| 1 |
Philip Webb <purslow@××××××××.net> wrote: |
| 2 |
> |
| 3 |
>> If you're willing to wait an hour, it might be able to come up |
| 4 |
>> with a list of ways you could resolve a conflict, but basically |
| 5 |
>> all of them will be wrong, eg suggestion #1, uninstall everything. |
| 6 |
> |
| 7 |
> Really, this is a flippant response to a serious issue, |
| 8 |
|
| 9 |
No. It is how facts are. Actually, making suggestions is an |
| 10 |
NP-complete problem which means that if already, say, 20 possible |
| 11 |
choices have influence to the solution (and cannot be trivially cut), |
| 12 |
it would take hundreds of years to complete. If you come up with an |
| 13 |
algorithm which can do essentially better, you have probably solved |
| 14 |
the P=NP millenium problem and can expect an incredible award. |
| 15 |
You can be very lucky if you get *one* suggestion (which is |
| 16 |
what portage sometimes finds with high value of --backtrack). |
| 17 |
Checking whether the suggestion is reasonable is the user's task. |
Archives