1 .. _managementstyle:
2
3 Linux kernel management style
4 =============================
5
6 This is a short document describing the prefer
7 on who you ask) management style for the linux
8 mirror the :ref:`process/coding-style.rst <cod
9 degree, and mainly written to avoid answering
10 questions over and over again.
11
12 Management style is very personal and much har
13 simple coding style rules, so this document ma
14 to do with reality. It started as a lark, but
15 might not actually be true. You'll have to dec
16
17 Btw, when talking about "kernel manager", it's
18 lead persons, not the people who do traditiona
19 companies. If you sign purchase orders or you
20 budget of your group, you're almost certainly
21 These suggestions may or may not apply to you.
22
23 First off, I'd suggest buying "Seven Habits of
24 People", and NOT read it. Burn it, it's a gre
25
26 .. [#f1] This document does so not so much by
27 making it painfully obvious to the questione
28 to what the answer is.
29
30 Anyway, here goes:
31
32 .. _decisions:
33
34 1) Decisions
35 ------------
36
37 Everybody thinks managers make decisions, and
38 important. The bigger and more painful the de
39 manager must be to make it. That's very deep
40 actually true.
41
42 The name of the game is to **avoid** having to
43 particular, if somebody tells you "choose (a)
44 to decide on this", you're in trouble as a man
45 manage had better know the details better than
46 you for a technical decision, you're screwed.
47 competent to make that decision for them.
48
49 (Corollary:if the people you manage don't know
50 you, you're also screwed, although for a total
51 Namely that you are in the wrong job, and that
52 your brilliance instead).
53
54 So the name of the game is to **avoid** decisi
55 painful ones. Making small and non-consequent
56 makes you look like you know what you're doing
57 needs to do is to turn the big and painful one
58 nobody really cares.
59
60 It helps to realize that the key difference be
61 small one is whether you can fix your decision
62 can be made small by just always making sure t
63 you **will** be wrong), you can always undo th
64 backtracking. Suddenly, you get to be doubly
65 **two** inconsequential decisions - the wrong
66
67 And people will even see that as true leadersh
68 *cough*).
69
70 Thus the key to avoiding big decisions becomes
71 things that can't be undone. Don't get ushere
72 you cannot escape. A cornered rat may be dang
73 is just pitiful.
74
75 It turns out that since nobody would be stupid
76 a kernel manager have huge fiscal responsibili
77 fairly easy to backtrack. Since you're not go
78 huge amounts of money that you might not be ab
79 thing you can backtrack on is a technical deci
80 back-tracking is very easy: just tell everybod
81 incompetent nincompoop, say you're sorry, and
82 work you had people work on for the last year.
83 you made a year ago wasn't a big decision afte
84 easily undone.
85
86 It turns out that some people have trouble wit
87 reasons:
88
89 - admitting you were an idiot is harder than
90 maintain appearances, and coming out in pub
91 wrong is sometimes very hard indeed.
92 - having somebody tell you that what you work
93 wasn't worthwhile after all can be hard on
94 too, and while the actual **work** was easy
95 deleting it, you may have irrevocably lost
96 engineer. And remember: "irrevocable" was
97 the first place, and your decision ended up
98 all.
99
100 Happily, both of these reasons can be mitigate
101 admitting up-front that you don't have a frigg
102 people ahead of the fact that your decision is
103 might be the wrong thing. You should always r
104 your mind, and make people very **aware** of t
105 to admit that you are stupid when you haven't
106 stupid thing.
107
108 Then, when it really does turn out to be stupi
109 eyes and say "Oops, not again".
110
111 This preemptive admission of incompetence migh
112 actually do the work also think twice about wh
113 not. After all, if **they** aren't certain wh
114 sure as hell shouldn't encourage them by promi
115 work on will be included. Make them at least
116 embark on a big endeavor.
117
118 Remember: they'd better know more about the de
119 they usually already think they have the answe
120 thing you can do as a manager is not to instil
121 healthy dose of critical thinking on what they
122
123 Btw, another way to avoid a decision is to pla
124 we just do both?" and look pitiful. Trust me,
125 clear which approach is better, they'll eventu
126 answer may end up being that both teams get so
127 situation that they just give up.
128
129 That may sound like a failure, but it's usuall
130 something wrong with both projects, and the re
131 couldn't decide was that they were both wrong.
132 smelling like roses, and you avoided yet anoth
133 have screwed up on.
134
135
136 2) People
137 ---------
138
139 Most people are idiots, and being a manager me
140 with it, and perhaps more importantly, that **
141 **you**.
142
143 It turns out that while it's easy to undo tech
144 as easy to undo personality disorders. You ju
145 theirs - and yours.
