[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
The names have been changed to protect the innocent...
As if ANY of us can claim innocence... :-)
Well I have completed the nomenclature changes. You can find
them in http://nonabelian.com/754/754r.{sxw,pdf}.
It was a bigger job than some people estimated but not as bad
as I thought. While there were a lot of terms to look at it
was only necessary to change about 10% of them.
I took notes as I made the changes & I include them here.
One note for each change.
Well, I think it reads better anyway.
See you at the style review on Tuesday.
Enjoy,
Dan
This is the nomenclature I used:
FLOATING-POINT DATUM - The most general case.
NaN - As a NaN has only been defined for floating-point
numbers, FLOATING-POINT NaN is unneeded. If I
come across an instance of some other kind of
NaN I will qualify it, eg: INTEGER NaN.
FLOATING-POINT NUMBER - The general numeric case.
INFINITY - Also does not need FLOATING-POINT.
FINITE NUMBER - This could be FINITE FLOATING-
POINT NUMBER but I'll go with FINITE
NUMBER if there is no danger of
confusion.
ZERO - Fine as is.
NON-ZERO NUMBER - Again, will only use
NON-ZERO FINITE FLOATING-POINT
NUMBER if there is danger of
confusion. NON-ZERO NUMBERS are
taken to be a subset of FINITE
NUMBERS.
NORMAL NUMBER - Fine as is.
SUBNORMAL NUMBER - This too.
We should consider making a definition entry for each.
In the following list of changes, I went through the document
once for each term (1406 in all), read the sentance involved,
and made changes needed to make that sentance conform to the new
nomenclature. Thus, expecially for early terms like 'number',
the term changed was not always the term searched for but, often
as not, something else in the same sentance or nearby. As you
might expect, this dropped off rapidly as the task unfolded & in
latter searches there were few, if any, terms changed other than
that searched for. The task also got easier. Thus, the 149
instances of 'number' took me something like 5 hours to complete
whereas later on the 500 instances of 'format' took only 2 hours.
Some things, like the 195 instances of 'NaN' required no changes
at all.
One result is that all instances of entity/entities have been
removed.
For all this, there were only 137 changes in all. A lot of work
for so little changed. I glanced through it at the end & most
of what I looked at seemed to read much more natural. I should
hope so.
The page numbers are from the SXW source &, as always, your
milage may vary. Go by the section numbers & you should get
close enough to see the change indicated.
The changes:
Number (149):
-------------
page 10, 1.2 Inclusions:
floating-point number formats -> floating-point data formats
floating-point numbers -> floating-point data
page 13, 3.2.4 canonical encoding:
representable value -> floating-point number
numbers -> finite numbers
page 14, 3.2.13: represented number -> floating-point number
page 14, 3.2.15:
number -> floating-point number
Suggestion: page 14, 3.2.16: the definition of a floating-point number
changes & the definition of a floaing-point datum should
be added. For that matter, we should put all the terms
in the nomenclature into definitions.
page 15: 3.2.31:
number -> floating-point number
page 18, 5.2, after table 2:
representable entity -> representable floating-point datum
representable entity -> floating-point datum
number represented -> floating-point number represented
page 18, 5.3:
representable entities -> floating-point data
entities representable -> floating-point data representable
those floating-point representations -> floating-point data
This next one reads funny:
page 20, 5.3:
following entities shall -> following floating-point data shall
numbers -> floating-point numbers
numbers -> floating-point numbers
numbers -> floating-point numbers
finite representable numbers -> finite number
page 20, Table 2:
representable numbers -> floating-point data
page 21, 5.4:
representable non-zero number -> non-zero number
number -> floating-point number
numbers in the binary -> floating-point data in the binary
page 22, 5.5:
representable number -> floating-point number
floating-point representations ->
floating-point number representations
number -> floating-point number
number's -> floating-point number's
number -> floating-point number
number -> floating-point number
floating-point representations -> floating-point numbers
non-zero number -> non-zero number's
one-digit number -> one-digit floating-point number
p-digit number -> p-digit floating-point number
n-digit number -> n-digit floating-point number
number -> floating-point number
numbers -> floating-point data
number -> floating-point number
page 26, after Table 6:
valid numbers -> valid decimal numbers
page 26, 5.6, Table 7:
representable numbers -> floating-point data
page 28, 6.2.1:
representable number -> floating-point number
representable numbers -> floating-point numbers
representable number -> floating-point number
representable numbers -> floating-point numbers
page 28, 6.2.2:
representable number -> floating-point number
representable number -> floating-point number
representable number -> floating-point number
page 30, 7.2:
representable number -> floating-point number
number's -> floating-point number's
page 30, 7.3.1:
representable number -> floating-point number
page 32, starting at minNum:
number -> floating-point number
number -> floating-point number
number -> floating-point number
number -> floating-point number
page 32, 7.3.2:
number -> floating-point number
number -> floating-point number
page 33, 7.3.3:
numbers -> finite numbers
numbers -> floating-point numbers
page 39, isCanonical:
number -> finite number
numbers -> finite numbers
page 44, 7.10
number -> floating-point number
number -> floating-point number
page 44, 7.11
compre two numbers ->
compare one floating-point datum to another
