1 |
>> There have been some really interesting points brought up recently |
2 |
>> about "where is Gentoo going?" I have been wondering that myself. |
3 |
|
4 |
I really appreciate that you had the initiative to start this conversation. |
5 |
I also agree that its a lingering problem that needs to be addressed |
6 |
clearly. |
7 |
|
8 |
>> Some people seem to think that Gentoo has the potential to be an |
9 |
>> enterprise player. I have not responded directly to those people, but |
10 |
>> I wonder if they know what they mean. I have worked in the enterprise |
11 |
>> UNIX market for 6 years. My code is running in places like NASA |
12 |
>> mission control, 9-1-1 call centers, and most of the telephone |
13 |
>> carriers. I've produced patches on weekends to close $800m deals. |
14 |
>> I now work in hp's Open Source and Linux Organization, mostly on RHEL |
15 |
>> and SLES, so I have a good idea of what it takes to be an enterprise |
16 |
>> player. |
17 |
>> |
18 |
|
19 |
I worked on NASA projects for several years and helped write the code |
20 |
to support a major scientific instrument onboard one research satellite(XTE), |
21 |
as well as contributing key sections of the health and safety, mission monitoring, |
22 |
and command generation subsystems associated with the science operating |
23 |
center for that satellite. I've also managed the infrastructure for several |
24 |
web hosting companies, even created my own startup which provided |
25 |
high-end systems administration services for DotComs which couldn't afford |
26 |
to ever go down. |
27 |
|
28 |
I am also a supporter of getting Gentoo into the enterprise. |
29 |
|
30 |
>> In my humble opinion, Gentoo is missing too many points to be an |
31 |
>> enterprise Linux. We commit to a live tree. We don't have true QA, |
32 |
>> testing or tinderbox. We don't have paid staff, alpha/beta/rc cycles. |
33 |
>> We don't really have product lifecycles, since we don't generally |
34 |
>> backport fixes to older versions, requiring instead for people to |
35 |
>> update to a more recent release. We don't have, and probably will |
36 |
>> never be able to offer, support contracts. We support as wide a range |
37 |
>> of hardware as the upstream kernel, plus hardware that requires |
38 |
>> external drivers; we don't have access to a great deal of the hardware |
39 |
>> for which we provide drivers. We understand when real life gets in |
40 |
>> the way of bug-fixing, because all our developers are volunteers. |
41 |
>> |
42 |
|
43 |
On the other hand, many of the other commercial distributions have their |
44 |
own problems. Redhat can generally be an expensive distribution |
45 |
to maintain in a server environment where applications are constantly |
46 |
changing and newer versions of open source applications are required |
47 |
to address performance or feature requirements. RPM hell still |
48 |
exists. I haven't played around with SUSE, but I wouldn't expect them |
49 |
to be any different. |
50 |
|
51 |
Furthermore, the cost of deploying commercial licenses can be |
52 |
cost prohibitive for small businesses with less than 25 servers |
53 |
who have an annual IT budget of $50K-$500K. These businesses |
54 |
still deserve a solidly reliable OS that is customized for them, but |
55 |
their budget needs to be allocated towards colocation facilities, |
56 |
bandwidth, and professional system administrators - not expensive |
57 |
software licenses and maintenance constracts. |
58 |
|
59 |
And, no, CentOS or Fedora are not an option for them currently. |
60 |
Fedora changes everything every six months or so, and support |
61 |
for old installs is questionable. Gentoo beats CentOS out of |
62 |
the water in terms of customizeability, community, and overall |
63 |
package availability. Plus RHEL, which CentOS is based on, |
64 |
has its own problems (setting up a simple server with it yesterday |
65 |
and having bonding enabled caused a kernel crash). |
66 |
|
67 |
A clueful sysadmin with gentoo is a far superior arrangement |
68 |
provided the rate of hardware installs isn't too much. For very large |
