GitSurvey2010

From Git SCM Wiki
Jump to: navigation, search

OBSOLETE CONTENT

This wiki has been archived and the content is no longer updated. Please visit git-scm.com/doc for up-to-date documentation.

Git User's Survey 2010 summary

The "Git User's Survey 2010" has been closed on October 15, 2010.

The survey could be found here:

The survey was be open from 1 September to 15 October, 2010 (for 1.5 months).

You can get raw data (individual responses) for the survey here (exported from Survs.com):

You can use survey_parse_Survs_CSV(num).com.perl script from http://github.com/jnareb/GitSurvey-scripts repository to analyze data from CSV export linked above.

The live results of the survey can be found at:


There were 8841 individual responses in this survey (the 5 responses in 'test' channel were removed), for 28001 views.
There were 3868 individual responses in GitSurvey2009 (the 30 responses in 'test' channel were removed), for 14106 views.
There were 3236 individual responses in GitSurvey2008 (including 21 responses in 'test' channel), for 12101 views.
There were 683 individual responses in GitSurvey2007.
There were around 117 responses in GitSurvey2006.



Table of contents:

Contents



Information about survey completion

Date of response

Histogram of number of responders by date


The graph above shows histogram of number of responders by date (in CET / GMT+2 / +0200 timezone).


Number of replies per question

Question Count Perc. per 8841 responses
About you
01. What country do you live in (country of residence)? 8470 95.8%
 
02. How old are you (in years)? 8437 95.4%
 
Getting started with Git
03. Have you found Git easy to learn? 8404 95.1%
 
04. Have you found Git easy to use? 8418 95.2%
 
05. Which Git version(s) are you using? 8481 95.9%
 
06. Rate your own proficiency with Git: 8546 96.7%
 
How you use Git
07. I use Git for (check all that apply): 8551 96.7%
 
08. How do/did you obtain Git (install and/or upgrade)? 8505 96.2%
 
09. On which operating system(s) do you use Git? 8541 96.6%
 
10. What Git interfaces, implementations and frontends do you use? 8467 95.8%
 
11. How often do you use following kinds of Git tools? 8429 95.3%
 
12. What Git GUIs (graphical user interfaces) do you use? 6919 78.3%
 
13. Which git hosting site(s) do you use for your project(s)? 7922 89.6%
 
14. How do you fetch/get changes from upstream repositories? 8354 94.5%
 
15. How do you publish/propagate your changes? 8272 93.6%
 
16. Which of the following features do you use? 7087 80.2%
 
17. Which of the following features would you like to see implemented in git? 7203 81.5%
 
18. Describe what features would you like to have in Git, 1187 13.4%
 
What you think of Git
19. Overall, how happy are you with Git? 8471 95.8%
 
20. In your opinion, which areas in Git need improvement? 8107 91.7%
 
Changes in Git (since year ago, or since you started using it)
21. Did you participate in previous Git User's Surveys? 8332 94.2%
 
22. How do you compare the current Git version with the version from one year ago? 8202 92.8%
 
Documentation. Getting help.
23. How useful have you found the following forms of Git documentation? 8240 93.2%
 
24. Have you tried to get help regarding Git from other people? 8294 93.8%
 
25. If yes, did you get these problems resolved quickly and to your liking? 5846 66.1%
 
26. What channel(s) did you use to request help? 5777 65.3%
 
27. Which communication channel(s) do you use? 1441 16.3%
 
About this survey.
28. How did you hear about this Git User's Survey? 8288 93.7%
 

Legend:

  free-form oneline
  free-form essay
  single choice
  multiple choice
  matrix


Number of questions answered per responder

Histogram of the number of answered questions by responder

Most people answered 26 questions, all but 2 questions out of 28 in total.

About you

01. What country do you live in (country of residence)?

(tabularized free-form single line)


Geomap of responders per country

This table is sorted alphabetically by country name.

The countries where number of responders was less than 10 (less than 0.1%) were gathered together and summed under "other countries" label (this applies both to the table below, and to the geomap visualization above).

Some responders put more than one country in response, or wrote 'multiple' (perhaps double citizenship?). There were 8 such responses, as can be seen in the row named "multiple".

Europe
 
North America
 
Asia
 

Country Count Perc.
Argentina 59 0.7%
Australia 283 3.3%
Austria 85 1.0%
Belarus 15 0.2%
Belgium 94 1.1%
Brazil 191 2.3%
Bulgaria 20 0.2%
Canada 356 4.2%
Chile 11 0.1%
China 70 0.8%
Colombia 13 0.2%
Croatia 12 0.1%
Czech Republic 196 2.3%
Denmark 65 0.8%
Estonia 22 0.3%
Finland 111 1.3%
France 518 6.1%
Germany 876 10.3%
Greece 33 0.4%
Hungary 51 0.6%
India 109 1.3%
Indonesia 16 0.2%
Ireland 37 0.4%
Israel 42 0.5%
Italy 120 1.4%
Japan 85 1.0%
Lithuania 19 0.2%
Mexico 22 0.3%
Netherlands 182 2.1%
New Zealand 86 1.0%
Norway 137 1.6%
Philippines 14 0.2%
Poland 169 2.0%
Portugal 29 0.3%
Romania 44 0.5%
Russian Federation 162 1.9%
Serbia 11 0.1%
Singapore 14 0.2%
Slovakia 27 0.3%
Slovenia 13 0.2%
South Africa 39 0.5%
Spain 100 1.2%
Sweden 302 3.6%
Switzerland 121 1.4%
Taiwan 15 0.2%
Thailand 18 0.2%
Turkey 16 0.2%
Ukraine 61 0.7%
United Kingdom 585 6.9%
United States 2655 31.3%
other countries (below 10 resp.) 136 1.6%
multiple 8 0.0%
Invalid answer 5 0.0%
Base 8470 / 8841


Analysis:
As one can easily see, slightly less than third of Git users (31.3%) are in the USA (those who answered this survey, and this question). Next of countries is Germany with around 10.3% (or 1/10th) responses. Third is United Kingdom of Great Britain (including England, Scotland, Wales and Northern Ireland) with 6.9%, and closely following on fourth place is France with 6.1%

The names of the countries were normalized using Locale::Country Perl module version 3.14

You can also take a look at Git Activity Map at Ohloh, where you can see on (large) Google Map locations of selected people who have Git in a stack, or locations of Git contributors (those with known location).


It is a bit pity that this survey did not include question "Does git.git repository include code produced by you?" (or equivalent), so we would be able to see where the developers are.