floating-point numbers -> floating-point data
page 46, 7.12:
numbers -> floating-point numbers
numbers -> floating-point numbers
binary internal number -> internal binary floating-point number
page 47, 7.12.1:
internal format numbers -> internal floating-point numbers
internal format numbers -> internal floating-point numbers
page 48, 7.12.3:
decimal numbers -> decimal floating-point numbers
number -> decimal number
target format -> target floating-point format
page 49, 7.12.3:
numbers -> decimal numbers
page 54, 9.4:
number -> finite number
page 54, 9.5:
number -> finite number
page 62, annex D:
representable numbers -> floating-point numbers
representable number -> floating-point number
representable numbers -> floating-point numbers
Value (112):
------------
page 13, 3.2.5:
floating-point representations ->
representations of floating-point numbers
page 13, 3.2.7:
format value -> floating-point number
format value -> floating-point number
page 15, 3.2.28:
a floating-point representation ->
the representation of a floating-point number
page 20, 5.3:
nonzero values -> non-zero floating-point numbers
page 21, 5.4:
corresponding representable entity value ->
value of the corresponding floating-point number
corresponding representable entity value ->
value of the corresponding floating-point number
page 32, 7.3.2:
value -> numerical value
value -> numerical value
page 43, 7.9:
floating-point representation -> floating-point number
floating-point representation -> floating-point number
floating-point representations -> floating-point numbers
Entity (8):
-----------
page 14, 3.2.22: symbolic entity -> symbol
page 21, 5.4:
floating-point representation ->
representation of the floating-point number
representable entity ->
value of the floating-point number represented
page 24, 5.5:
floating-point representation ->
representation of the floating-point number
representable entity ->
value of the floating-point number represented
page 30, 7.2:
representable entity -> value
representable entities -> floating-point values
Entities (11):
--------------
page 14, 3.2.23:
numerical representable entities -> floating-point numbers
page 18, 5.2, table 1:
representable entities -> floating-point data
floating-point representations ->
representations of floating-point data
page 20, 5.3:
entities -> floating-point data
nonzero representable entities -> non-zero numbers
page 26, 5.6:
the representable entities -> floating-point data
page 44, 7.10:
same entities -> same floating-point datum
same representation -> same floating-point datum
(There are now no more entity/entities in the draft.)
Represented (10):
-----------------
(Remaining cases are all verb forms.)
Representable (49):
-------------------
page 14, 3.2.25:
representable nonzero finite floating-point number ->
non-zero finite number
page 18, 5.2:
representable floating-point datum ->
representation of a floating-point datum
page 32, nextAfter:
representable neighbor of ->
floating-point number that neighbors
page 62, annex D:
representable -> a floating-point number
representable -> a floating-point number
representable -> a floating-point number
representable -> a floating-point number
representable -> a floating-point number
Representation (59):
--------------------
page 13, 3.2.1:
floating-point representations ->
representations of floating-point data
page 18, 5.2, table 1:
floating-point representations -> floating-point data
page 18, 5.2, after table 1:
floating-point representations ->
representations of floating-point data
floating-point representations of a format ->
floating-point data in a format
floating-point representation -> floating-point datum
floating-point representations -> floating-point data
page 22, 5.5:
floating-point representations -> representations
page 24, 5.5, after table 4, b):
infinity representations -> representations of infinity
page 30, 7.2:
floating-point representation ->
representation of floating-point data
floating-point representation for the result ->
representation for a floating-point result
a floating-point representation ->
the representation of the floating-point number
page 43, 7.9:
floating-point representation -> floating-point number
page 46, 7.12:
floating-point representation -> floating-point number
page 47, 7.12.1:
sequence representations for -> sequences representing
page 48, 7.12.2:
sequence representations for -> sequences representing
page 48, 7.12.3:
sequence representations for -> sequences representing
page 49, 7.12.3, after table 11:
external representation -> external character sequence
Format (500 even, wow):
-----------------------
page 10, 1.2:
Binary and decimal floating-point data formats ->
Formats for binary and decimal floaring-point data
page 34, 7.4.1, in formatOf squareRoot:
decimal format results -> decimal results
decimal format results -> decimal results
page 34, 7.4.1, in formatOf division:
decimal format results -> decimal results
decimal format results -> decimal results
page 34, 7.4.1, in formatOf fusedMultiplyAdd:
decimal format results -> decimal results
decimal format results -> decimal results
page 35, 7.4.2, in formatOf convert:
decimal format results -> decimal results
page 48, 7.12.2:
an external hexadecimal character sequence format ->
external hexadecimal character sequences
NaN (195):
----------
(no changes)
Infinite (32):
--------------
(no changes)
Infinity (28):
--------------
(no changes)
Finite (61):
------------
(no changes)
zero (97):
----------
(no changes)
non-zero (3):
-------------
(no changes)
nonzero (16):
-------------
(no changes)
Normal (50):
------------
(no changes)
Subnormal (24):
---------------
(no changes)
Denormal (2, that's bad):
-------------------------
first occurance, page 13, 3.2.10: OK, backward reference to 1985.
second occurance, page 15, 3.2.35: OK, also backward reference.
(no changes)