1 |
-----BEGIN PGP SIGNED MESSAGE----- |
2 |
Hash: SHA1 |
3 |
|
4 |
Jan, |
5 |
|
6 |
There appeared to be a quoting issue with your email, so I have |
7 |
reformatted it for my reply. I hope I have the attributions correct, if |
8 |
not I apologise in advance. |
9 |
|
10 |
On 2008.02.03 03:39, Jan Bilek wrote: |
11 |
> Dear Roy |
12 |
> |
13 |
> |
14 |
> > On Feb 3, 2008 1:04 AM, Roy Bamford <neddyseagoon@g.o |
15 |
> > wrote: |
16 |
> |
17 |
> > You need a balance between the formal, which produces formal |
18 |
> > records and the informal, that does not. |
19 |
> |
20 |
> Yes - sure. I wrote that under the impression that there is too much |
21 |
> of a 'formal side' in Gentoo recently. |
22 |
[snip] |
23 |
As others have said, you only see the informal part if you are taking |
24 |
part in it. If you think Gentoo (any of it) is too formal, which |
25 |
information would you as a user be prepared to go without? |
26 |
The premise here is that users only see the formal information. |
27 |
> |
28 |
> |
29 |
> > I think the Gentoo council, which was set up to be a technical |
30 |
> > body is getting bogged down in politics from time to time, which |
31 |
> > impedes its technical decision making process. This is where the |
32 |
> > Gentoo Foundation can help, by taking on all the political aspects |
33 |
> > of our community. |
34 |
> |
35 |
> This might be very difficult task. How are you going to handle it if |
36 |
> elected? I might be totally out of proportions but do you think there |
37 |
> can be some kind of a fight for competences between trustees and |
38 |
> council? I am sorry if that is a stupid question. |
39 |
First, there are no stupid questions, except the one you do not ask. Be |
40 |
on your guard for stupid answers though. |
41 |
I do not think there will be a "fight for competencies" because the |
42 |
council and foundation were established with different aims which need |
43 |
volunteers with different skill sets. The council has had to deal with |
44 |
the politics of Gentoo because the Foundation has not been very active. |
45 |
I do not intend to suggest that the Foundation will snatch any |
46 |
responsibilities away from the council. It will be by agreement with |
47 |
the council. |
48 |
> |
49 |
> |
50 |
> > I hope that's not the intent. My intent is to have the council and |
51 |
> > foundation work together such that the council does not spend its |
52 |
> > time on politics and is free to focus on technical things. This |
53 |
> > will lead to more flexible decision making. |
54 |
> |
55 |
> I am sure that's not the intent. But... and I don't want to sound too |
56 |
> pessimistic... my experience is that these things (regular meetings, |
57 |
> regular summaries etc.), despite good intentions, tend to turn into |
58 |
> meaningless routine which is more of a barrier than any help. |
59 |
> Hopefully this time I am wrong. |
60 |
I would like to model the Foundation on the current council process |
61 |
that seems to work quite well. Allow me to refer you to the council |
62 |
meeting logs and summaries. |
63 |
|
64 |
> |
65 |
> > Gentoo has reached a size where central control can at best, only |
66 |
> > set a direction. It cannot manage details. Gentoo does have some of |
67 |
> > the structure in place for these things to happen - the separate |
68 |
> > projects and herds. Gentoo is in need of middle management - |
69 |
> > perhaps it can come from the Foundation. |
70 |
> |
71 |
> Yes - I think Foundation might be very helpful. But I keep thinking |
72 |
> about structural matters - these things are closely related to |
73 |
> management and efficiency of the system. |
74 |
Gentoo is already a loosely co-operating set of projects. These |
75 |
individual projects seem to work well. |
76 |
|
77 |
> I think very flat and very free and open organizational structure can |
78 |
> be more efficient although more difficult to manage - to set up smart |
79 |
> rules enabling more of a self-managed system (yes, I have read too |
80 |
> much about learning organizations and evolution) - that's a good goal |
81 |
> IMO. |
82 |
In practice (in the workplace anyway) its not possible for one person |
83 |
to manage more that abouot 10 others. At that sort of scale, the one |
84 |
person becomes a full time manager. Gentoo has the same scaling |
85 |
problem. Projects work - they fit the above relationships. |
86 |
|
87 |
> |
88 |
> There is no money in Gentoo so it has to be about fun and freedom - |
89 |
> people have to have options when it doesn't feel like fun. If you |