Continent Count Perc.
Africa 52 0.6%
 
Asia 441 5.2%
 
Europe 4232 50.0%
 
North America 3049 36.0%
 
South America 288 3.4%
 
Oceania 370 4.4%
 

In the table above "Oceania" includes (among others) Australia and New Zealand. As you can see most responders are from Europe (50.0%, a half of responses), followed by North America (36.0% responders).

The conversion from country name to continent was done using Locale::Object::Country Perl module from CPAN (version 0.78), which in turn uses data from http://www.worldatlas.com/cntycont.htm

02. How old are you (in years)?

(tabularized free-form single line)

Histogram of age of responders


Age Count Perc.
 < 18 147 1.7%
 
18-21 711 8.4%
 
22-25 1970 23.3%
 
26-30 2645 31.4%
 
31-40 2293 27.2%
 
41-50 514 6.1%
 
51-75 144 1.7%
 
76+ 7 0.1%
 
Not a number 6 0.1%
 
Base 8437 / 8841


Analysis:
Discounting joke responses (or errors in parsing), like 3 or 4 years old on one hand side, and 1000000 years old on other side, youngest user who answered this question is 10 years old (next oldest is 12 years old), oldest is 87 (next to oldest is 81). That is quite a span.

Most Git users (slightly less than one third: 31.4%) fall in the 26-30 range. The age of 25 got most responses (616 for 8437 responses, or 7.3%).

Getting started with Git

03. Have you found Git easy to learn?

(single choice)

Answer Count Perc.  
Very easy 408 4.9%
 
Easy 1984 23.6%
 
Reasonably easy 4460 53.1%
 
Hard 1413 16.8%
 
Very hard 139 1.7%
 
Base 8404 / 8841

Nice Gaussian curve. Most users find it reasonably easy to learn, slightly biased towards easy.

04. Have you found Git easy to use?

(single choice)

Answer Count Perc.  
Very easy 932 11.1%
 
Easy 3168 37.6%
 
Reasonably easy 3575 42.5%
 
Hard 658 7.8%
 
Very hard 85 1.0%
 
Base 8418 / 8841

Analysis:
Most users find it reasonably easy to easy to use Git.

When analyzing this data (and the data for the question before) you should take into account that somebody considering or finding Git too hard to learn wouldn't be probably Git user, and thus wouldn't fill this survey. So because it is Git User's Survey we should consider that results can be skewed towards lower value (easier).

What's interesting is comparing (percentage) results for questions 3. and 4.; how hard is git to learn versus how hard is to use. It seems like Git is reasonably easy to learn, and reasonably easy to easy to use. So it looks like Git just have somewhat steep learning curve, and the difficulty to learn pays in being more powerful to use.


Correlations:
Below there is table of correlations between question "04. Have you found Git easy to use?" and "03. Have you found Git easy to learn?"

Bubble plot of table of correlations between questions '03. Have you found Git easy to learn?' and '04. Have you found Git easy to use?'
Very easy Easy Reasonable Hard Very hard
Very easy 3.2%
(5.933)
1.3%
(0.717)
0.3%
(0.152)
0.0%
(0.000)
0.0%
(0.489)
Easy 4.4%
(1.678)
14.1%
(1.585)
4.9%
(0.491))
0.2%
(0.097)
0.0%
(0.000)
Reasonably easy 3.2%
(0.536)
19.3%
(0.966)
28.2%
(1.255)
2.3%
(0.554)
0.1%
(0.111)
Hard 0.3%
(0.185)
2.9%
(0.455)
8.7%
(1.218)
4.6%
(3.490)
0.3%
(1.955)
Very hard 0.0%
(0.130)
0.1%
(0.096)
0.2%
(0.325)
0.7%
(5.758)
0.6%
(35.673)
Base 8369 / 8841


Percentages in the table above are of number of responders that selected given answer to both questions. For example 19.3% in third row, second column means that out of responders who answered both questions 19.3% found Git "reasonably easy" to learn, while "easy" to use.

The small number in parentheses below percentage is ratio of said percentage to the percentage as if those answers were probabilistically independent.

In the bubble plot on the right hand side of table it is the area of bubble that represents correlation (number of responders that selected given answers)

  • Bubble-corr gt xy.png means that correlation is greater than for independent answers
  • Bubble-corr lt xy.png means that correlation is smaller than for independent answers


Different correlation coefficients

  • Pearson's r = 0.55 [-1, 1] (linear dependence)
  • Spearman's ρ = 0.52 [-1, 1] (monotonic dependence)
  • Kendall's τb = 0.48 [-1, 1] (rank correlation)

show that answers to those questions are correlated quite well; if one finds Git easy to use, he/she would find it easy to use.

Pearson's coefficient was calculated using least_squares_fit from Statistics::Descriptive Perl module,
while both Spearman's and Kendall's coeficients were calculated using Statistics::RankCorrelation

05. Which Git version(s) are you using?

(multiple choice, with other)

Version used Count Perc.  
pre 1.3 12 0.1%
 
1.3.x 16 0.2%
 
1.4.x 39 0.5%
 
1.5.x 530 6.2%
 
1.6.x 2009 23.7%
 
1.7.x 7024 82.8%
 
minor (maintenance) release 1.x.y.z 1463 17.3%
 
'master' branch of official git repository 304 3.6%
 
'next' branch of official git repository 54 0.6%
 
'pu' branch of official git repository 13 0.2%
 
version from msysGit / Git for Windows fork repository 811 9.6%
 
git-bigfiles fork 12 0.1%
 
JGit implementation (e.g. via Eclipse plugin, EGit) 404 4.8%
 
other git implementation (Git#, Dulwich, etc.) 75 0.9%
 
other, please specify 157 1.9%
 
Base 8481 / 8841


Description:
You can find git version by using "git --version" or "git version".

"Minor release" is additional specification, so if you for example use git version 1.6.3.3, please check both "1.6.x" and "minor release".


Analysis:
As you can see from above results great majority (83%) run latest 1.7.x series of git (note: because it is multiple-choice question it means 'at least on one machine'). Only a few use 1.3.x or pre 1.3 versions somewhere. The 1.4.x version gets only 0.5%.

Only around 17.3% (less than one fifth) of responders use (somewhere) maintenance releases (1.x.y.z). That, or the question wasn't formulated clear enough.

