Gommas?
Wilhelm B. Kloke
here is the context, as found on Google:
0. Introduction
... 0.3.12 Features removed. Certain features, such as proceduring, gommas
and formal bounds, have not been included in the revision. ...
burks.brighton.ac.uk/burks/language/other/a68rr/rr0.htm
What the heck is this?
It seems that the word may be inspired by "Finnegan's Wake":
... Count the hemi-, 9. semidemicolons! Screamer caps and invented gommas, quoites,
10. puntlost, forced to farce! The pipette will say anything at all for, 11. ...
www.trentu.ca/jjoyce/fw-374.htm
Does anybody know, what it meant in promming language syntax?
--
Dipl.-Math. Wilhelm Bernhard Kloke
Institut fuer Arbeitsphysiologie an der Universitaet Dortmund
Ardeystrasse 67, D-44139 Dortmund, Tel. 0231-1084-257
Dr A. N. Walker
Wilhelm B. Kloke <w...@yorikke.arb-phys.uni-dortmund.de> wrote:
>In the introduction to the Algol68 Revised Report, I found the word "gommas".
>... 0.3.12 Features removed. Certain features, such as proceduring, gommas
>and formal bounds, have not been included in the revision. ...
>What the heck is this?
Firstly, somehow the paras have been misnumbered on the
Burks site, and this is actually 0.3.11; secondly, "gommas" is
in the ordinary typeface on the site and is in bold in the RR,
representing therefore some [proto]notion[s] of the RR [or, in
this case, *not* of the RR].
>It seems that the word may be inspired by "Finnegan's Wake":
Much is inspired by FW, but I think this particular word
was simply a portmanteau. My copy of MR101 is AWOL, but IIRC it
was a go-on token [ie semicolon] or a comma token, possibly used
in some context such as serial clauses being units separated by
gommas. May have allowed completers [full stops followed by
labels] as well? I forget.
>Does anybody know, what it meant in promming language syntax?
Nothing in English!
--
Andy Walker, School of MathSci., Univ. of Nott'm, UK.
a...@maths.nott.ac.uk
Charles Lindsey
>In the introduction to the Algol68 Revised Report, I found the word "gommas".
>here is the context, as found on Google:
>0. Introduction
>... 0.3.12 Features removed. Certain features, such as proceduring, gommas
>and formal bounds, have not been included in the revision. ...
>burks.brighton.ac.uk/burks/language/other/a68rr/rr0.htm
In the original report, 5.4.1.d:
gomma : go on symbol ; comma symbol.
It allowed you to use ';'s in place of ','s in actual-parameters, thus
forcing them to be elaborated serially instead of collaterally.
--
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl
Email: c...@clw.cs.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
John English
>
> Firstly, somehow the paras have been misnumbered on the
> Burks site, and this is actually 0.3.11;
I'd just like to point out that the version on the BURKS site is just
a copy of the version that was available when the CDs were mastered in
summer 2001 at http://vestein.arb-phys.uni-dortmund.de/~wb/RR/, and
it's not my fault, honest!
-----------------------------------------------------------------
John English | mailto:j...@brighton.ac.uk
Senior Lecturer | http://www.it.bton.ac.uk/staff/je
Dept. of Computing | ** NON-PROFIT CD FOR CS STUDENTS **
University of Brighton | -- see http://burks.bton.ac.uk
-----------------------------------------------------------------
Wilhelm B. Kloke
John English <j...@brighton.ac.uk> wrote:
>"Dr A. N. Walker" wrote:
>>
>> Firstly, somehow the paras have been misnumbered on the
>> Burks site, and this is actually 0.3.11;
>
>I'd just like to point out that the version on the BURKS site is just
>a copy of the version that was available when the CDs were mastered in
>summer 2001 at http://vestein.arb-phys.uni-dortmund.de/~wb/RR/, and
>it's not my fault, honest!
I am going to look this up. IMHO I did not change the numbering at any time.
Perhaps MR101 is not identical to the final version in this respect.
The remark on the font used for the word "gomma" may be not applicable,
as gomma was a "notion" in the original report, but not in the RR anymore.
I just wanted to know what (feature) it was.
Thanks to Andy Walker and Charles Lindsey for their explanations.
