves.

We communicate with systems with programs that are unstructured
text-files even though programs are conceptually highly structured entities.

Likewise we communicate with each other by this obscenely obsolete
textual mode that I am using right now when rich text formats have been
available for decades.

Some of my more detailed writings on this:

http://blog.languager.org/2013/09/poorest-computer-users-are-programmers.html

http://blog.languager.org/2012/10/html-is-why-mess-in-programming-syntax.html



More information about the Python-list mailing list