06. Rate your own proficiency with Git:

(single choice)

Answer Count Perc.  
1. novice 341 4.0%
 
2. casual, needs advice 1625 19.0%
 
3. everyday use 3308 38.7%
 
4. can offer advice 2703 31.6%
 
5. know it very well 569 6.7%
 
Base 8546 / 8841


Description:
You can think of it as 1-5 numerical grade of your proficiency in Git.

Analysis:
As you can see most responders know Git enough for everyday use (39%), or can even offer some Git advice (32%). If we treat possible answers as a proficiency grade, the average proficiency is around 3.2 ± 0.9.

Either Git users don't stay novices long (4%), or survey announcement(s) didn't reach many novice users.

Correlations:
Below there is table of correlations to "04. Have you found Git easy to use?".

Bubble plot of table of correlations between questions '06. Rate your own proficiency with Git' and '04. Have you found Git easy to use?'
Very easy Easy Reasonable Hard Very hard
1. novice 0.2%
(0.410)
1.0%
(0.642)
1.8%
(1.052)
0.9%
(2.861)
0.2%
(4.227)
2. casual, needs advice 1.3%
(0.624)
5.9%
(0.820)
9.1%
(1.128)
2.3%
(1.557)
0.4%
(2.130)
3. everyday use 4.5%
(1.048)
14.3%
(0.982)
16.6%
(1.011)
3.0%
(0.981)
0.3%
(0.831)
4. can offer advice 3.9%
(1.115)
13.5%
(1.131)
12.8%
(0.950)
1.4%
(0.567)
0.1%
(0.300)
5. know it very well 1.2%
(1.599)
3.0%
(1.205)
2.2%
(0.777)
0.3%
(0.480)
0.0%
(0.178)
Base 8395 / 8841


Percentages in the table above are of number of responders that selected given answer to both questions, relative to number of responders that answered both questions (e.g. the 1.0% in first row, second column of above table means that those responders find Git "easy" to use, while thinking about themselves as being "novice" with Git).

The small number in parentheses below percentage (e.g. the "(0.642)" in in first row, second column of above table) is ratio of responders who selected given answers to both question to the product of number of responsers that chose given answer to first question with number of responders that chose given answer to second question. This means that value of 1.000 means that the correlation is the same as if distributions were independent: correlation would be the product of distributions in those two questions.

One can see from the percentages that the center of above matrix is in 'easy' to 'reasonable' and 'everyday use' to 'can offer advice'. The percentages together with correlation ratios shows that answers to those questions are somewhat anti-correlated: more novice users find git harder to use (which is easy to understand). Though it is more about "novice" users finding it "very hard" to use Git (correlation ratio of 4.227, though only 0.2% of responders) and "know it very well" not finding it "very hard" to use (correlation ratio of 0.178, and below 0.05% of responses), rather than "know it very well" finding it "very easy" to use (correlation ratio of about 1.6, with 1.2% of responders).

The correlations coefficient support this observation of slight correlation between answers to those questions:

  • Pearson's r = -0.19 [-1, 1] (linear dependence)
  • Spearman's ρ = -0.18 [-1, 1] (monotonic dependence)
  • Kendall's τb = -0.16 [-1, 1] (rank correlation)

Pearson's coefficient was calculated using least_squares_fit from Statistics::Descriptive Perl module, while both Spearman's and Kendall's coeficients were calculated using Statistics::RankCorrelation

In the bubble plot on the right hand side of table it is the area of bubble that represents correlation (number of responders that selected given answers).


Bubble plot of table of correlations between questions '06. Rate your own proficiency with Git' and '03. Have you found Git easy to learn?'
Very easy Easy Reasonable Hard Very hard
1. novice 0.1%
(0.626)
0.7%
(0.703)
1.9%
(0.885)
1.0%
(1.538)
0.3%
(4.537)
2. casual, needs advice 0.8%
(0.820)
3.7%
(0.829)
10.0%
(0.991)
4.0%
(1.250)
0.5%
(1.699)
3. everyday use 2.0%
(1.091)
9.5%
(1.044)
20.5%
(0.998)
6.0%
(0.925)
0.6%
(0.930)
4. can offer advice 1.4%
(0.933)
7.9%
(1.059)
17.4%
(1.036)
4.8%
(0.891)
0.2%
(0.318)
5. know it very well 0.5%
(1.532)
1.8%
(1.129)
3.3%
(0.935)
1.0%
(0.918)
0.1%
(0.542)
Base 8375 / 8841


Examining correlations between question "06. Rate your own proficiency with Git:" and "03. Have you found Git easy to learn?" we see that answers to those questions are almost completly unrelated.

  • Pearson's r = -0.10 [-1, 1] (linear dependence)
  • Spearman's ρ = -0.09 [-1, 1] (monotonic dependence)
  • Kendall's τb = -0.08 [-1, 1] (rank correlation)

The very faint hint of anti-correlation might be caused by the fact that people for which Git is very hard to learn do not become Git guru's (and do not know it very well).

How you use Git

07. I use Git for (check all that apply):

(multiple choice, with other)

Answer Count Perc.
work projects 6905 80.8%
 
unpaid projects 6623 77.5%
 
proprietary projects 3365 39.4%
 
OSS development 5556 65.0%
 
private (unpublished) code 6335 74.1%
 
code (programming) 7819 91.4%
 
personal data 2337 27.3%
 
documents (data) 2528 29.6%
 
static website 2477 29.0%
 
web app 3892 45.5%
 
large (>1 MB) binary files 1484 17.4%
 
often changing binary files 611 7.1%
 
sharing data or sync 1724 20.2%
 
managing configuration files 2850 33.3%
 
backup 1664 19.5%
 
backend for wiki, blog, or other web app 922 10.8%
 
frontend to other SCM (e.g. git-svn) 1809 21.2%
 
other (please specify) 154 1.8%
 
Base 8551 / 8841


Description:
Note that above choices are neither orthogonal nor exclusive. You might want to check multiple answers even for a single repository.

08. How do/did you obtain Git (install and/or upgrade)?

(multiple choice, with other)

Answer Count Perc.
binary package 6275 73.8%
 
source package or script 2004 23.6%
 
source tarball 1211 14.2%
 
pull from (main) repository 984 11.6%
 
preinstalled / sysadmin job 789 9.3%
 
other - please specify (if none of the above apply) 414 4.9%
 
Base 8505 / 8841


Description:
Explanation: "binary package" covers pre-compiled binary (e.g. from rpm or deb binary packages); "source package" covers things like deb-src and SRPMS/*.src.rpm; "source script" is meant to cover installation in source-based distributions, like 'emerge' in Gentoo.

