Re: Weird spelling fixes in 2.1.107

Peter Swain (swine@softway.com.au)
Fri, 26 Jun 1998 23:44:39 +1000 (EST)


> > > - "cpu family\t: %c\n"
> > > + "CPU family\t: %c\n"
> > > The worst one -- this breaks interface between the kernel and apps.
> > The /proc/cpuinfo file is normally read by humans. I think applications
> > which use it should be adapted to what is best suited for humans.
> This is total CRAP, sorry. It was bad enough that some dork removed
> the underscores, making automated parsing *MUCH* harder, now this...
> I totally agree here, there are for example a bunch of scripts for the
> see if an SMP kernel should be installed etc. You just can't break
> this sort of thing.

<sermon>

this is 2.1.xxx
you *can* and should break things here
so 2.2 gets a consistent interface, in which the doco doesn't have to
apologise for the way things just happen to have been coded.

CPU is in caps
central processing unit is not

either we abandon caps completely
or we get them right

most of the little niggly grammar-police patches
move toward a more tightly defined keyword set,
allowing glossaries to shrink by removing redundant redundant
terms (cpu: see CPU; filesys: see file system; recursion: see recursion).

caps are not for SHOUTING, but they do have legitimate uses in English:
they signal (esp. to non-native speakers) a collection of letters for which
phonetic pronunciation and/or dictionary lookup is likely futile.

i can't speak for the grammar-police, but i suspect their aims can be
summarised concisely (would someone utter the URL of their Manifesto, please?)
- base language is simple, correct english.
- each concept has *one* consistent spelling and usage
- all non-obvious# jargon to be documented somewhere (linux/Doc???)

# 'obvious' to either
1/ native speaker with little-or-no technical vocab
2/ non-native speaker with non-technical dictionary

Those of us who enjoy the privilege of honing our English in
day-to-day conversation don't notice the extra effort we don't
have to make translating comments, identifiers, etc. from the
original contributor's tongue.
We could have Linus' tecnonerdish jokes embedded in the code in
Finnish/Swedish, much of the networking code commented in Russian,
video drivers writ in an unpredictable mix of color/colour greyscale/grayscale,
the ISDN layers in German, ...

It could so easily have been otherwise.
Europeans would settle on a polyglot mix of core languages (most write 2
and read 3..6) to develop the kernel in -- this would hamper them a bit,
but totally outwierd (say) americans, who just couldn't cope.

if you're one of the anal-rebellious who insist on kewl spelling becoz
anything correct smells faintly of having yr ass whupped at skool by a
spelling-nerd teacher, keep yer hat on. An international software project
is one of the very few things for which this really does matter.
Anyone who *always* gets pedantic about these things in daily life has quite
possibly lost the plot, but it does have its place.

we need spelling pedants.
we need security pedants.
we need intellectual property pedants.
we need architecture-independance pedants.
we need damn-fast-code pedants.
we need future-proofing pedants.

patches from each should be welcomed.

</sermon>

we probably don't need my OTT response, but the thread was reminiscent of too
many arguments of this style.

it isn't directed at any of today's posters, more a splenetic explosion
at the spelling-dork i live with, whose neat, tidy little world is quite
unconnected with reality.
On this issue he's right for all the wrong reasons.
i HATE that, so i'm spelling out my reasons.

^..^ piglet
(qp) no, the above don't comply with its' own guidelines, it only needs
to be said to *some* native anglophones, and they can parse it.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu