1 |
On Mon, 4 Dec 2017 03:58:40 +0100, tuxic@××××××.de wrote: |
2 |
|
3 |
> what could fail, when doing the change to PIE-enabled applications |
4 |
> on base of the regular updates? |
5 |
> Compilation may fail, if libs are included and not flagged as to be |
6 |
> recompiled, which are of the "old standard"... |
7 |
> What else can fail? What may be the worst scenario? |
8 |
|
9 |
Anything. You are rememrging packages that may have changed, or their |
10 |
dependencies, so you are bound to find the odd problem. |
11 |
|
12 |
> Is there a way to do a "emerge -e @world" but only for the system |
13 |
> applications? |
14 |
|
15 |
emerge @system, but |
16 |
> |
17 |
> Would it be possible to do a "emerge -e @world" for the system |
18 |
> applications and then update the rest of the applications via the |
19 |
> regular updates of the system (and recompile failing components |
20 |
> manually because one obviously already know the reason) ? |
21 |
|
22 |
How do you know which packages are important? By doing a partial update |
23 |
you risk more problems and time wastage than just re-emerging @world in |
24 |
one go. |
25 |
|
26 |
> Do I have to do a "emerge -e @world" from a certain kind of |
27 |
> "reduced system" i.e. starting the system without a desktop |
28 |
> first or boot into an even more reduced state aka "maintance |
29 |
> mode" (via grub) and make the disk rw by hand? |
30 |
|
31 |
No, just do it. Add --keep-going to the emerge command and it will spit |
32 |
out a list of any failed packages at the end. Then you can deal with |
33 |
those as and when you see fit. In the time you have spent worrying about |
34 |
this, I have updated two systems, one after the other I wasn't brave |
35 |
enough to try in parallel, and dealt with the build failures. You are |
36 |
just creating more work for yourself. |
37 |
|
38 |
|
39 |
-- |
40 |
Neil Bothwick |
41 |
|
42 |
Sometimes too much to drink is not enough. |