90 |
> don't like to work with someone anymore you should be supported to |
91 |
> build up - easily - your own independent team inside Gentoo |
92 |
> (supporting internal competition instead of internal fighting) - even |
93 |
> if the team would work on the same thing, seemingly wasting time - |
94 |
> *let them do what they want* strategy - to me it seems somehow |
95 |
> contradictory to 'named positions with responsibilities' - I might be |
96 |
> wrong. |
97 |
> Again - foundation itself is very specific, but speaking about Gentoo |
98 |
> generally - it's not company and it cannot be organized as a company |
99 |
> - the most important thing is fun. |
100 |
I agree with most of that, especially the fun part. |
101 |
However workable orgainsations can be applied to any project. The |
102 |
difference between Gentoo and a company is that the company is supposed |
103 |
to make money. |
104 |
> |
105 |
> |
106 |
> > That's fine for the individual projects but what about the wider |
107 |
> > community and the bigger projects that need to know what is |
108 |
> > happening to the projects that they use. In particular, I'm |
109 |
> > thinking of Release Engineering who are putting together the 2008.0 |
110 |
> > LiveCD. They need to know that the various parts will be ready on |
111 |
> > time. Users like to know whats happening too - how would that |
112 |
> > information be circulated without summaries and reports ? |
113 |
> |
114 |
> Yes - I agree with you. And I am not against summaries and reports |
115 |
> generally. I just think it's good idea to be very careful with |
116 |
> institutionalizing these things. If you have something to tell just |
117 |
> write blog post or comment and there is no need to make it regular or |
118 |
> somehow mandatory... that's what I meant. |
119 |
If reports are irregular and on blogs they will mostly be missed by |
120 |
readers. Blogs are OK for individuals making statements on their own |
121 |
behalf but they can never be offical statements on behalf of Gentoo. |
122 |
|
123 |
> |
124 |
> > > Allowing and promoting funny competition between smaller teams |
125 |
> > > instead of demotivating (because unsolvable) fights inside huge |
126 |
> > > teams frozen in official ways of doing things. |
127 |
> > I have never seen this - can you provide an example please ? |
128 |
> |
129 |
> I am afraid I cannot - it's just an impression I got - I used to read |
130 |
> devel MLs and sometimes I saw technical disputes turned into personal |
131 |
> attacks and some people even left - and I felt it was out of |
132 |
> frustration that those extremely valuable people had no other way to |
133 |
> do anything constructive with their disagreement - I felt it could be |
134 |
> somehow organizational failure and not just the personal one. But I |
135 |
> have to admit that I am probably not competent enough to assess these |
136 |
> things - I have never been a developer. |
137 |
> |
138 |
> |
139 |
> > Some technical discussions really do only have a single solution. |
140 |
> |
141 |
> No doubt about it. But to have opportunity to try my solution and |
142 |
> fail is much better than never-ending fights. |
143 |
> |
144 |
> |
145 |
> > Like the portage, plaudis, pkgcore developments in progress at |
146 |
> > them moment perhaps ? |
147 |
> |
148 |
> Perhaps... but maybe there is a need for some kind of a fair and |
149 |
> transparent mechanism how to decide which one gets to be in official |
150 |
> tree. |
151 |
I'm missing something here. The offical tree is only limited by what |
152 |
developers want to commit and there are lots of competing packages |
153 |
there already. KDE vs GNOME, any number of window managers, several |
154 |
Office applications, plaudis, portage (not sure about pkgcore) |
155 |
|
156 |
> It's great that Gentoo offers possibility to have many |
157 |
> different overlays and many different ways to do things but it's also |
158 |
> great that there is no need to fight with different repos. I mean - |
159 |
> it's not just about apps - all of these apps can be in official tree |
160 |
> - I thought more about overlays- how to make sure that the best stuff |
161 |
> gets to official tree - without fights and delays? |
162 |
The only requirement for a package to be in the offcial tree is that a |
163 |
Gentoo developer wants to put it there and maintain it. |
164 |
|
165 |
> Generally - it's good to have many possibilities and its good to have |
166 |
> an efficient mechanism how to combine as many of them as possible to |