BTW. Some reasons why proceduring and formal bounds had been removed,
have been discussed in Lindsey's language history article in the ACM
SIGPLAN paper, but not the gommas.
Wilhelm B. Kloke
John English <j...@brighton.ac.uk> wrote:
>"Dr A. N. Walker" wrote:
>>
>> Firstly, somehow the paras have been misnumbered on the
>> Burks site, and this is actually 0.3.11;
>
>I'd just like to point out that the version on the BURKS site is just
>a copy of the version that was available when the CDs were mastered in
>summer 2001 at http://vestein.arb-phys.uni-dortmund.de/~wb/RR/, and
>it's not my fault, honest!
Probably mine. I am going to look this up.
I did not change the numbering at any time.
So you may expect that this will be corrected in some days.
The remark on the font used for the word "gomma" may be not applicable,
as gomma was a "notion" in the original report, but not in the RR anymore.
Perhaps there are minor differences in the 3 (?) versions of RR published.
I just wanted to know what (feature) it was.
Thanks to Andy Walker and Charles Lindsey for their explanations.
BTW. Some reasons why proceduring and formal bounds had been removed,
have been discussed in Lindsey's language history article in the ACM
SIGPLAN paper, but not the gommas. Perhaps I cite some sentences from
this paper as a remark.
ngocnguy...@gmail.com
James Harris
news:31eb0f3c-a89a-4012...@googlegroups.com...
V�o 05:19:56 UTC-8 Th? s�u, ng�y 29 th�ng mu?i m?t nam 2002, Wilhelm B.
Kloke d� vi?t:
> "gommas".
> here is the context, as found on Google:
>
> 0. Introduction
> ... 0.3.12 Features removed. Certain features, such as proceduring, gommas
> and formal bounds, have not been included in the revision. ...
> burks.brighton.ac.uk/burks/language/other/a68rr/rr0.htm
>
> What the heck is this?
> It seems that the word may be inspired by "Finnegan's Wake":
>
> ... Count the hemi-, 9. semidemicolons! Screamer caps and invented gommas,
> quoites,
> 10. puntlost, forced to farce! The pipette will say anything at all for,
> 11. ...
> www.trentu.ca/jjoyce/fw-374.htm
>
> Does anybody know, what it meant in promming language syntax?
that it was intended to be a way of forcing an order on the evaluation of
arguments. The footnote of the report continues on the next page and says
"gomma : go on symbol ; comma symbol.
It allowed you to use ';'s in place of ','s in actual-parameters, thus
forcing
them to be elaborated serially instead of collaterally."
Some I can think of:
1. actual parameters, as above
2. booleans as with && and || in C
3. protection in arithmetic expressions (to avoid unwanted issues such as
divide by zero and maybe some combinations with NaNs)
4. performance or semantics in arithmetic expressions such as A * Func(...)
where we don't want Func to be called if A is zero.
I did at one point consider supporting some of the situations above by
defining the keyword "then" to only proceed if sensible, and the keyword
"also" to force evaluation of a later expression. For example, depending on
what the default was the programmer might have wanted to encode
A() && then B() ;only invoke B if A is true
A() && also B() ;invoke B always
A() || then B() ;only invoke B if A is false
A() || also B() ;invoke B always
A() * then B() ;only invoke B if A returns nonzero
A() * also B() ;invoke B always
Alternatively the words may have been better preceding the operation as in
A() then && B()
...
A() also * B()
However, I thought that such options were too complex and the gains were not
general enough.
It may be better for a language to define points that do not get reordered.
In the case of calling a function with arguments that are themselves the
results of non-pure functions (or some other volatile)
A(B(), C())
AIUI the C language, for example, doesn't define an order so doesn't specify
whether B or C would be evaluated first. Different compiles could generate
different code.
If an order is required on the evaluation of B and C it may be better for a
language design to specify that statements will appear to be executed in
order and to have the programmer write
b := B()
A(b, C())
or
c := C()
A(B(), c)
In that way, the order of evaluation would be defined. The 'gomma' feature
that was dropped from the aforementioned draft would only have allowed one
of those two sequences to have been specified by the language so although it
might have been useful on occasion IMHO they were wise to avoid the
complexity.
Another thought. Instead of defining that instructions would execute in
order the compiler could be allowed to reorder statements unless a special
statement appeared such as "then" as in
c := C()
then
A(B(), c)
James