Automatic update (apt, yum, etc.) in most cases means binary package install; unless one uses source-based distribution like Gentoo, CRUX, or SourceMage, where automatic update means using source package (or source script).

The option named "preinstalled / sysadmin job" means that either you didn't need to install git because it was preinstalled (and you didn't upgrade); or that you have to ask system administrator to have git installed or upgraded.


Note that this question is multiple choices question because one can install Git in different ways on different machines or on different operating systems.

Analysis:
Most people (73.8%) use ready binary packages, which was kind of expected; that is the easiest way.

09. On which operating system(s) do you use Git?

(multiple choice, with other)

Operating System Count Perc.
GNU/Linux 7243 84.8%
 
*BSD (FreeBSD, OpenBSD, NetBSD, etc.) 639 7.5%
 
MacOS X (Darwin) 3932 46.0%
 
MS Windows (any)[1] 2671 31.3%
 
 MS Windows/Cygwin 823 9.6%
 
 MS Windows/msysGit (MINGW) 2151 25.2%
 
other Unix 211 2.5%
 
Other, please specify 92 1.1%
 
Base 8541 / 8841

Notes:

  1. Number of responders that selected "MS Windows/Cygwin", or "MS Windows/msysGit (MINGW)", or both

Description:
On Unix-based operating system you can get the name of operation system by running 'uname' (or 'uname -o').

Analysis:
Most common used operating system is GNU/Linux, next is MacOS X, and then MS Windows (on MS Window dominates msysGit version). This is quite understandable, as Git was created on Linux and for Linux, and it works best on POSIX-compliant operating systems.

Analysis of "other, please specify" responses:
Below there is table with analysis of write-in responses in "other".

Operating System Count Perc.
EXPLANATION 24 0.3%
GNU/Linux 9 0.1%
GNU/Linux distribution 17 0.2%
GNU/Linux (it was "Linux", not "GNU/Linux") 4 0.0%
GNU/Hurd 2? 0.0%
GNU/kFreeBSD 1 0.0%
*BSD (FreeBSD, OpenBSD, NetBSD, etc.) 1 0.0%
MacOS X (Darwin) 2 0.0%
MacOX X (Homebrew) 1 0.0%
MacOS 9/Lamp 1 0.0%
MS Windows/unknown 8 0.1%
MS Windows (without Cygwin/msysGit) 1 0.0%
MS Windows msysGit from Cygwin (?) 1 0.0%
MS Windows/GitExtensions 1 0.0%
MS Windows/TortoiseGit 6 0.1%
MS Windows/GitSharp 1 0.0%
MS Windows/JGit 6 0.1%
MS Windows/virtual machine with GNU/Linux 3 0.0%
JGit 9 0.1%
Emacs 1 0.0%
Haiku (BeOS) 2 0.0%
OpenVMS 1 0.0%
Maemo 1 0.0%
iPhone OS (jailbroken) 2 0.0%
other Unix 14 0.2%
   AIX 1 0.0%
   HP-UX 1 0.0%
   IRIX 1 0.0%
   SunOS 6 0.1%
   Solaris 24 0.3%
     Joyent 2 0.0%
     OpenIndiana 2 0.0%
     OpenSolaris 12 0.1%
comment 2 0.0%
comment about survey 1 0.0%
don't know 1 0.0%

Here "EXPLANATION" means that code parsing and tabularizing "other, please specify" response found category which was already marked (used). This most probably meant that responder used "other, please specify" field to explain / describe in more detail their response.

'GNU/Linux (it was "Linux", not "GNU/Linux")' response means people who used "other, please specify" response to tell that it should be 'GNU/Linux' (it was before change of answer from 'Linux' to 'GNU/Linux').

Quite a few responders knew that they use e.g. Ubuntu, but didn't choose 'GNU/Linux' response...

10. What Git interfaces, implementations and frontends do you use?

(multiple choice, with other)

Answer Count Perc.
git (core) commandline 8285 97.9%
 
JGit (Java implementation), e.g. via Eclipse plugin 745 8.8%
 
Cogito (DEPRECATED) 7 0.1%
 
Easy Git 33 0.4%
 
Pyrite 7 0.1%
 
StGIT 84 1.0%
 
Guilt 14 0.2%
 
TopGit 34 0.4%
 
pg aka Patchy Git (DEPRECATED) 1 0.0%
 
I don't know, I use only (graphical) tools 86 1.0%
 
Other (please specify) 1706 20.1%
 
Base 8467 / 8841

Analysis:
Using (at least among other implementations) git command line tool dominates with 97.9% responses... but there are responders (2.1%) who do not use command line tools at all. JGit users have almost one tenth of responses (8.8%).

This question has very large amount of "other (please specify)" responses, 20.1% which is around fifth of responders.

11. How often do you use following kinds of Git tools?

(matrix)

Answer never rarely sometimes often (almost) always Avg. / Count  
editor/IDE VC integration[1] 4792 56.9% 1457 17.3% 951 11.3% 737 8.7% 492 5.8% 1.9 (5) / 8429
 
 
 
 
 
filemanager / shell extension[2] 6477 76.8% 804 9.5% 501 5.9% 325 3.9% 237 2.8% 1.4 (5) / 8344
 
 
 
 
 
graphical history viewer/browser[3] 1794 21.3% 1496 17.7% 1873 22.2% 1912 22.7% 1183 14.0% 2.9 (5) / 8258
 
 
 
 
 
graphical commit tool[4] 4710 55.9% 1265 15.0% 785 9.3% 757 9.0% 678 8.0% 2.0 (5) / 8195
 
 
 
 
 
graphical diff and/or merge tool[5] 3551 42.1% 1301 15.4% 1391 16.5% 1230 14.6% 673 8.0% 2.3 (5) / 8146
 
 
 
 
 
graphical blame or pickaxe tool[6] 6012 71.3% 1091 12.9% 642 7.6% 288 3.4% 80 0.9% 1.4 (5) / 8113
 
 
 
 
 
git-instaweb, or self-hosted web interface 5889 69.9% 836 9.9% 646 7.7% 519 6.2% 200 2.4% 1.6 (5) / 8090
 
 
 
 
 
tool to manage git repositories[7] 5383 63.9% 689 8.2% 668 7.9% 750 8.9% 558 6.6% 1.8 (5) / 8048
 
 
 
 
 
my own scripts (for daily use) 4551 54.0% 1104 13.1% 1037 12.3% 779 9.2% 446 5.3% 1.9 (5) / 7917
 
 
 
 
 
my own scripts (for special tasks) 4110 48.8% 1292 15.3% 1166 13.8% 576 6.8% 295 3.5% 1.9 (5) / 7439
 
 
 
 
 
Base 8429 / 8841


Notes (examples):

  1. For example EGit (for Eclipse), TextMate Git bundle, magit (for GNU Emacs)
  2. For example git-cheetah, TortoiseGit
  3. For example gitk
  4. For example git-gui
  5. For example Kompare, Meld
  6. For example 'git gui blame'
  7. For example Gitosis or Gitolite


Description:
Here graphics diff tool means tools such as Kompare, and graphical merge tool means tools such as Meld and KDiff3. Those answers include graphical merge and diff tools used by programmers editors and IDEs.

