1 |
On 26/09/2015 17:00, lee wrote: |
2 |
> Alan McKinnon <alan.mckinnon@×××××.com> writes: |
3 |
> |
4 |
>> On 20/09/2015 17:28, lee wrote: |
5 |
>>> Neil Bothwick <neil@××××××××××.uk> writes: |
6 |
>>> |
7 |
>>>> On Sat, 19 Sep 2015 21:36:06 +0200, lee wrote: |
8 |
> |
9 |
>> [...] |
10 |
>>>>> !!! Multiple package instances within a single package slot have been |
11 |
>>>>> pulled !!! into the dependency graph, resulting in a slot conflict: |
12 |
>> [...] |
13 |
>>>> These are unimportant, it is simply portage telling you it is not |
14 |
>>>> updating some packages to the latest available and why. Personally, I |
15 |
>>>> believe this sort of output should only be shown when using --verbose. |
16 |
>>> |
17 |
>> [...] |
18 |
>>> Should I always ignore such messages? |
19 |
>> |
20 |
>> No, you should not ignore such messages. They are printed for a reason. |
21 |
> |
22 |
> Well, what can I do other than ignore them? With dependencies as they |
23 |
> are, and given that I don't want to remove packages, some of the |
24 |
> packages that could be upgraded to newer versions won't be upgraded |
25 |
> because otherwise things might be broken. There's nothing I could do |
26 |
> about that, or is there? |
27 |
|
28 |
Look, you are over-complicating this and making it way more difficult |
29 |
than it needs to be. |
30 |
|
31 |
We all agree portage would be easier to use if it was less wordy, and if |
32 |
it drew a better distinction between debug, info, error and warning |
33 |
messages. But right now it's not there, so unless you can step up with a |
34 |
high quality patch to improve matters, you have to deal with what is there. |
35 |
|
36 |
Look at the output, take each thing portage is saying and eveluate it on |
37 |
it's own merits. Maybe you need to do something, maybe not. But you have |
38 |
to read them and decide. |
39 |
|
40 |
Your question was should you always ignore such messages, and I forget |
41 |
what the such was. Obviously, no, you must not globally ignore what |
42 |
software is telling you. |
43 |
|
44 |
So clam down, take a chill pill or whatever and deal with portage on |
45 |
it's own terms |
46 |
|
47 |
> |
48 |
>> You have a SLOT conflict and whether that prevents you from proceeding |
49 |
>> or not doesn't change the fact that portage knows you have that conflict. |
50 |
> |
51 |
> Is it possible to solve this conflict without removing packages? |
52 |
|
53 |
NO. YOU DO NOT JUST REMOVE PACKAGES WILLY-NILLY. |
54 |
|
55 |
Neil already explained what a slot conflict is. Portage wants to install |
56 |
two versions from the same slot. Find out why and deal with that. |
57 |
Oftentimes the message is a mere info, telling you why portage won't |
58 |
install the latest. This is actually the same thing as yesterday's |
59 |
question on nvidia-drivers that I already answered. You treat SLOTs and |
60 |
packages the same way, a SLOT is just a subset of all versions of a |
61 |
packages. |
62 |
|
63 |
> |
64 |
>> In your specific case today, I believe portage will simply install the |
65 |
>> lesser version and be done with it, but it will only do that when you |
66 |
>> fix the USE issue (a whole separate issue) |
67 |
> |
68 |
> Probably --- yet it tells me about conflicts, makes them appear to be |
69 |
> important, and leaves me wondering how to solve them. |
70 |
|
71 |
A conflict is just a conflict, doesn't have to be serius. Maybe portage |
72 |
can solve it, maybe not. Either way, you get to read and understand the |
73 |
output. |
74 |
|
75 |
> |
76 |
>> [...] |
77 |
>> The USE conflict for sure. Maybe the SLOT conflict but I think portage |
78 |
>> will just deal with that one |
79 |
>> [...] |
80 |
>>> This one doesn't look very important, or does it? |
81 |
>> |
82 |
>> Chill dude, seriously. The sky is not about to fall on your head and the |
83 |
>> bits on your disk are not going to miraculously re-arrange themselves |
84 |
>> into Windows just because you can't do this update. |
85 |
> |
86 |
> Sure, yet why make unimportant messages look important and important |
87 |
> ones unimportant? |
88 |
|
89 |
Because the devs are human. Ask them. |
90 |
|
91 |
> |
92 |
>> Portage is what it is, deal with it. |
93 |
>> |
94 |
>> The portage team are all unpaid volunteers just liek everyone else and |
95 |
>> none of us have any right at all to make demands of them. Especially not |
96 |
>> you and I who are not active contribution solutions. |
97 |
> |
98 |
> I know --- however, making a suggestion to improve the messages is a |
