11 May
2005
11 May
'05
4:16 p.m.
No, I can not remember those quotes either.
Must have been something else, Paul.
For what I remember, BEATHE looked a bit like PASCAL; all the
statements ended with a semicolon (;).
Admitted, memories are vague after 30 years not using BEATHE :~)
but I would certainly remember something as idiot as a quote for
every letter of a reserved word of the language.
- Henk, PA8PDP.
On Wed, 11 May 2005, Paul Koning wrote:
I only used BEATHE briefly; my recollection is that it
used quotes to
mark ALGOL keywords rather than reserved words, i.e.,
'i'f i=1 't'h'e'n ...
instead of
if i=1 then ...
which added no real value but did use up a lot more cards and
finger-power...
You have GOT to be kidding. Who's the genius that came up with that
scheme?
--
Sellam Ismail Vintage Computer
Festival
11 May
11 May
4:51 p.m.
Gooijen H wrote:
No, I can not remember those quotes either.
Must have been something else, Paul.
For what I remember, BEATHE looked a bit like PASCAL; all the
statements ended with a semicolon (;).
Admitted, memories are vague after 30 years not using BEATHE :~)
but I would certainly remember something as idiot as a quote for
every letter of a reserved word of the language.
I have not used algol but I looked at the tiny algol for the PDP-8.
I think that had quotes around the entire reserved word.
Ben alias woodelf
4:53 p.m.
>>>> "Gooijen" == Gooijen H
<Gooijen> writes:
Gooijen> No, I can not remember those quotes either. Must have been
Gooijen> something else, Paul. For what I remember, BEATHE looked a
Gooijen> bit like PASCAL; all the statements ended with a semicolon
Gooijen> (;). Admitted, memories are vague after 30 years not using
Gooijen> BEATHE :~) but I would certainly remember something as idiot
Gooijen> as a quote for every letter of a reserved word of the
Gooijen> language.
No -- neither in Pascal nor in Algol do statements end in semicolon.
Instead, statements are SEPARATED by semicolons. An important
difference, which PL/1 and C both got wrong, in different ways.
I may have the quoting thing mixed up. It probably was quotes around
the keywords, i.e.,
'begin' 'int' i; 'if' j=1 'then' .....
The reason for this notation is that traditional Algol does not have
any notion of reserved words (the spec has two different typefaces,
one for keywords, one for identifiers). The THE system Algol used
underlining for keywords (using Flexowriter punch-tape machines where
underline was a non-spacing character). So BEATHE was created to
preserve that "no reserved words" property. It may have done other
things as well, again for the purpose of being maximally compatible
with the older language.
paul
11:58 p.m.
No -- neither in Pascal nor in Algol do statements end
in semicolon.
Instead, statements are SEPARATED by semicolons. An important
difference, which PL/1 and C both got wrong, in different ways.
How did C get it wrong? I think Pascal got it wrong and C got it
right, because C did it much more the way humans tend to think of it
(and that, after all, is really what a programming language is all
about: communicating from a human to a computer.)
Of course, in C, semicolons don't terminate statements. They terminate
*some* statements - specifically, expression statements. (Well, that's
the big one. There are also null statements, the do-while,
break/continue/return, and (if you count them) declarations, and
semicolons occur embedded in a few places like the for statement. And
I may have forgotten something.)
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
12 May
12 May
9:30 a.m.
>>>> "der" == der Mouse <mouse
at rodents.montreal.qc.ca> writes:
> No -- neither in Pascal nor in Algol do statements
end in
> semicolon. Instead, statements are SEPARATED by semicolons. An
> important difference, which PL/1 and C both got wrong, in
> different ways.
der> How did C get it wrong? I think Pascal got it wrong and C got
der> it right, because C did it much more the way humans tend to
der> think of it (and that, after all, is really what a programming
der> language is all about: communicating from a human to a
der> computer.)
der> Of course, in C, semicolons don't terminate statements. They
der> terminate *some* statements - specifically, expression
der> statements.
Exactly. C has a random grammar -- the random rules for semicolons is
just one example. It's especially bad with K&R function headers -- I
dropped those like a hot potato when I learned about ANSI headers,
where I actually had a chance to get the syntax right on the first
attempt.
The Algol/Pascal rule is very simple: semicolons between statements.
Since Algol at least has null statements, if you throw in an extra one
it doesn't hurt, but that's the rule and it works very well.
Thus the program that Cameron quoted as
begin;
writeln('You had best not be dissing Pascal, boy');
end.
should be written without either of the semicolons.
paul
9:38 a.m.
