1 |
On 17:38 Wed 14 Jan , Daniel Campbell wrote: |
2 |
> I'm not a developer, but I am interested in becoming one. That said, I |
3 |
> don't understand why relevance or focus are important. Is Gentoo not a |
4 |
> distribution made by people who feel that Gentoo meets their needs and |
5 |
> (for those who contribute) want to improve it? One of the core values |
6 |
> of Gentoo is the concept of choice. Without choice, none of the three |
7 |
> things you listed on the blog post are possible. |
8 |
|
9 |
Daniel, |
10 |
|
11 |
Thanks for your comments. Gentoo has been about choice where it makes |
12 |
sense and is pragmatic — not about choice for the sake of choice. This |
13 |
subtlety is often lost through the mists of time. But it drives other |
14 |
key tenets of Gentoo like packaging proprietary software, and all the |
15 |
weird restrictions that enable you to package things that can't be |
16 |
mirrored, etc. |
17 |
|
18 |
Here's one way to think of my proposal. Imagine a REST API. The provider |
19 |
of that API might also have a few SDKs for some popular languages or |
20 |
frameworks, along with documentation to make it really easy for them. |
21 |
But *the REST API is still there*, usable by anyone in any language. It |
22 |
just takes a bit more work. For Gentoo, this would work in much the same |
23 |
way — it essentially requires the existence of an underlying platform, |
24 |
but the focus would be on the use cases. |
25 |
|
26 |
There are two major problems that drive this. One is that Gentoo's |
27 |
popularity has greatly declined in the past decade. While you could |
28 |
argue that fewer users isn't necessarily a problem in its own right, |
29 |
it's creating severe difficulties for us in recruiting new developers. |
30 |
That is a problem, and it's an existential one. We need to change, |
31 |
because doing what we're doing will just keep us on the same road. |
32 |
|
33 |
Second, and this is one thing I've learned from my professional life, is |
34 |
that platforms are too ephemeral of a concept in isolation. Platforms |
35 |
(or metadistributions) succeed when it becomes crystal clear exactly how |
36 |
to use them, typically through creation of sample applications, early |
37 |
public customers, or something along those lines. |
38 |
|
39 |
> Maybe I'm misunderstanding you, but it seems like "focus" is referring |
40 |
> to actively marketing or advertising Gentoo to garner more interest. |
41 |
|
42 |
Focus includes but goes far beyond marketing. Here are some examples of |
43 |
what this kind of focus would imply: |
44 |
|
45 |
- The council would base its decision-making on these use cases |
46 |
- The QA project would help test that these use cases worked well |
47 |
- PMS enhancements would be driven in part by these use cases |
48 |
- Releng goals could be to provide image/package types to support them |
49 |
- Individual developers would make ebuild-level decisions based on them |
50 |
- We could use them as a rallying cry to unite Gentoo devs |
51 |
- And yes, the website would tell these stories to draw in new users |
52 |
|
53 |
> The real problem, based on what I've read here on gentoo-dev, |
54 |
> gentoo-project, and so on, is a lack of manpower. |
55 |
|
56 |
This is kind of a problem that isn't a problem. We've supposedly had a |
57 |
lack of manpower ever since Gentoo's founding, because there are more |
58 |
things that we could be doing but aren't. |
59 |
|
60 |
> The current issues I think Gentoo faces are more related to manpower |
61 |
> than relevance, focus, or usefulness. Maybe some new blood (to assist |
62 |
> the older, knowledgeable blood) is needed to help Gentoo move forward |
63 |
> faster. |
64 |
|
65 |
And where will that new blood come from? |
66 |
|
67 |
There's a limited number of ways to improve the conversion rate through |
68 |
our recruitment funnel. You can either lower the barriers along the way |
69 |
or pour more into the top of it. And while lowering the barriers would |
70 |
certainly get people their commit rights faster, I very rarely see |
71 |
people who give up halfway through because it's too frustrating. |
72 |
(Granted I haven't checked lately but I've been an on-and-off |
73 |
recruiter.) |
74 |
|
75 |
-- |
76 |
Thanks, |
77 |
Donnie |
78 |
|
79 |
Donnie Berkholz |
80 |
Council Member / Sr. Developer, Gentoo Linux <http://dberkholz.com> |
81 |
Analyst, RedMonk <http://redmonk.com/dberkholz/> |