| 1 |
Am 26.03.2016 um 16:40 schrieb Rich Freeman: |
| 2 |
> On Sat, Mar 26, 2016 at 10:48 AM, Volker Armin Hemmann |
| 3 |
> <volkerarmin@××××××××××.com> wrote: |
| 4 |
>> because it is broken by design, a security nightmare and seriously not |
| 5 |
>> needed at all? |
| 6 |
>> |
| 7 |
> While there is general interest in a better design, Linus believes it |
| 8 |
> is in fact needed and intends to merge the ultimate result. The |
| 9 |
> concern is with the design of kdbus itself, not the concept. It is |
| 10 |
> just a more rigorous form of IPC. |
| 11 |
> |
| 12 |
> Others are of course welcome to disagree. |
| 13 |
> |
| 14 |
|
| 15 |
hm, back then and everytime kdbus came up on lkml the consensus was |
| 16 |
'speed? you do it for speed? Get userspace dbus in order and the speed |
| 17 |
argument collapses'. Pretty much everybody also voiced problems with |
| 18 |
security (none) and the statefulness of dbus. |
| 19 |
|
| 20 |
All problems, blissfully ignored by the kdbus bunch. |
Archives