The Algol/Pascal rule is very simple: semicolons
between statements.
Since Algol at least has null statements, if you throw in an extra one
it doesn't hurt, but that's the rule and it works very well.
Thus the program that Cameron quoted as
begin;
writeln('You had best not be dissing Pascal, boy');
end.
should be written without either of the semicolons.
Why? (Unless you're talking about Algol.)
--
---------------------------------- personal: http://www.armory.com/~spectre/ --
Cameron Kaiser, Floodgap Systems Ltd * So. Calif., USA * ckaiser at floodgap.com
-- Flat text is just *never* what you want. -- stephen p spackman -------------
10:28 a.m.
>>>> "Cameron" == Cameron Kaiser
<spectre at floodgap.com> writes:
> The Algol/Pascal rule is very simple: semicolons
between
> statements. Since Algol at least has null statements, if you
> throw in an extra one it doesn't hurt, but that's the rule and it
> works very well.
>
> Thus the program that Cameron quoted as begin; writeln('You had
> best not be dissing Pascal, boy'); end.
>
> should be written without either of the semicolons.
Cameron> Why? (Unless you're talking about Algol.)
Because Pascal uses the same rule of semicolons as Algol does.
paul
10:38 a.m.
> > The Algol/Pascal rule is very simple:
semicolons between
> > statements. Since Algol at least has null statements, if you
> > throw in an extra one it doesn't hurt, but that's the rule and it
> > works very well.
> >
> > Thus the program that Cameron quoted as begin; writeln('You had
> > best not be dissing Pascal, boy'); end.
> >
> > should be written without either of the semicolons.
> Why? (Unless you're talking about Algol.)
Because Pascal uses the same rule of semicolons as
Algol does.
I was all set to disagree with you, so I ran it through Turbo Pascal without
the semicolons and ... it worked.
However, when you say semicolons between statements, that's what I did. So
why could I have dropped the semicolons, unless I have the wrong idea about
what Pascal (and presumably Algol) considers a statement?
--
---------------------------------- personal: http://www.armory.com/~spectre/ --
Cameron Kaiser, Floodgap Systems Ltd * So. Calif., USA * ckaiser at floodgap.com
-- FORTUNE: The moon is in Venus' house. This will make no difference. --------
10:45 a.m.
>>>> "Cameron" == Cameron Kaiser
<spectre at floodgap.com> writes:
> > > should be written without either of the
semicolons.
> > Why? (Unless you're talking about Algol.)
> Because Pascal uses the same rule of semicolons as
Algol does.
Cameron> I was all set to disagree with you, so I ran it through
Cameron> Turbo Pascal without the semicolons and ... it worked.
Cameron> However, when you say semicolons between statements, that's
Cameron> what I did. So why could I have dropped the semicolons,
Cameron> unless I have the wrong idea about what Pascal (and
Cameron> presumably Algol) considers a statement?
"begin" and "end" are not statements -- they are keywords. Think of
them as parentheses. (In fact, in Algol 68 they are synonymous with
parentheses.) Similarly, you don't put semicolons into the middle of
an if/then/else either.
paul
10:54 a.m.
> However, when you say semicolons between
statements, that's
> what I did. So why could I have dropped the semicolons,
> unless I have the wrong idea about what Pascal (and
> presumably Algol) considers a statement?
"begin" and "end" are not
statements -- they are keywords. Think of
them as parentheses. (In fact, in Algol 68 they are synonymous with
parentheses.) Similarly, you don't put semicolons into the middle of
an if/then/else either.
Ahhh, okay. That explains it. Thanks for the clarification.
--
---------------------------------- personal: http://www.armory.com/~spectre/ --
Cameron Kaiser, Floodgap Systems Ltd * So. Calif., USA * ckaiser at floodgap.com
-- NOBODY EXPECTS THE SPANISH INQUISITION! ------------------------------------
13 May
13 May
1:58 a.m.
New subject: Algol, Pascal sytax
On Thu, 12 May 2005, Cameron Kaiser wrote:
I don't doubt that this assertion is true, but the behavior of one
compiler doesn't have any bearing on the language definition.
Because Pascal
uses the same rule of semicolons as Algol does.
I was all set to disagree with you, so I ran it through Turbo Pascal without
the semicolons and ... it worked.
7011
days inactive
7013
days old
10 comments
6 participants
participants (6)
-
bfranchuk＠jetnet.ab.ca -
GOOI＠oce.nl -
mouse＠Rodents.Montreal.QC.CA -
pkoning＠equallogic.com -
spectre＠floodgap.com -
tomj＠wps.com