| [hatari-devel] Re: Colloquialisms |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/hatari-devel Archives
]
Hi,
On 10/18/18 6:03 PM, Roger Burrows wrote:
On 18 Oct 2018 at 9:48, Thomas Huth wrote:
As a native English speaker (currently living in Canada, but originally from
the UK), I'd say that:
1. you would almost always use "it's" in spoken language
2. you would almost always use "it is" in a printed document
Online is somehow between the two, so it's hard to be dogmatic. For me, "it
is" creates an impression of greater accuracy & precision, but less
friendliness. So I would use "it's" in a FAQ, but "it is" in a manual. But
that's just an opinion, and "honest men may differ".
>
By the way, looks like we're rather informal in the manual:
$ grep -i " it's " manual.html | wc -l
22
$ grep -i " it is " manual.html | wc -l
11
... maybe we should standardize on one of the two spellings?
I think that is definitely something that should be fixed.
Looking whose latest commits use the contracted form:
for ext in html 1 txt; do
echo "*** *.$ext ***";
hg files "glob:**.$ext" | xargs hg blame -u |\
fgrep -i -e "it's" -e "they're" -e "isn't" -e "aren't" \
-e "can't" -e "hasn't" -e "haven't" -e "doesn't" -e "don't" |\
cut -d: -f1 | tr -d ' ' | sort -n | uniq -c | sort -nr;
done
*** *.html ***
92 oak
38 eerot
9 Nicolas
7 Thomas
2 simonsunnyboy
1 thothy
*** *.1 ***
14 oak
4 eerot
1 Nicolas
*** *.txt ***
86 oak
30 eerot
8 Nicolas
5 thothy
2 Thomas
2 huth
1 npomarede
And the more formal forms:
for ext in html 1 txt; do
echo "*** *.$ext ***";
hg files "glob:**.$ext" | xargs hg blame -u |\
grep -i -e "it *is" -e "they *are" -e "is *not" \
-e "are *not" -e "can *not" -e "has *not" \
-e "have *not" -e "does *not" -e "do *not" |\
cut -d: -f1 | tr -d ' ' | sort -n | uniq -c | sort -nr;
done
*** *.html ***
29 oak
23 Nicolas
21 Thomas
12 simonsunnyboy
11 eerot
4 thothy
4 npomarede
2 huth
*** *.1 ***
5 oak
2 Thomas
2 eerot
*** *.txt ***
33 thothy
12 oak
6 Nicolas
5 Thomas
4 eerot
It's clear who are the sources for these forms. :-)
As expected, I've used contracted forms nearly exclusively,
and Thomas the formal ones.
(I'm fine with both forms, but colloquial form is more natural to me.
I read a lot of English fiction, but I don't write or review academic
papers.)
Looking at who has last touched specific lines in documentation:
for ext in html 1 txt; do
echo "*** *.$ext ***";
hg files "glob:**.$ext" | fgrep -v -e CMake -e /results/ |\
xargs hg blame -u | cut -d: -f1 | tr -d ' ' |\
sort -n | uniq -c | sort -nr;
done
*** *.html ***
3469 eerot
3006 oak
1100 Thomas
593 Nicolas
293 thothy
276 simonsunnyboy
195 huth
164 npomarede
16 laurent
6 laurent29
*** *.1 ***
760 oak
631 eerot
352 thothy
48 Nicolas
22 Thomas
9 npomarede
7 huth
*** *.txt ***
2360 oak
1295 eerot
816 thothy
507 Nicolas
332 Thomas
154 huth
41 npomarede
2 Laurent
1 laurent29
1 laurent
As I've touched most documentation lines, there were also more
instances of the contracted form. I don't think I can promise to
consistently use the formal form without automated tests for it.
Automated test should be able to detect contracted forms more
reliably than formal ones (as they use a single word). I'm fine
if somebody converts all docs and adds such a test.
- Eero
PS. stats for last person touching given line of code:
for ext in c h py sh; do
echo "*** *.$ext ***";
hg files "glob:**.$ext" | xargs hg blame -u |\
cut -d: -f1 | tr -d ' ' | sort -n | uniq -c | sort -nr;
done
*** *.c ***
71376 Nicolas
30937 thothy
21349 Thomas
18004 eerot
15579 oak
13205 Laurent
9256 laurent
5345 huth
4471 npomarede
1617 laurent29
1300 admin
150 jb
130 Uwe
80 lafrog
56 simonsunnyboy
35 marndt
30 emanne
13 markus
11 clafou
9 laurentS
6 laurents
3 Huth
*** *.h ***
9450 Nicolas
3943 thothy
3523 Laurent
2921 Thomas
1033 eerot
793 oak
342 npomarede
257 laurent
256 huth
120 clafou
86 admin
63 laurent29
34 simonsunnyboy
15 emanne
7 marndt
6 lafrog
4 Uwe
4 markus
*** *.py ***
3850 oak
3845 eerot
13 Nicolas
6 Thomas
*** *.sh ***
678 oak
477 eerot
290 Thomas
19 huth
2 Nicolas