"graphical history browser (any)" covers tools such as gitk, QGit, Giggle, tig etc., but also built-in git commands such as "git log --graph" and "git show-branch". If you use one of mentioned tools as history browser, mark both a tool and "graphical history browser (any)"; if you use some graphical history viewer not listed here, please both mark this answer and specify it in the "other tool" answer.

Similarly for other answers marked "(any)".

12. What Git GUIs (graphical user interfaces) do you use?

(multiple choice, with other)

Answer Count Perc.
gitk 4042 58.4%
 
git gui 1729 25.0%
 
tig 513 7.4%
 
QGit 449 6.5%
 
Qct 6 0.1%
 
Giggle 312 4.5%
 
GitView 51 0.7%
 
gitg 507 7.3%
 
git-cola 174 2.5%
 
Pyrite 6 0.1%
 
GitX 1925 27.8%
 
GitNub 150 2.2%
 
Katana 4 0.1%
 
RepoWatch 5 0.1%
 
teamGit 5 0.1%
 
gitSafe 4 0.1%
 
SmartGit 159 2.3%
 
StupidGit 10 0.1%
 
git-cheetah 81 1.2%
 
TortoiseGit 799 11.5%
 
Git Extensions 247 3.6%
 
EGit (Eclipse) 640 9.2%
 
other IDE with git support 461 6.7%
 
editor with git support 642 9.3%
 
Other, please specify 488 7.1%
 
Base 6919 / 8841


Description:
Please select GUI that you use somewhat regularly (at least from time to time).

Similar issue applies to all other "What do you use?" kind of questions.

Analysis
Most popular graphical tool is gitk, a graphical history viewer, with 58% of responses. Most probably it is because it is the default history viewer, bundled with Git (though if you install git from packages, it usually is in a separate package).

git gui, a commit tool (a git GUI), also distributed with Git, is third with 25%, closely following GitX, a git GUI for MacOS X.

13. Which git hosting site(s) do you use for your project(s)?

(Please check only hosting sites where you publish/push to)

(multiple choice, with other)

The section of table with generic git hosting sites (first section) is sorted alphabetically by site name. Write-in (written in "Other (please specify)" are marked using different style and use different color of histogram (if it is visible). Not all write-ins were added.

Answer Count Perc.
A2Hosting 3 0.0%
 
Assembla 141 1.8%
 
Beanstalk 101 1.3%
 
BerliOS 11 0.1%
 
BitBucket 4 0.1%
 
Branchable 2 0.0%
 
CipherHive 2 0.0%
 
Codaset 54 0.7%
 
Codebase 118 1.5%
 
Codesion 13 0.2%
 
Gerrit 6 0.1%
 
GitFarm 7 0.1%
 
GitHub 6107 77.1%
 
Gitorious 936 11.8%
 
Google Code 4 0.1%
 
Heroku 34 0.4%
 
InDefero 61 0.8%
 
Launchpad 3 0.0%
 
ProjectLocker 47 0.6%
 
Redmine 9 0.1%
 
repositoryhosting 19 0.2%
 
repo.or.cz 305 3.9%
 
Savannah 68 0.9%
 
SourceRepo 3 0.0%
 
SourceForge 400 5.0%
 
The Chaw 9 0.1%
 
TuxFamily 7 0.1%
 
Unfuddle 349 4.4%
 
XP-Dev 4 0.1%
 
kernel.org 105 1.3%
 
freedesktop.org 76 1.0%
 
git.gnome.org 27 0.3%
 
Alioth 121 1.5%
 
debian-hosted 100 1.3%
 
Fedora Hosted 50 0.6%
 
company internal 34 0.4%
 
git hosting site for set of related projects (e.g. OLPC) 123 1.6%
 
generic site without git support 166 2.1%
 
self hosted 3496 44.1%
 
Other (please specify) 377 4.8%
 
Base 7922 / 8841


Description:
You can find those and other git hosting sites listed on GitHosting page.


Disclaimer:
Note that replies might be affected by the fact that announcement about this survey was shown at some git hosting sites, but not at others. Additionally users of some git hosting sites may skip welcome page more often than for other hosting site; also not all hosting sites have blog or a news / announcements section (or some hosting sites might have better mechanism for announcements)

Note also that there are git hosting sites, like for example Codaset, which were not present as one of choices in the ready list of answers. They were "write-in", its numbers retrieved by parsing contents of "other (please specify)" option. Nevertheless those write-ins might be under-represented.

The "company internal" answer was forgotten when creating this survey, and is also write-in.


Analysis:
The leader here is (like in previous years) GitHub with 77% of responders (up from 62% in 2009); "self hosted" is second with 44%, but it is quite lower than 57% in 2009.

The high position of GitHub might be at least partially caused by the fact that GitHub has good announcement system for new features, which was used to announce "Git User's Survey 2010" to GitHub users.


Below there is table showing how many responders selected how many options. Note that it does not analyse "other, please specify" answer, which might describe more than one git hosting site on one side, and might say something like "don't use git hosting site" on the other side.

Multiplicity / 24 Count Perc.
0 919 11.6%
 
1 4129 52.1%
 
2 2676 33.8%
 
3 831 10.5%
 
4 202 2.5%
 
5 54 0.7%
 
6 17 0.2%
 
7 9 0.1%
 
8 1 0.0%
 
9 2 0.0%
 
Base 7922 / 8841

Multiplicity analysis:
Most responders (52%) use only one git hosting site, around a third (34%) use two git hosting site (or selected one git hosting site and wrote something in "other, please specify" choice).

See also:

  • Survey results: Open source developer preferences at CodePlex weblog shows, based on around 500 responses, and shows repository hosting site oreferences and version control system preference by what is preferred operating system for development.

