devblog
This commit is contained in:
parent
0226122842
commit
d25efeda6a
1 changed files with 56 additions and 0 deletions
56
doc/devblog/day_360__results_of_2015_user_survey.mdwn
Normal file
56
doc/devblog/day_360__results_of_2015_user_survey.mdwn
Normal file
|
@ -0,0 +1,56 @@
|
|||
The [2015 git-annex user servey](http://git-annex-survey.branchable.com/polls/2015/)
|
||||
is over with, and I'm reading through it and comparing with the
|
||||
[2013 survey](http://git-annex-survey.branchable.com/polls/2013/).
|
||||
|
||||
37% fewer users responded to the 2015 survey than in 2013.
|
||||
It's hard to tell if this has anything to do with the total number of
|
||||
git-annex users; Debian's popcon suggests the number of users has doubled
|
||||
since 2013, although its graph also suggests the number of users has
|
||||
flattened off since 2014. The difference may just be that I
|
||||
promoted the 2013 survey better than the 2015 survey, perhaps reaching
|
||||
kickstarter backers who I was in touch with back then.
|
||||
|
||||
25% use the assistant. Of those, 20% use XMPP, which is good to know as I'd
|
||||
like to get rid of it.
|
||||
|
||||
Android use has quardrupled, and Windows use has doubled; both are now at
|
||||
4%. It's not surprising that Android and Windows users still think more
|
||||
porting work is needed for those OSes. iOS is the only unsupported OS that
|
||||
more than 1% of users want. Embedded and NAS systems were mentioned much
|
||||
less than in 2013; probably the
|
||||
[[arm_tarball_build|forum/new_linux_arm_tarball_build]] met many such
|
||||
needs.
|
||||
|
||||
About the same percentage of users prefer direct mode in 2015 as did in 2013,
|
||||
and ditto for indirect mode. But, more users in 2015 only use direct mode
|
||||
on platforms that force its use. Correlating with the OS percentages
|
||||
suggests that many of these users are using removable media with the
|
||||
FAT filesystem, rather than an OS like Windows or Android. Hopefully
|
||||
v6 unlocked files will eventually better meet those user's needs.
|
||||
|
||||
The percent of users installing git-annex from source has halved since 2013,
|
||||
and it seems that builds from this website have taken up most of that slack;
|
||||
I would have expected more installs from Debian, Homebrew etc, but that
|
||||
seems not to have increased.
|
||||
|
||||
The number of repositories per user has gone up quite a lot since 2013,
|
||||
when only 7% of users had more than 10 repos. Now, 23% of users do. And,
|
||||
2% of users have more than 100 repos! This probably involves both more
|
||||
repositories for different purposes, and cloning of repositories
|
||||
to more devices.
|
||||
|
||||
Similarly, the amount of data stored has gone up. 34% have more than 1
|
||||
terabyte stored, up from 18% in 2013. 2% have more than 16 terabytes.
|
||||
|
||||
There's some indications of more users sharing repositories or
|
||||
otherwise using it in teams of larger groups, although most users still use
|
||||
it by themselves.
|
||||
|
||||
Users seem happier with git-annex now than in 2013. 16% call it "one of my
|
||||
favorite applications of all time". And, significantly fewer find it too
|
||||
hard to use than in 2013.
|
||||
|
||||
The main blocking problems are documentation, performance with many
|
||||
files (a general git problem), and various issues with the assistant.
|
||||
Respondants suggest more focus on making it easier for nontechnical users,
|
||||
and for use in larger groups/organizations.
|
Loading…
Reference in a new issue