| 1 |
Jan Kundrát wrote: |
| 2 |
> Steve Long wrote: |
| 3 |
>>> insinto /usr/share/doc/${P}/examples |
| 4 |
>> Is there any chance we can start using correctly quoted filenames across |
| 5 |
>> the board? |
| 6 |
> |
| 7 |
> Since when is ${P} allowed to have spaces? |
| 8 |
> |
| 9 |
I believe I answered this in my prior post. |
| 10 |
|
| 11 |
>> Besides being faster (quote the whole thing) |
| 12 |
> |
| 13 |
> Have you done a benchmark certifying that "/usr/share/doc/${P}/examples" |
| 14 |
> is faster than /usr/share/doc/${P}/examples? |
| 15 |
> |
| 16 |
Here you go (this is on an old machine, so you'll get much quicker times if |
| 17 |
you try this at home): |
| 18 |
[igli@box ~]$ echo "$(<run)" |
| 19 |
#!/bin/bash |
| 20 |
P='some-crap/god-i-hate-asshats' |
| 21 |
for ((i=0;i<100000;i++)); do echo /usr/share/doc/${P}/examples > /dev/null; |
| 22 |
done |
| 23 |
[igli@box ~]$ echo "$(<run2)" |
| 24 |
#!/bin/bash |
| 25 |
P='some-crap/god-i-hate-asshats' |
| 26 |
for ((i=0;i<100000;i++)); do echo "/usr/share/doc/$P/examples" > /dev/null; |
| 27 |
done |
| 28 |
|
| 29 |
[igli@box ~]$ for ((l=0;l<5;l++)); do time -p ./run; done |
| 30 |
real 11.25 |
| 31 |
user 9.96 |
| 32 |
sys 1.13 |
| 33 |
real 11.12 |
| 34 |
user 9.82 |
| 35 |
sys 1.16 |
| 36 |
real 10.99 |
| 37 |
user 9.70 |
| 38 |
sys 1.15 |
| 39 |
real 11.25 |
| 40 |
user 10.02 |
| 41 |
sys 1.07 |
| 42 |
real 11.36 |
| 43 |
user 10.06 |
| 44 |
sys 1.16 |
| 45 |
|
| 46 |
[igli@box ~]$ for ((l=0;l<5;l++)); do time -p ./run2; done |
| 47 |
real 9.24 |
| 48 |
user 8.00 |
| 49 |
sys 1.11 |
| 50 |
real 9.22 |
| 51 |
user 8.02 |
| 52 |
sys 1.08 |
| 53 |
real 9.08 |
| 54 |
user 7.90 |
| 55 |
sys 1.06 |
| 56 |
real 9.23 |
| 57 |
user 7.96 |
| 58 |
sys 1.15 |
| 59 |
real 9.18 |
| 60 |
user 7.98 |
| 61 |
sys 1.09 |
| 62 |
|
| 63 |
>> more useful error messages in the case of a dev/user slip-up |
| 64 |
> |
| 65 |
> Could you elaborate? |
| 66 |
> |
| 67 |
File not found "foo/bar bar/" as opposed to two separate lines; for elib |
| 68 |
functions, one would expect there to be an error handler which would be |
| 69 |
wrapping that, per-parameter. |
| 70 |
|
| 71 |
>> They also highlight better in a capable editor.[1] |
| 72 |
> |
| 73 |
> Please elaborate. |
| 74 |
> |
| 75 |
I did. |
Archives