99 |
> contribution. |
100 |
|
101 |
But freaking out and complaining helps no-one. |
102 |
|
103 |
You appear to not fully understand the nature of the problem and your |
104 |
emotional outbursts are not helping. You keep going round the same |
105 |
circle, complaining about how the output doesn't suit you, but I don;t |
106 |
see evidence yet that you are actually reading it. You need to read it. |
107 |
|
108 |
> |
109 |
>> [...] |
110 |
>>> How about adding comments to such messages, like "You don't need to do |
111 |
>>> anything to be able to proceed." and "You need to fix this before you |
112 |
>>> could proceed."? |
113 |
>> |
114 |
>> If emerge exited then you need to fix something in your config. |
115 |
>> If emerge does not exit then your config can be used as-is. |
116 |
> |
117 |
> Messages more helpful could make it easier to figure out what needs to |
118 |
> be fixed. |
119 |
|
120 |
Learn python, submit a high-quality patch. |
121 |
|
122 |
> |
123 |
>> [...] |
124 |
>>> The last sync I did before the one yesterday wasn't the day before |
125 |
>>> yesterday but over three months ago, so don't ask me today (or next |
126 |
>>> weekend or whenever I give it another try) when that exactly was. See |
127 |
>>> what I mean? Asking me to mask all packages to a certain point in time |
128 |
>>> is like asking me to do much of the package management by myself. |
129 |
>> |
130 |
>> Exactly. You DO need to do the package management yourself. The Gentoo |
131 |
>> devs provide useful tools in the form of portage and the tree with it's |
132 |
>> ebuilds and eclasses, plus some amazing automation. |
133 |
>> |
134 |
>> But, are here's the bit where so many people move away from Gentoo: |
135 |
|
136 |
So what? Gentoo is what it is. There are hundreds of Linux distros out |
137 |
there. If some users are not prepared to do what it takes to run Gentoo, |
138 |
and find something more suited to their needs then they should use that |
139 |
and the Gentoo community will wish them well for the future. |
140 |
|
141 |
>> You are required to do the management yourself, including most of the |
142 |
>> thinking and all of the sweeping up of broken pieces. That's what you |
143 |
>> signed up for when using Gentoo. |
144 |
> |
145 |
> Perhaps not so many people would move away if the messages were |
146 |
> improved. |
147 |
|
148 |
Gentoo is not here to be your personal distro. |
149 |
|
150 |
I think you might be happier with Arch, come to think of it. |
151 |
|
152 |
> |
153 |
>> If you want to roll back the tree, then you need to implement a |
154 |
>> solution that will let you do it as Gentoo does nto provide one. Git |
155 |
>> now makes this easier. |
156 |
> |
157 |
> Converting to btrfs might work for that, if I can boot from it. |
158 |
> |
159 |
>> However, tree rollbacks are inadvisable for excellent technical reasons |
160 |
>> - see if you can figure them out. Better to snapshot your entire system |
161 |
>> and revert the snapshot if it goes south. |
162 |
> |
163 |
> That's not even advisable with sources, though IIRC, the reasons for |
164 |
> that might not apply here. However, it's weird that a system like git |
165 |
> makes it inadvisable to undo something, considering that being able to |
166 |
> undo something very easily, is one important reason to invent and use |
167 |
> such a system in the first place. |
168 |
|
169 |
See below. You are considering the wrong problem. |
170 |
|
171 |
> |
172 |
> Using snapshots for undoing things git is quite an application of |
173 |
> overengineering. |
174 |
|
175 |
The problem is not the tree. The problem is what you do with the tree, |
176 |
and then desire to undo THAT. By far the most common reason to want to |
177 |
rewind the tree is to have used it first. Emerging something usually |
178 |
takes you past the point of no return. |
179 |
|
180 |
p.s. Package management is hard. Really truly fucking annoyingly |
181 |
frustratingly hard. And difficult to solve. That's why Windows doesn't |
182 |
even try, MacOS shoves a walled garden down your throat, Debian thinks |
183 |
their view must be perfect just because, and Red Hat charges lots of |
184 |
money. Other teams try with varying success, giving us abominations like |
185 |
npm. |
186 |
|
187 |
Gentoo is one of the very few (perhaps only) distros that have the balls |
188 |
to tackle this problem head on. Cut it some slack, please. Stop whinging. |
189 |
|
190 |
|
191 |
-- |
192 |
Alan McKinnon |
193 |
alan.mckinnon@×××××.com |