1 |
On Mon, 31 Aug 2015 20:33:42 -0400 |
2 |
Fernando Rodriguez <frodriguez.developer@×××××××.com> wrote: |
3 |
|
4 |
> On Tuesday, September 01, 2015 12:13:25 AM Alan McKinnon wrote: |
5 |
> > On 31/08/2015 23:13, walt wrote: |
6 |
> > > I ask this strange question because this (badly broken) machine |
7 |
> > > once again flipped between 6.0 and 6.0-r1 after rsyncing this |
8 |
> > > morning. |
9 |
> > > |
10 |
> > > First, it emerged 6.0, which turned out to be almost catastrophic |
11 |
> > > because the qmerge phase of the emerge failed (although it claimed |
12 |
> > > success afterwards) and deleted the entire /usr/share/terminfo |
13 |
> > > subdirectory. That was fun, but I won't bore you with the |
14 |
> > > details. (The ncurses-6.0 files in /lib64 are dated August 28, |
15 |
> > > BTW.) |
16 |
> > > |
17 |
> > > Right now emerge tries to install ncurses-6.0-r1 but the 32-bit |
18 |
> > > part of the build fails because emerge never ran make in the |
19 |
> > > work/cross/progs directory, and so the 32-bit tools didn't get |
20 |
> > > compiled. |
21 |
> > > |
22 |
> > > I hacked around this by running make in that directory manually, |
23 |
> > > which allowed the ebuild install and ebuild package phases to |
24 |
> > > succeed. |
25 |
> > > |
26 |
> > > Now I have an ncurses-6.0-r1 binary package available but I'm too |
27 |
> > > scared to install it because I might need to kill myself |
28 |
> > > afterwards :/ |
29 |
> > > |
30 |
> > > Any suggestions before I take the plunge? Is ncurses-6.0-r1 the |
31 |
> > > right version as of today, Aug 31? |
32 |
> > > |
33 |
> > > Thanks. |
34 |
> > > |
35 |
> > > |
36 |
> > > |
37 |
> > |
38 |
> > |
39 |
> > |
40 |
> > This machine was entirely unaffected by all the recent ncurses |
41 |
> > issues: |
42 |
> > |
43 |
> > [I] sys-libs/ncurses |
44 |
> > Available versions: |
45 |
> > (0) 5.9-r3 (~)5.9-r4 5.9-r5(0/5) (~)6.0-r1(0/6) |
46 |
> > (5) 5.9-r99(5/5) (~)5.9-r101(5/5) (~)6.0(5/6) |
47 |
> > {ada +cxx debug doc gpm minimal profile static-libs test |
48 |
> > threads tinfo trace unicode ABI_MIPS="n32 n64 o32" ABI_PPC="32 64" |
49 |
> > ABI_S390="32 64" ABI_X86="32 64 x32"} |
50 |
> > Installed versions: 6.0-r1(12:52:29 30/08/2015)(cxx gpm |
51 |
> > threads unicode -ada -debug -doc -minimal -profile -static-libs |
52 |
> > -test -tinfo -trace ABI_MIPS="-n32 -n64 -o32" ABI_PPC="-32 -64" |
53 |
> > ABI_S390="-32 -64" ABI_X86="32 64 -x32") |
54 |
> > Homepage: https://www.gnu.org/software/ncurses/ |
55 |
> > http://dickey.his.com/ncurses/ |
56 |
> > Description: console display library |
57 |
> > |
58 |
> > So 6.0-r1 works completely relaible on at least one Gentoo machine |
59 |
> > in this world :-) |
60 |
> |
61 |
> Hmm, I keyworded ncurses and this is what portage wants to do: |
62 |
> |
63 |
> [ebuild r U ~] sys-libs/ncurses-6.0-r1:0/6::gentoo |
64 |
> [5.9-r5:0/5::fernan] USE="cxx doc gpm tinfo unicode -ada -debug |
65 |
> -minimal -profile -static-libs {- test%} -threads% -trace" |
66 |
> ABI_X86="32 (64) -x32" 3,059 KiB [ebuild U ~] |
67 |
> sys-libs/ncurses-5.9-r101:5::gentoo [5.9-r99:5::gentoo] USE="gpm |
68 |
> tinfo unicode (-ada%) (-cxx%*) (-static-libs%)" ABI_X86="32 (64) - |
69 |
> x32" 0 KiB [ebuild rR ~] sys-devel/gdb-7.10::gentoo USE="client |
70 |
> expat python server zlib -lzma -multitarget -nls {-test} -vanilla" |
71 |
> PYTHON_SINGLE_TARGET="python2_7 -python3_3 -python3_4" |
72 |
> PYTHON_TARGETS="python2_7 python3_4 -python3_3" 0 KiB [ebuild rR |
73 |
> ] app-misc/screen-4.3.1::gentoo USE="pam -debug -multiuser - nethack |
74 |
> -selinux" 0 KiB [ebuild rR ] app-emulation/wine-1.6.2::gentoo |
75 |
> USE="X alsa cups custom- cflags fontconfig gecko jpeg lcms ldap mp3 |
76 |
> ncurses openal opengl perl png prelink pulseaudio run-exes samba ssl |
77 |
> threads truetype udisks v4l xcomposite xinerama xml -capi -dos |
78 |
> -gphoto2 -gsm -gstreamer -mono -nls -odbc -opencl - osmesa -oss |
79 |
> -realtime -scanner -selinux {-test}" ABI_X86="32 64 -x32" |
80 |
> LINGUAS="-ar -bg -ca -cs -da -de -el -en -en_US -eo -es -fa -fi -fr |
81 |
> -he -hi -hr -hu -it -ja -ko -lt -ml -nb_NO -nl -or -pa -pl -pt_BR |
82 |
> -pt_PT -rm -ro -ru -sk - sl -sr_RS@cyrillic -sr_RS@latin -sv -te -th |
83 |
> -tr -uk -wa -zh_CN -zh_TW" 0 KiB |
84 |
> |
85 |
> |
86 |
> That looks dangerous to me because the first build will upgrade my |
87 |
> 5.9 installation to 6.0 and the second will reinstall 5.9. |
88 |
|
89 |
That's exactly what happened to me last week and it was a disaster. |
90 |
Don't allow that to happen. After hours of frustration I finally |
91 |
got 6.0-r1 installed and everything Just Works again, but 6.0 was |
92 |
another disaster. Do whatever you need to do to avoid 6.0. |
93 |
|
94 |
> So what happens in between when I have no 5.9 installed but |
95 |
> everything is linked against it? Won't it need bash to build the |
96 |
> second one? What if the 2nd build fails? Will stuff linked against |
97 |
> 5.9 work with 6.0? |
98 |
|
99 |
No, but packages linked against 5.9 will continue to work if portage |
100 |
doesn't delete the files from 5.9 (@preserved-rebuild, etc) |
101 |
|
102 |
I suggest you quickpkg whatever ncurses you have now before you do |
103 |
anything else. |
104 |
|
105 |
I also suggest you see what portage wants to do if you |
106 |
|
107 |
emerge -p =sys-libs/ncurses-6.0-r1 |
108 |
|
109 |
If the output from that looks reasonable then build 6.0-r1 as a binary |
110 |
package *before* you install it so you have the binary packages for |
111 |
both versions ready to go. |
112 |
|
113 |
I ran into other problems during the ncurses upgrade but I managed to |
114 |
work through them, and you know more about this stuff than I do :) |