[Info-vax] VMS-based InfoServer

[Phillip Helbig---undress to reply]

In article
<3b22ae2e-4779-4994-a8ca-b49f79e3e682 at 14g2000yqa.googlegroups.com>,
MetaEd <metaed at gmail.com> writes: 

> On Aug 4, 7:34=A0pm, VAXman-  @SendSpamHere.ORG wrote:
> > As far as I am concerned, DCL is text. =A0There's no need to convert it
> > to a usenet hostile and non-standard encoding. =A0Period.

Note that in the text above you quote from VAXman, "=A0" appears.  There
is absolutely no reason for this.  IIRC, it is a non-breaking space.
However, there is no reason to think it was intended as such in the
original text.  In fact, the convention of two spaces after final
punctuation is to accentuate the separation of sentences.  This is not
necessary of the end of the sentence is at the end of the line.  So
there is no reason not to break a line there.  Also, such text already
has the intended line breaks, so there is no reason to reformat it. 
Finally, it is TWO SPACES.  There isn't even any such thing as a
non-breaking space in 7-bit ASCII, which is what the original text was. 

> Last year, I studied how Google Groups codes messages for transmission
> over Usenet. I found that Google Groups uses pure 7-bit ASCII to code
> messages whenever possible. 

That's good.

> In other words, when there is no need to
> use a non-standard encoding, Google Groups does not use one.

That's good.

However, what is a "need"?  Surely two spaces after final punctuation is 
not such a need.  No way, José!

> The offending message, however, was not representable with pure 7-bit
> ASCII. There was a non-breaking space character in the text.

No.  There were two consecutive ASCII 20 (hex; 32 decimal) characters.  
That is NOT a non-breaking space.  If anything, it is the opposite (see 
above).

> Naturally, one wonders who put a non-breaking space character in a
> text which was otherwise representable with pure 7-bit ASCII. 

Indeed.

> Google
> uses NBSP to code space characters when you enter a row of spaces
> (more than one space). 

Why?

> This is a hint to the reader's terminal not to
> collapse or wrap within the row of spaces.

Can't this newfangled funky software differentiate between collapsing
and wrapping?  What have all those propellor-headed Java programmers
been getting paid for?  There is a simpler solution: if you don't want
to collapse it, DON'T COLLAPSE IT.  The original text has NOTHING to
indicate that it shouldn't be wrapped.  (Again, a) if anything it is OK
to wrap it but b) it is already correctly formatted so that there is no 
need to wrap it.)