| 1 |
Kurt Lieber wrote: |
| 2 |
> How about if we posted a question on survey.gentoo.org 3-6 months after the |
| 3 |
> service has been up that simply asks, "Do you find planet.gentoo.org to be |
| 4 |
> valuable?" If the "yes" rate is something less than X% (I'd suggest 33%) |
| 5 |
> then we kill it. If not, we leave it up and consider it a success. This |
| 6 |
> seems like a very fair, impartial way of gauging the impact that this |
| 7 |
> service has had on our community. |
| 8 |
|
| 9 |
I am much keener on this idea, thanks for the suggestion! What I'm going to |
| 10 |
propose in the next GLEP revision is: |
| 11 |
|
| 12 |
1. user survey is taken (as you describe) on planet.gentoo.org, with question |
| 13 |
along the lines of "Do you find Planet Gentoo to be valuable?" - answers being |
| 14 |
"Yes", "No", or "Never read it". |
| 15 |
|
| 16 |
2. developer survey is taken with similar question. this has also been |
| 17 |
discussed in this thread, but i think its important that the developer survey |
| 18 |
is done independantly. if the developer and user results are combined, then |
| 19 |
the views of the developer minority won't appear in the stats. (example: of |
| 20 |
the people who took the survey, 30 out of 30 developers say yes its useful, |
| 21 |
and 5000 out of 5000 users said no its not. the overall figures will say its |
| 22 |
not useful, but infact the view of the developer community is the opposite). |
| 23 |
|
| 24 |
3. results are somehow evaluated against our criteria. this is what i'm not |
| 25 |
sure about. Any suggestions how we would do this? And as Kurt says, we do need |
| 26 |
some criteria that we can decide upon now rather than later. (zhen, input |
| 27 |
would be appreciated here!) |
| 28 |
|
| 29 |
Thanks, |
| 30 |
Daniel |
| 31 |
|
| 32 |
-- |
| 33 |
gentoo-dev@g.o mailing list |
Archives