14. How do you fetch/get changes from upstream repositories?

(multiple choice, with other)

Answer Count Perc.
git protocol (e.g. git://example.com/repo.git) 6485 77.6%
 
ssh (e.g. ssh://example.com/repo.git or user@example.com:/repo.git) 5924 70.9%
 
http (e.g. http://example.com/repo.git) 3416 40.9%
 
rsync (DEPRECATED) (e.g. rsync://example.com/repo.git) 18 0.2%
 
filesystem (e.g. /path/to/repo.git or file:///path/to/repo.git) 1687 20.2%
 
via git-bundle 118 1.4%
 
foreign SCM import / script like git-svn 1299 15.5%
 
Other, please specify 87 1.0%
 
Base 8354 / 8841


Description:
This question asks about how do you get changes (updates) from projects you follow into your local repository. It is not about how do you get latest version of Git.

Fetching (or rather cloning) via bundle could mean that project publishes ready for download bundles to reduce traffic and load on server (HTTP download [of bundle] can be resumed, git-clone currently cannot; one can also distribute bundle using P2P).


Analysis:
git:// protocol and SSH are almost the same often used to fetch changes: git protocol with 78%, SSH with 71%. Only 41% responders use HTTP for fetching (from somewhere).

15. How do you publish/propagate your changes?

(multiple choice, with other)

Answer Count Perc.
push 7882 95.3%
 
pull request (+ any form of announcement / notification) 2765 33.4%
 
format-patch + email 1322 16.0%
 
format-patch + other (e.g. reviewboard, issue tracker or forum) 661 8.0%
 
git bundle 144 1.7%
 
git-svn (to Subversion repository) 1503 18.2%
 
git-p4 (to Perforce repository) 80 1.0%
 
foreign SCM interface (other than mentioned above) 126 1.5%
 
other - please specify 155 1.9%
 
Base 8272 / 8841


Description:
Publishing via bundle could mean sending bundle via email, or posting it on review board (or forum).

Instead of git-format-patch there can be utilized other way of generating patch (e.g. git-show, git-diff, 'patch' view from gitweb, etc.)

16. Which of the following features do you use?

(multiple choice, with other)

Answer Count Perc.
git bundle (off-line transport) 488 6.9%
 
"smart" HTTP support[1] 902 12.7%
 
eol conversion (crlf) 1307 18.4%
 
custom diff/merge driver 697 9.8%
 
textconv 82 1.2%
 
gitattributes (other than above[2]) 443 6.3%
 
submodules (subprojects) 2336 33.0%
 
subtree merge (optionally git-subtree[3]) 364 5.1%
 
partial (sparse) checkout 350 4.9%
 
assume-unchanged bit 144 2.0%
 
separate worktree / core.worktree 178 2.5%
 
multiple worktrees (git-new-workdir[4]) 203 2.9%
 
alternates mechanism (sharing object database) 160 2.3%
 
stash (optionally "git stash --keep-index") 4703 66.4%
 
shallow clone (e.g. "git clone --depth=<n>") 496 7.0%
 
detaching HEAD (e.g. "git checkout HEAD^0") 1367 19.3%
 
interactive rebase (small scale history editing) 3195 45.1%
 
interactive commit / per-hunk comitting / partial commit[5] 2869 40.5%
 
commit message templates 458 6.5%
 
git-filter-branch or equivalent (large history rewriting) 756 10.7%
 
bisect (optionally "git bisect run <script>") 1981 28.0%
 
committing with dirty tree (keeping some changes uncommitted) 3199 45.1%
 
non-default hooks (from contrib/hooks/ or other place) 1008 14.2%
 
shell completion of commands 3684 52.0%
 
git-aware shell prompt 2503 35.3%
 
git aliases, shell aliases for git, or own git scripts 2905 41.0%
 
one-way interaction with other SCMs (from SCM to git) 911 12.9%
 
two-way interaction with other SCMs (git as fat client) 1213 17.1%
 
git cvsserver 33 0.5%
 
git notes 237 3.3%
 
Other, please specify 100 1.4%
 
Base 7087 / 8841

Notes:

  1. Note that '"smart" HTTP support' is a feature that one can be using without knowing about it; it is enough if one uses new enough git, there is new enough git on server, and one uses http:// protocol for fetching.
  2. This means gitattributes other than `text`, `eol`, `diff.<name>.command`, `diff.<name>.textconf`, `merge.<name>.command` etc., which means `ident`, `filter`, custom diff hunk-header and word diff, selecting pre-defined merge drivers, `conflict-marker-size`, `whitespace`, `export-ignore`, `export-subst`, `delta`, `encoding`. See gitattributes(5) manpage.
  3. git-subtree is an external tool making it easier to use workflow which uses "subtree" merge strategy. It was at the time of this survey managed as external project.
  4. git-new-workdir is a shell script in contrib/worktree/ which can be used to manage multiple worktrees (working areas) for a single repository. Uses symbolic links.
  5. "interactive / partial commit" includes using commands like "git commit --interactive" or per-chunk comitting in "git gui".


Analysis: (does not include "other, please specify")
This question, to avoid complicating it, does not distinguish between features that are used often and features that are used rarely. Besides, some features are by their very nature to be used more rarely than other features.

Most popular feature, by quite a large margin, is stash, with two thirds (66.4%) of responses. "git stash" first appeared in git version 1.5.3, from September 2007.

Second is "shell completion of commands", with more than half of responses (52.0%).

Then there are, with percentage of responses from 40% to 45%, comitting with dirty tree and interactive / partial commit, interactive rebase, and aliases (git aliases, shell aliases, one's own git scripts).

More than third of responses have also git-aware shell prompt with 35%, and submodules with 33%.

Least used, with only 0.5% of responses, is "git cvsserver".

