Gentoo Archives: gentoo-dev

From: R0b0t1 <r030t1@×××××.com>
To: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] New item for sys-kernel/hardened-sources removal
Date: Sun, 20 Aug 2017 06:05:28
Message-Id: CAAD4mYi1C2Y7USyEaSdem5rsw6p9MLGyLS_hXbq35LznqJ_dRA@mail.gmail.com
In Reply to: Re: [gentoo-dev] New item for sys-kernel/hardened-sources removal by R0b0t1
1 On Sun, Aug 20, 2017 at 12:39 AM, R0b0t1 <r030t1@×××××.com> wrote:
2 > On Sat, Aug 19, 2017 at 6:34 AM, Francisco Blas Izquierdo Riera
3 > (klondike) <klondike@g.o> wrote:
4 >> El 19/08/17 a las 13:18, Aaron W. Swenson escribió:
5 >>> On 2017-08-19 13:01, Francisco Blas Izquierdo Riera (klondike) wrote:
6 >>>> El 19/08/17 a las 12:37, Aaron W. Swenson escribió:
7 >>>>> On 2017-08-15 17:01, Francisco Blas Izquierdo Riera (klondike) wrote:
8 >>>>>> Hi!
9 >>>>>>
10 >>>>>> I'd like to get this one up by Saturday so that we can proceed with
11 >>>>>> masking and removing of the hardened-sources after upstream stopped
12 >>>>>> releasing new patches.
13 >>>>> I hope I’m not too late.
14 >>>>>
15 >>>>>> We'd like to note that all the userspace hardening and MAC support
16 >>>>>> for SELinux provided by Gentoo Hardened will still remain there and
17 >>>>>> is unaffected by this removal.
18 >>>>> Where is there? I think you’re talking about the packages, but the news
19 >>>>> item is about the kernels. It would help to be more specific here.
20 >>>>>
21 >>>>> That’s all I had that the others hadn’t touched on.
22 >>>> Do you think something like that is better then?
23 >>>>
24 >>>> We'd like to note that all the userspace hardening and MAC support
25 >>>> for SELinux provided by Gentoo Hardened will still remain available
26 >>>> on the portage. Keep in mind though that the security provided by
27 >>>> these features will be weakened a bit when using
28 >>>> sys-kernel/gentoo-sources. Also, all PaX related packages other than
29 >>>> the hardened-sources will remain available for the time being.
30 >>>>
31 >>>>
32 >>> Much better. We should mention that we’re specifically discussing
33 >>> packages and not portage itself. At least, that’s my understanding from
34 >>> your edit.
35 >>>
36 >>> Here’s my take on it:
37 >>>
38 >>> We'd like to note that all the userspace hardening and MAC support for
39 >>> SELinux provided by Gentoo Hardened will still remain in the packages
40 >>> found in portage. Keep in mind, though, that the security provided by
41 >>> these features will be weakened a bit when using
42 >>> sys-kernel/gentoo-sources. Also, all PaX related packages, except
43 >>> sys-kernel/hardened-sources, will remain available for the time being.
44 >>
45 >> I updated the news item with your propossal. Thanks a lot :)
46 >>
47 >
48 > The discussion is nice but no one has actually touched on the
49 > technical merits of removing the packages besides "they are old."
50 > There's plenty of old software in portage. Why not remove it first?
51 >
52 > I had a similar issue with the GCC developer who removed GCJ support.
53 > I asked him for any justification at all for the removal and he had
54 > none but some vague statements about it creating work. I would have
55 > taken any more specific example he gave at face value, but he didn't
56 > want to give one. I was left to conclude he didn't have one to give.
57 >
58 > So I ask again: On what basis are the hardened sources being removed
59 > from the tree?
60 >
61 > At this point I am far less interested in making sure the sources stay
62 > in the tree than I am in forcing you to justify your actions, because
63 > I suspect your attempt to do so will be entertaining.
64 >
65
66 I just had a bad day so perhaps that last bit was a tad blunt.
67 Consider replacing it with this:
68
69 There is nothing that holds you accountable to me. However, I am
70 honestly trying to understand why you are doing what you are doing and
71 would like you to explain your decision making process to me. If you
72 can't explain it to me, then how do you know that you have selected
73 the best course of action?
74
75 If it was a matter of opinion I can accept you will probably go "I'm a
76 developer" and then ignore me. However I don't think it has gotten to
77 that point yet, and you are doing the thing being discussed for what
78 seems to be nebulous and poorly defined reasons.
79
80 R0b0t1.