| 1 |
2010/1/13 basile <basile@××××××××××××××.edu> |
| 2 |
|
| 3 |
> Hi guys, |
| 4 |
> |
| 5 |
> I'm emailing because the profile issue came up again in IRC. I'd like |
| 6 |
> to continue the discussion here. Let me try to get it started. |
| 7 |
> |
| 8 |
> Here's some general issues with the current profile stucture: |
| 9 |
> |
| 10 |
> 1) It is horribly complex and difficult to read the inheritance |
| 11 |
> strucutre. Its not clear the inheritance even works. As a result, the |
| 12 |
> user is not sure what is going on. This ambiguity makes it difficult |
| 13 |
> to even start a coherent criticism! |
| 14 |
> |
| 15 |
> 2) There doesn't appear to be a good structure for seperation of various |
| 16 |
> features. In OO language, I can't choose what to inherit. I wind up |
| 17 |
> getting stuff from other profiles which I don't want and can't control |
| 18 |
> this, so I'm tempted to just USE="-*" and start from scratch, which is |
| 19 |
> not a good thing. |
| 20 |
> |
| 21 |
> 3) There is a clear bias towards the desktop. If you go that route, you |
| 22 |
> get what you need/want. When you deviate, you start to get more things |
| 23 |
> that you don't want/need and have to struggle against points 1 and 2. |
| 24 |
> |
| 25 |
> This effects hardened and hardened+server most. Comments? |
| 26 |
> |
| 27 |
|
| 28 |
I don't really get the productive side of this message, but I do agree with |
| 29 |
all that points. |
| 30 |
|
| 31 |
|
| 32 |
-- |
| 33 |
Pierre. |
| 34 |
"Sometimes when I'm talking, my words can't keep up with my thoughts. I |
| 35 |
wonder why we think faster than we speak. Probably so we can think twice." - |
| 36 |
Bill Watterson |
Archives