| 1 |
On 11/14/12 9:09 PM, Diego Elio Pettenò wrote: |
| 2 |
> On 14/11/2012 21:02, "Paweł Hajdan, Jr." wrote: |
| 3 |
>> My _limited_ testing seems to indicate it's "working". But if you |
| 4 |
>> people can just do your tests, that would be most welcome. Please |
| 5 |
>> report back either success or failure. |
| 6 |
> |
| 7 |
> What kind of testing are you thinking of? Execution or build? |
| 8 |
|
| 9 |
Build. But also possibly the upgrade paths, e.g. make sure the ABI is |
| 10 |
not changed on gcc upgrade to 4.7 and similar cases (well actually ABI |
| 11 |
change with an SONAME change is fine; it's mostly changing ABI without |
| 12 |
changing SONAME which is a very bad thing). |
| 13 |
|
| 14 |
> And I assume this is still under the idea that C++11 support is kept |
| 15 |
> _disabled_ for now. |
| 16 |
|
| 17 |
-r1 disables C++11 completely, while -r2 just allows reverse |
| 18 |
dependencies of icu not to support C++11, while possibly making icu use |
| 19 |
C++11 features internally (I was asked by an upstream icu developer to |
| 20 |
test that; if this still causes problems, we should report them to him). |
| 21 |
|
| 22 |
> Anyway I'll run an icu tinderbox soon as the current one is done |
| 23 |
> (libav-9). |
| 24 |
|
| 25 |
Sounds good! |
Archives