summaryrefslogtreecommitdiff
blob: 872f1545f0a010509f4561b4ea3eb9b34e40fc09 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
[22:22:02] <dilfridge> ssuominen: around=
[22:22:04] <dilfridge> ?
[22:43:31] <Anarchy> jlec, that will be fine no rush on it.
[22:55:58] <ssuominen> dilfridge: for a moment
[22:57:27] <dilfridge> I have a question on how to best bump an ebuild to the tree
[22:57:34] <dilfridge> about "required stability in ~arch"
[22:57:35] <robbat2|na> dilfridge, sup?
[22:58:04] <dilfridge> I would like to bump app-admin/collectd to the tre
[22:58:11] <dilfridge> http://git.overlays.gentoo.org/gitweb/?p=user/dilfridge.git;a=tree;f=app-admin/collectd;hb=HEAD
[22:58:30] <dilfridge> this has a huge number of plugins, and some require specific hardware / software
[22:58:56] <dilfridge> Plugin building is controlled by useflags, but I cant test all
[22:59:29] <dilfridge> dberkhol suggested that I have two ebuilds, a ~arch which has only the "tested building plugins" and a masked with all
[22:59:47] <chithead> you cannot build-test all?
[22:59:56] <dilfridge> is there a way to mask e.g. -r51 .. -r99 for all versions?
[23:00:02] <robbat2|na> "to the tree"? bump is not the word at all. migrate to the tree
[23:00:11] <robbat2|na> bump = version bump or revision bump
[23:00:15] <ssuominen> build-test all would be enough to stick it into ~arch
[23:00:18] <dilfridge> chithead: that might be possible 
[23:00:30] <robbat2|na> yes, just build-test them all and put that as ~arch
[23:00:45] <dilfridge> ok then I will target that and "activate" all useflags that I have build-tested. 
[23:01:05] <dilfridge> thx guys problem solved, will take a few more days though...
[23:01:23] <robbat2|na> and that ebuild looks like it's going to want USE_EXPAND instead of the manual cd_ stuff
[23:01:27] <dilfridge> yes true
[23:01:47] <dilfridge> how much buerocracy loops are needed for USE_EXPAND?
[23:02:05] <dilfridge> s/buerocracy/bureaucracy/
[23:02:36] <dilfridge> just ask on dev ml?
[23:03:05] <robbat2|na> yup
[23:03:11] <dilfridge> ok will do
[23:03:24] <robbat2|na> also, the errors you spew when linux_config_exists is false are very misleading
[23:03:33] <dilfridge> ok
[23:03:46] <robbat2|na> unless i'm misreading something
[23:03:47] <robbat2|na> one sec
[23:04:36] <darkside_> uhm, what? # hal is autodetected by configure, so there is no point in a hal useflag.
[23:05:08] <dilfridge> darkside_: yes not patched yet, but this affects only _one_ plugin
[23:05:45] <dilfridge> robbat2|na: this tries to figure out what kernel options might be improved for the plugins to work
[23:05:46] -*- ssuominen -C slocate
[23:06:06] <robbat2|na> collectd_plugin_kernel_linux <--- i'd refactor that, to have a specific error for linux_config_exists failure. also your linux_chkconfig_present will only correctly check the first flag 
if multiple are specified
[23:06:41] <dilfridge> ok will reread
[23:06:57] <robbat2|na> if linux_chkconfig_present ${opt}; then return 0; fi
[23:07:09] <ssuominen> there's a requirement to ask on dev ML before adding USE_EXPAND?
[23:07:22] <robbat2|na> so you return out of the function on the first successful option, no way to specify multiple options must be set together
[23:07:39] <robbat2|na> ssuominen, non written, but every past one was asked about
[23:08:00] <ssuominen> robbat2|na: except the one I added just week ago, then ;-)
[23:08:07] <ssuominen> one for XFCE_PLUGINS
[23:09:11] <robbat2|na> dilfridge, lastly, why a DEPEND on sys-kernel/vserver-sources? that's a perfectly valid kernel to be prebuilt and have no kernel sources on a machine
[23:09:42] <dilfridge> robbat2|na: yes that's true but why not go for "check one option" for simplicity's sake
[23:10:23] <robbat2|na> dilfridge, ok, just document that in the function header then
[23:10:34] <dilfridge> it is :)
[23:11:01] <dilfridge> robbat2|na: about the vserver, simply because I did not get to that plugin yet (and have no experience with that)
[23:11:12] <robbat2|na> use the eclass doc syntax?
[23:11:19] <dilfridge> ok
[23:11:36] <dilfridge> will do
[23:12:59] <dilfridge> ok give me a few days and I will clean things up and send mail :) thanks for your help


 544   * libnatapp (optional)
 545     Required for the “netapp” plugin.
 546     This library is part of the “Manage ONTAP SDK” published by NetApp.

 560   * liboping (optional)
 561     Used by the `ping' plugin to send and receive ICMP packets.
 562     <http://verplant.org/liboping/>

 586   * librouteros (optional)
 587     Used by the `routeros' plugin to connect to a device running `RouterOS'.
 588     <http://verplant.org/librouteros/>

 596   * librt, libsocket, libkstat, libdevinfo (optional)
 597     Various standard Solaris libraries which provide system functions.
 598     <http://developers.sun.com/solaris/>

 625   * libxmms (optional)
 626     <http://www.xmms.org/>

 628   * libyajl (optional)
 629     Parse JSON data. This is needed for the `curl_json' plugin.
 630     <http://github.com/lloyd/yajl>