17. Which of the following features would you like to see implemented in git?

(multiple choice)

Answer Count Perc.
better support for big files (large media)[1] 2024 28.1%
 
resumable clone/fetch (and other remote operations) 1738 24.1%
 
GitTorrent Protocol, or git-mirror 998 13.9%
 
lazy clone / on-demand fetching of object 814 11.3%
 
subtree clone 1167 16.2%
 
support for tracking empty directories 2532 35.2%
 
environment variables in config 612 8.5%
 
better undo/abort/continue, and for more commands 1672 23.2%
 
'-n' like option for each command, which describes what would happen 2413 33.5%
 
warn before/when rewriting published history 1619 22.5%
 
side-by-side diffs and/or color-words diff in gitweb 1386 19.2%
 
admin and/or write features in gitweb 390 5.4%
 
graphical history view in gitweb 1188 16.5%
 
GUI for rebase in git-gui 1302 18.1%
 
GUI for creating repository in git-gui 493 6.8%
 
graphical diff/merge tool integrated with git-gui 1069 14.8%
 
syntax highlighting in git-gui 935 13.0%
 
filename encoding (in repository vs in filesystem) 397 5.5%
 
git push --create 1848 25.7%
 
localization of command-line messages (i18n) 258 3.6%
 
wholesame directory rename detection 1149 16.0%
 
union checkouts (some files from one branch, some from other) 517 7.2%
 
advisory locking / "this file is being edited" 467 6.5%
 
"commands issued" (or "command equivalents") in git-gui / gitk 420 5.8%
 
built-in gitjour/bananajour support 376 5.2%
 
better support for submodules 2070 28.7%
 
other (described below) 519 7.2%
 
Base 7203 / 8841


Notes:

  1. There is git-bigfiles git fork, which means to make life bearable for people using Git on projects hosting very large files (hundreds of megabytes), and to merge back as many changes as possible into upstream Git once they're of acceptable quality


Analysis:
One of the most requested features is native support for tracking empty directories, with 35.2%, followed closely by request for '-n' like option for each command, which would print what would happen, but do not actually execute command, with 33.5%. Both have more than third of responses.

Then there are requests for better support for submodules with 28.7%, and for big files (large media) with 28.1%. Then there is request for "git push --create" i.e. native way of creating repositories remotely, with 25.7% of responders. Following it closely is request for resumable fetch, clone, and other remote operations, with 24.1% responses. All those have got than fourth of responses (or close to it).

Least requested are lazy clone with 11.3%, and support for environment variables like $HOME in config file, with 8.5%.

18. Describe what features would you like to have in Git,

if they are not present on the list above (in previous question)

(free-form essay)

 TO TABULARIZE
 1187 / 8841 non-empty responses


What you think of Git

19. Overall, how happy are you with Git?

(single choice)

Answer Count Perc.
unhappy 88 1.0%
 
not so happy 323 3.8%
 
happy 1897 22.4%
 
very happy 4592 54.2%
 
completely ecstatic 1571 18.5%
 
Base 8471 / 8841

Analysis:
There were some complaints during planning stage of 2009 survey that the answers for this question are not symmetric; nevertheless it is set of answers that was used in previous surveys, and it would help comparing data with previous surveys to keep the current form.

Most responders are "very happy" with Git, and this answer seems to be a center of responses. One should take into account that if one is unhappy with Git, it is less likely that one would continue using it (unless other circumstances would force using it, like the project one contributes to using Git as DVCS of choice), thus less likely to be Git user and to participate in this Git User's Survey. There can be bias because it is Git survey; it might be different if it was generic survey about (distributed) version control systems.

20. In your opinion, which areas in Git need improvement?

Please state your preference.

(matrix)

Answer don't care don't need a little some much Avg. / Count
user-interface 1180 14.6% 1209 14.9% 2248 27.7% 2090 25.8% 1380 17.0% 3.2 / 8107
 
 
 
 
 
 
documentation 420 5.2% 1582 19.5% 2844 35.1% 2279 28.1% 861 10.6% 3.2 / 7986
 
 
 
 
 
 
performance 1167 14.4% 4735 58.4% 1370 16.9% 460 5.7% 145 1.8% 2.2 / 7877
 
 
 
 
 
 
more features 1203 14.8% 2864 35.3% 2680 33.1% 913 11.3% 161 2.0% 2.5 / 7821
 
 
 
 
 
 
tools (e.g. GUI) 1628 20.1% 1551 19.1% 1722 21.2% 1750 21.6% 1125 13.9% 2.9 / 7776
 
 
 
 
 
 
localization (translation) 4417 54.5% 2315 28.6% 535 6.6% 293 3.6% 176 2.2% 1.6 / 7736
 
 
 
 
 
 
community (mailing list) 4170 51.4% 2425 29.9% 799 9.9% 203 2.5% 67 0.8% 1.6 / 7664
 
 
 
 
 
 
community (IRC) 4173 51.5% 2261 27.9% 734 9.1% 181 2.2% 64 0.8% 1.6 / 7413
 
 
 
 
 
 
Base 8107 / 8841


Analysis:
From the table and histogram (graph) above one can see that:

  • User interface, tools and documentation all need improvement;
  • Performance rather doesn't need to be improved;
  • More features gets mainly "don't need" and "a little" as answers;
  • People mostly don't care about improving communities or localization (but please remember that this survey was in English, and announced on English-speaking sites / lists).

Changes in Git (since year ago, or since you started using it)

21. Did you participate in previous Git User's Surveys?

(multiple choice)

Answer Count Perc.
in 2006 64 0.8%
 
in 2007 164 2.0%
 
in 2008 467 5.6%
 
in 2009 912 10.9%
 
I don't remember, but I have participated in the past 969 11.6%
 
none of above (first Git User's Survey) 6492 77.9%
 
Base 8332 / 8841

Analysis:
Most responders (77.9%) didn't participate in previous surveys. This answer is explicit rather than implicit (by not answering the question) like in previous surveys.

It is good that "I don't remember, but I have participated in the past" got added (thanks to comments during planning stage of this survey); it has most answers (11.6%) next to not answering any of previous surveys.

22. How do you compare the current Git version with the version from one year ago?

(single choice)

Answer Count Perc.
better 3136 38.2%
 
no changes 675 8.2%
 
worse 9 0.1%
 
cannot say 4382 53.4%
 
Base 8202 / 8841


Description:
The version from approximately one year ago is 1.6.4 from 29-07-2009 (1.6.3 is from 06-05-2009),

Changes since this version include:

  • refs/replaces/* mechanism
  • git notes
  • "smart" HTTP transport
  • sparse checkout
  • "git push" safeties like refusing to push into checked out branch
  • "git status" is no longer "git commit --dry-run"
  • various ways to access and change upstream info
  • git log --glob=<pattern>
  • more hints, ability to squelch verbose "advice" messages
  • word diff support
  • TAP-ified test output
  • and more...

(see individual RelNotes for more details)

Disclaimer:
The original title of this question when starting this survey was "How do you compare the current version with the version from one year ago?". It was noticed during the time when survey was already open that this question is ambiguous and confusing: it is not clear whether 'version' refers to Git or to Git User's Survey. This was fixed, but in meantime many people seen older version.