69 |
environments with 100+ boxes, I'd definitly agree with you that |
70 |
gentoo has a long way to go. |
71 |
|
72 |
Regarding your points on gentoo's weaknesses: |
73 |
Live Tree - not really a problem, thats what snapshots are for. |
74 |
QA Testing - not reallly a problem, thats what snapshots and a test server |
75 |
is for. |
76 |
Backporting of patches - A real concern, but I believe that gentoo is |
77 |
such a good solution that eventually some startup biz will address it. |
78 |
Support contracts - I'll take professional sysadmins over contracts |
79 |
any day, but this can also eventually be addressed by an eventual |
80 |
startup biz dedicated to it. I'd pay $10/month per box for it and could |
81 |
instantly refer 50-100 boxes assuming the solution was properly implemented |
82 |
and I could convince my clients to go with it. |
83 |
|
84 |
|
85 |
>> I think that attempting to take Gentoo in the "enterprise" direction |
86 |
>> is a mistake. I think that we are a hobbyist distribution. This |
87 |
>> doesn't mean that we should not strive to meet some of the enterprise |
88 |
>> goals. Those things can be important to hobbyists too. But I don't |
89 |
>> think we should be aiming for corporate America. |
90 |
>> |
91 |
|
92 |
I think Gentoo shouldn't rule out providing some support and flexibility |
93 |
for any need that a significant amount of its userbase is interested in. |
94 |
And, I know there are a significant number of devs already who |
95 |
have at least some interest in enterprise support do to conversations |
96 |
I've had via IRC. |
97 |
|
98 |
I definitly don't expect that the entire gentoo community or dev base |
99 |
should go substantially out of their way or change organizational structure |
100 |
to facilitate enterprise capabilities. Just allow some startup biz that |
101 |
eventually comes along to be able to provide a backported snapshot |
102 |
based tree for their own customers. |
103 |
|
104 |
>> I don't even understand why that goal appeals to people. Let other |
105 |
>> distros go there! I want Gentoo to run in people's homes, in student |
106 |
>> dorm rooms, etc. Places where people want a fun distribution that |
107 |
>> they can tailor and work on easily. |
108 |
>> |
109 |
|
110 |
Other distros suck in different ways. Gentoo definitly has enterprise |
111 |
advantages. |
112 |
|
113 |
>> I like the idea of Gentoo on alternative arches and in embedded |
114 |
>> environments. Not because I want Sony to start using Gentoo on |
115 |
>> walkmans, but purely because the idea of running Linux on a PDA is |
116 |
>> cool. I'd like Gentoo to be a place where neat things are developed. |
117 |
>> If RH or SuSE (or another for-profit Linux vendor) wants to take some |
118 |
>> of those developments and use them to make a profit, that's fine with |
119 |
>> me. We're over here having fun. |
120 |
>> |
121 |
|
122 |
|
123 |
RH has burned alot of people's trust. Some devs probably want nothing |
124 |
to do with them. |
125 |
|
126 |
>> Also I find it amusing when people say that Gentoo exists for the |
127 |
>> users. I think that is wrong. Gentoo exists for the *developers*. |
128 |
>> It's our playground, and it's the reason we use a live tree rather |
129 |
>> than switching to an actually sane approach. The users are cool |
130 |
>> because they point out bugs, help solve problems on bugzilla, suggest |
131 |
>> enhancements, provide patches, and notify us of package updates. |
132 |
>> Sometimes they become developers. But the truth is that Gentoo sees |
133 |
>> improvement and maintenance in the areas that appeal to the |
134 |
>> developers. And that is why Gentoo exists for the developers first, |
135 |
>> the users second. |
136 |
>> |
137 |
|
138 |
Thats part of the reason that I'm a developer - because I like interacting |
139 |
with the dev community here. But, the users have their own role and |
140 |
the above could be taken the wrong way. |
141 |
|
142 |
>> Regards, |
143 |
>> Aron |
144 |
|
145 |
MattM |
146 |
|
147 |
-- |
148 |
gentoo-dev@g.o mailing list |