146
147 However, in order to prepare yourself as a ker
148 remember not to burn any bridges, bomb any inn
149 alienate too many kernel developers. It turns
150 is fairly easy, and un-alienating them is hard
151 immediately falls under the heading of "not re
152 no-no according to :ref:`decisions`.
153
154 There's just a few simple rules here:
155
156 (1) don't call people d*ckheads (at least not
157 (2) learn how to apologize when you forgot ru
158
159 The problem with #1 is that it's very easy to
160 "you're a d*ckhead" in millions of different w
161 even realizing it, and almost always with a wh
162 you are right.
163
164 And the more convinced you are that you are ri
165 you can call just about **anybody** a d*ckhead
166 right), the harder it ends up being to apologi
167
168 To solve this problem, you really only have tw
169
170 - get really good at apologies
171 - spread the "love" out so evenly that nobody
172 like they get unfairly targeted. Make it i
173 might even be amused.
174
175 The option of being unfailingly polite really
176 trust somebody who is so clearly hiding their
177
178 .. [#f2] Paul Simon sang "Fifty Ways to Leave
179 frankly, "A Million Ways to Tell a Developer
180 scan nearly as well. But I'm sure he though
181
182
183 3) People II - the Good Kind
184 ----------------------------
185
186 While it turns out that most people are idiots
187 sadly that you are one too, and that while we
188 knowledge that we're better than the average p
189 nobody ever believes that they're average or b
190 also admit that we're not the sharpest knife a
191 other people that are less of an idiot than yo
192
193 Some people react badly to smart people. Othe
194
195 Make sure that you, as a kernel maintainer, ar
196 Suck up to them, because they are the people w
197 easier. In particular, they'll be able to make
198 which is what the game is all about.
199
200 So when you find somebody smarter than you are
201 management responsibilities largely become one
202 good idea - go wild", or "That sounds good, bu
203 second version in particular is a great way to
204 new about "xxx" or seem **extra** managerial b
205 smarter person hadn't thought about. In eithe
206
207 One thing to look out for is to realize that g
208 not necessarily translate to other areas. So
209 specific directions, but let's face it, they m
210 do, and suck at everything else. The good new
211 naturally gravitate back to what they are good
212 are doing something irreversible when you **do
213 direction, just don't push too hard.
214
215
216 4) Placing blame
217 ----------------
218
219 Things will go wrong, and people want somebody
220
221 It's not actually that hard to accept the blam
222 kind of realize that it wasn't **all** your fa
223 best way of taking the blame: do it for someon
224 for taking the fall, they'll feel good about n
225 person who lost their whole 36GB porn-collecti
226 incompetence will grudgingly admit that you at
227 out of it.
228
229 Then make the developer who really screwed up
230 **in private** that they screwed up. Not just
231 future, but so that they know they owe you one
232 importantly, they're also likely the person wh
233 face it, it sure ain't you.
234
235 Taking the blame is also why you get to be man
236 It's part of what makes people trust you, and
237 glory, because you're the one who gets to say
238 you've followed the previous rules, you'll be
239 by now.
240
241
242 5) Things to avoid
243 ------------------
244
245 There's one thing people hate even more than b
246 and that is being called a "d*ckhead" in a san
247 first you can apologize for, the second one yo
248 chance. They likely will no longer be listeni
249 do a good job.
250
251 We all think we're better than anybody else, w
252 somebody else puts on airs, it **really** rubs
253 be morally and intellectually superior to ever
254 don't try to make it too obvious unless you re
255 somebody [#f3]_.
256
257 Similarly, don't be too polite or subtle about
258 ends up going overboard and hiding the problem
259 internet, nobody can hear you being subtle". U
260 hammer the point in, because you can't really
261 your point otherwise.
262
263 Some humor can help pad both the bluntness and
264 overboard to the point of being ridiculous can
265 without making it painful to the recipient, wh
266 silly. It can thus help get through the perso
267 have about criticism.
268
269 .. [#f3] Hint: internet newsgroups that are no
270 are great ways to take out your frustrations
271 insulting posts with a sneer just to get int
272 a while, and you'll feel cleansed. Just don'
273
274
275 6) Why me?
276 ----------
277
278 Since your main responsibility seems to be to
279 peoples mistakes, and make it painfully obviou
280 you're incompetent, the obvious question becom
281 first place?
282
283 First off, while you may or may not get scream
284 boys, let's not be judgmental or sexist here)
285 room door, you **will** get an immense feeling
286 for being "in charge". Never mind the fact th
287 by trying to keep up with everybody else and r
288 as you can. Everybody will still think you're
289
290 It's a great job if you can hack it.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.