It might be the cause of the fact that more than half of responders answered 'cannot say'.

Analysis:
Most people (53.4%, more than half) cannot say if Git improved since year ago or not. They might have chosen this answer because they didn't use Git a year ago, and that was the reason they could not do a comparison. This would indicate large number of new Git users participating in this survey.

But this is not the only explanation. It is possible that at least some of responders selected 'cannot say' answer because of ambiguity of changes: some are for better, some are for worse. Having the question stated in unclear way (see "Disclaimer" section above) for most of survey duration didn't help matters any. Yet another possibility that people do not remember version from year ago. Those different issues should be split as a separate options / separate answers in this single-choice question.

Discounting 'cannot say' answers, most people (38.2% of replies to this question) think that Git improved since year ago; almost none (0.1%, 9 in 8202 responses to this question) think that Git is worse than it was year ago.

Documentation. Getting help.

23. How useful have you found the following forms of Git documentation?

(matrix)

Answer never used not useful somewhat useful Avg. / Count  
Git Wiki 3308 40.1% 266 3.2% 2786 33.8% 1880 22.8% 2.4 (4) / 8240
 
 
 
 
on-line help 1308 15.9% 223 2.7% 3026 36.7% 3586 43.5% 3.1 (4) / 8143
 
 
 
 
help distributed with git 1125 13.7% 467 5.7% 2798 34.0% 3582 43.5% 3.1 (4) / 7972
 
 
 
 
Base 8240 / 8841


Description:


Analysis:
Git Wiki is least used form of documentation, with 40% of responders never using it. Generally, all documentation is seen as somewhat useful (34-37%) or useful (44%). Only Git Wiki, with many people not using it, has only 23% considering it useful and not somewhat useful.

24. Have you tried to get help regarding Git from other people?

(single choice)

Answer Count Perc.
Yes 5387 65.0%
 
No 2907 35.0%
 
Base 8294 / 8841

25. If yes, did you get these problems resolved quickly and to your liking?

(single choice)

Answer Count Perc.
Yes 3610 61.8%
 

[1]

Somewhat 1835 31.4%
 

[1]

No 401 6.9%
 

[1]

Base 5846 / 8841
 

[2]

Notes:

  1. 1.0 1.1 1.2 relative to numbers of people who answered this question
  2. relative to numbers of all survey responders

Analysis:
Consistency check: 5387 people answered that they tried to get help about Git from other people (in previous question), but we have 5846 responses for this question...

It looks like we have good community surrounding Git, if 62% of Git questions are resolved quickly and accurately, and only 7% couldn't get quick response and one to their liking. Those results are very similar to those from 2009 survey.

26. What channel(s) did you use to request help?

(multiple choice, with other)

Channel Count Perc.
git mailing list (git@vger.kernel.org) 415 7.2%
 
msysGit mailing list / Google Group 81 1.4%
 
"Git for Human Beings" Google Group 138 2.4%
 
IRC (#git) 1043 18.1%
 
IRC (other git/SCM related, e.g. #github) 287 5.0%
 
IRC (other than above) 632 10.9%
 
request in blog post or on wiki[1] 409 7.1%
 
asking git guru/colleague 3400 58.9%
 
project mailing list, or IRC, or forum 533 9.2%
 
Twitter or other microblogging platform 801 13.9%
 
instant messaging (IM) like XMPP/Jabber 1017 17.6%
 
StackOverflow[2] 2157 37.3%
 
other (please specify) 432 7.5%
 
Base 5777 / 8841

Notes:

  1. Here I mainly meant asking a question on one's blog, and waiting for response in blog comments, like e.g. in this blog entry: "Migrating to git from Subversion", or asking on Git Wiki (as if it was help forum), or on Talk page of some other wiki (no example).
  2. StackOverflow is a community driven programming-related Q&A site
    http://stackoverflow.com/questions/tagged/git

Analysis:
Most used 'channel' is simply to ask somebody better versed in Git personally, same as in previous surveys, with almost 60%. This year StackOverflow (a Q&A site) came as second with 37%, moving up from 17% last year when it was a new thing.

The #git IRC channel with usually fast real-time response is now third at 18%, down from 31% in 2009 and 37% in 2008. Together with other IRC channels (git/SCM related with 7%, other with 11%) it comes up to 34%, close to StackOverflow 37%. One wonders how disruptive StackOverflow chat would be...

Close fourth, with almost 18% of responses is instant messaging (XMPP/Jabber, Google Talk, AIM, QQ, ICQ, etc.), also with real-time response; 2009 survey had similarly 19% of IM users.

Git mailing list is 9th on the list with only 7.2% of replies, down from 12% in 2009 and 19% in 2008. The "Git for Human Beings" Google Group is down with 2.4%, in next to last place.

27. Which communication channel(s) do you use?

Do you read the mailing list, or watch IRC channel?

(multiple choice)

Channel Count Perc.
git@vger.kernel.org (main) 503 34.9%
 
Git for Human Beings (Google Group) 178 12.4%
 
msysGit 136 9.4%
 
#git IRC channel 789 54.8%
 
#git-devel IRC channel 36 2.5%
 
#github or #gitorious IRC channel 315 21.9%
 
#revctrl IRC channel 7 0.5%
 
Base 1441 / 8841


Disclaimer:
Note that percentage is relative to the (small) number of replies to this question, not relative to the number of all responders.

Analysis:
Among listed channels, most commonly used are #git IRC channel on FreeNode with 55%, and git mailing list (git@vger.kernel.org) with 34% (down from 42% in 2009). Third is #github and #gitorious IRC channels together, with 22% or replies to this question (same as in 2009).

About this survey.

28. How did you hear about this Git User's Survey?

(single choice, with other)

Answer Count Perc.
git mailing list 313 3.8%
 
git-related mailing list (msysGit, Git for Human Beings, ...) 72 0.9%
 
mailing list or forum of some project 566 6.8%
 
#git IRC channel topic 42 0.5%
 
announcement on IRC channel 61 0.7%
 
git homepage 408 4.9%
 
git wiki 147 1.8%
 
git hosting site 1629 19.7%
 
software-related web site 563 6.8%
 
news web site or social news site (e.g. Digg, Reddit) 835 10.1%
 
blog (or blog planet) 885 10.7%
 
other kind of web site 166 2.0%
 
Twitter or other microblogging platform 1211 14.6%
 
other - please specify 1390 16.8%
 
Base 8288 / 8841
Personal tools