167 |
> some kind of main/default tree for 'basic users' - where there are |
168 |
> compatibility issues solved, quality checked and no need to dig in to |
169 |
> choosing the right overlays and dealing with specific configuration. |
170 |
That exists - its the arch tree. |
171 |
There will always be some configuration, but as long as you select your |
172 |
profile, the system will build a working Gentoo for you. |
173 |
Overlays are never required for Gentoo to work. |
174 |
|
175 |
> And btw, Sunrise overlay - IMO absolutely great thing but so many |
176 |
> users don't even know it exists! |
177 |
How do we fix that? |
178 |
I think we tell users about overlays as they ask and the need arises. |
179 |
The quality of software in overlays varies considerably. |
180 |
|
181 |
Many new users come to gentoo as their first Linux. Many of them are |
182 |
poorly informed about both Linux and Gentoo - I don't know how to reach |
183 |
this group *before* they start asking for help in #gentoo and on the |
184 |
forums. Thats the sort problem we have to solve. |
185 |
|
186 |
> |
187 |
> And speaking about Paludis - I don't know how to write this - I |
188 |
> respect McCreesh, he is very talented, but I think Gentoo needs some |
189 |
> sort of organizational change that would let him stay inside Gentoo |
190 |
> without need for some other great developers to leave Gentoo - do you |
191 |
> know what I mean? I realize that I am in no position to criticise |
192 |
> anyone - anyone don't take me too seriously please. |
193 |
I know what you mean, other projects have similar issues. The kernel |
194 |
manages these issues by structuring to keep such groups/individuals |
195 |
apart, so they continue to contribute to their areas of interest. |
196 |
Its more of a problem to Gentoo because any developer may contribute |
197 |
anywhere. Gentoo has that open structure. |
198 |
|
199 |
> |
200 |
> > > To find the mechanism that would allow to maintain functionality |
201 |
> > > of Gentoo as whole, solve compatibility issues etc. without too |
202 |
> > > much of a huge organization that needs more and more energy to |
203 |
> > > keep itself going... writing summaries and attending meetings |
204 |
> > > while there is less and less time left to do the actual work - |
205 |
> > > that is the problem. |
206 |
|
207 |
> > Its a question of balance. Producing the information needed to |
208 |
> > keep users and other developers informed without doing to much, at |
209 |
> > the same time making sure that enough paperwork is produced to be |
210 |
> > able to use the products made by developers and understand the |
211 |
> > decisions they made at sometime in the future, when changes need to |
212 |
> > be made. |
213 |
> |
214 |
> Yes - but the trustees will have very difficult task achieving the |
215 |
> balance and in my humble opinion they will have to distinguish legal |
216 |
> stuff and the rest of their part of leading Gentoo and come up with |
217 |
> very different styles of management for each of them. |
218 |
I'm sure it won't be easy, if it were, it would not be fun. I think |
219 |
that we can learn from what others are already doing. In the Open |
220 |
Source world, its very easy to build on what others have already |
221 |
achieved - thats what the Open means. |
222 |
This reminds me of a quote attributed to Issac Newton "If I have seen |
223 |
further than other men, it is because I stand on the shouders of |
224 |
giants" |
225 |
Gentoo can build in the same way. |
226 |
|
227 |
> |
228 |
> Please consider that I am not an insider and I do realize that I can |
229 |
> be wrong with anything I say. I just thought it might be useful to |
230 |
> give you user's point of view. Great respect for developers. |
231 |
> |
232 |
> |
233 |
> Thanks a lot, Jan. |
234 |
> |
235 |
> -- |
236 |
> gentoo-nfp@l.g.o mailing list |
237 |
> |
238 |
Its always good to hear others points of view. Let me leave you with a |
239 |
link to my ideas for the Foundation |
240 |
http://dev.gentoo.org/~neddyseagoon/docs/manifesto.xml |
241 |
|
242 |
- -- |
243 |
Regards, |
244 |
|
245 |
Roy Bamford |
246 |
(NeddySeagoon) a member of |
247 |
gentoo-ops |
248 |
forum-mods |
249 |
treecleaners |
250 |
|
251 |
-----BEGIN PGP SIGNATURE----- |
252 |
Version: GnuPG v2.0.7 (GNU/Linux) |
253 |
|
254 |
iD8DBQFHpcZcTE4/y7nJvasRAsHtAKDOU3eVyUE1EkvLKLvotT1Ev+nJrgCeJy7o |
255 |
v8f5X034tbAA3xGtAlu6ppI= |
256 |
=pX/I |
257 |
-----END PGP SIGNATURE----- |
258 |
|
259 |
-- |
260 |
gentoo-nfp@l.g.o mailing list |