1 |
On Mon, 2011-03-21 at 18:37 +0100, darkdefende@×××××.com wrote: |
2 |
> Fabian Groffen writes: |
3 |
> |
4 |
> > How hard do you think it would be to detect that you are actually |
5 |
> > dealing with one of those? |
6 |
> > |
7 |
> That depends on how much the build process differs from the standard way |
8 |
> of doing things. And as you said if it does a good job in most cases perhaps |
9 |
> that's good enough. |
10 |
> |
11 |
|
12 |
I think that most of the standard ways should be easily detected |
13 |
|
14 |
|
15 |
> > Why do you think a user knows? |
16 |
|
17 |
That user could be a dev for the app, but may not be a gentoo user. |
18 |
Just helping to create an ebuild to get his/her pkg into gentoo. |
19 |
|
20 |
Probably only if it is web based, and gets auto-submitted to a holding |
21 |
area that users/devs can review, download, etc. |
22 |
|
23 |
> > |
24 |
> Hmm, you have a good point there. If they already know then perhaps it would |
25 |
> be easier for them to write the ebuild by hand instead. |
26 |
> So a generator with minimal user input might be the way to go here. |
27 |
> |
28 |
> > I challenge you to think of a way to unpack a tarball, analyse its |
29 |
> > files and then conclude it's aparently a autoconf-based package, so |
30 |
> > configure, make and make install will likely be the way to do it. |
31 |
> > |
32 |
> If you wan't to be my mentor then I'll gladly accept your challenge and |
33 |
> look forward working with you! |
34 |
> |
35 |
|
36 |
Basic ability doesn't sound like much of a challenge to me. What would |
37 |
probably be best (and a challenge) is to have some kind of learning |
38 |
mechanism, that the generator could use to better refine it's detection |
39 |
and solutions. After all there is a whole tree, plus many overlays that |
40 |
can be run through as examples for it to learn from. (yes I know there |
41 |
are many that could/should be improved) |
42 |
|
43 |
There is an app called shogun-toolbox http://www.shogun-toolbox.org/ |
44 |
|
45 |
that is designed specifically for machine learning. It has many |
46 |
interface capabilities including python (which they state has the most |
47 |
complete documentation). There is also many examples for it's use. |
48 |
|
49 |
I personally don't know nearly enough about it to know what modeling |
50 |
methods might be suitable or even if there is one suitable for this type |
51 |
of model. But I think should be investigated. Implemented this would |
52 |
have the potential to increase it's success rate over time, reducing |
53 |
developer load even more. |
54 |
|
55 |
It is licensed GPL-3 |
56 |
|
57 |
|
58 |
> > This sounds more like making an example, not a use-case that describes |
59 |
> > how your technique should be used, and why it is really useful/much |
60 |
> > better than the current way of doing things. |
61 |
> > |
62 |
> Indeed. Now that I've thought about it perhaps your suggestion is the |
63 |
> way to go. :) |
64 |
> |
65 |
|
66 |
:) more ideas for thought |
67 |
-- |
68 |
Brian Dolbec <brian.dolbec@×××××.com> |