{"thread":{"id":"10261","subject":"Git User's Survey 2007 summary - comparison with previous survey","startedAt":"2007-10-12T22:07:08Z","lastAt":"2007-10-12T22:07:08Z","messageCount":1,"participants":["Jakub Narebski"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"55575","messageId":"8fe92b430710121507h6b9b684bw9f39dbf3c3b14432@mail.gmail.com","threadId":"10261","inReplyTo":null,"subject":"Git User's Survey 2007 summary - comparison with previous survey","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2007-10-12T22:07:08Z","receivedAt":"2007-10-12T22:07:08Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"It's been more than a year since last Git User's Survey. It would be\ninteresting to find what changed since then. Therefore the idea to\nhave another survey.\n\nThis is partial comparison of Git User's Survey 2007,\n(ending at state from 28 September 2007) with previous\nsurvey.\n\nThe current survey can be found here:\n http://www.survey.net.nz/survey.php?94e135ff41e871a1ea5bcda3ee1856d9\n http://tinyurl.com/26774s\n\nThe data for current, 2007 survey can be found here:\n\nThe data for previous, 2006 survey was taken from following wiki page\n http://git.or.cz/gitwiki/GitSurvey2006\n\n---\nThere were 683 individual responses in 2007 survey.\nThere were around 117 responses in 2006 survey.\n\nThere are 62 questions in 2007 survey, and 31 questions (half of\nthe number) in 2006 survey.\n\n\nAbout you\n~~~~~~~~~\n\n01. What country are you in?\n\n Answer | Old | Count\n ------------------------------------------------\n Algeria | | 1\n Argentina | | 3\n Australia | 3 | 25\n Austria | 2 | 9\n Belarus | 1 |\n Belgium | | 5\n Brazil | 2 | 20\n Bulgaria | | 1\n Canada | 3 | 44\n Chile | 1 | 2\n China | 2 | 4\n Colombia | | 2\n Czech Republic | 2 | 10\n Denmark | 4 | 7\n Ecuador | | 1\n Estonia | 1 | 1\n Europe | 1 | 1\n Finland | 5 | 23\n France | 6 | 36\n Germany | 14 | 64\n Greece | | 3\n Hungary | | 2\n India | 1 | 13\n Ireland | | 2\n Israel | | 6\n Italy | 3 | 14\n Japan | | 4\n Jersey | | 1\n Latvia | | 1\n Lithuania | 1 |\n Malaysia | | 1\n Mexico | | 1\n Netherlands | 3 | 15\n New Zealand | | 5\n Norway | 1 | 14\n Philippines | 1 | 3\n Poland | 3 | 6\n Portugal | | 2\n Puerto Rico | | 1\n Romania | | 1\n Russian Federation | 2 | 6\n Samoa | | 1\n Serbia | | 1\n Singapore | | 2\n Slovak Republic | | 1\n Slovenia | | 2\n South Africa | 1 | 4\n Spain | 2 | 11\n Sri Lanka | | 1\n Sweden | 6 | 14\n Switzerland | 1 | 15\n UAE | 1 |\n UK / US | | 1\n United Kingdom | 8 | 40\n United States of America | 35 | 218\n Venezuela | | 1\n Vietnam | 1 | 1\n ------------------------------------------------\n Base | | 673 / 683\n Total (sum) | 117 | 673\n\nEngland, Scotland and British Isles counts as United Kingdom here.\nTable is sorted in alphabetical order.\n\nAs before most Git users are in the USA.\n\nThere are quite a bit of new countries, only two vanished. Note that\ncurrent survey has much more responses, so it is expected.\n\n\n02. What is your preferred non-programming language?\n\nThis is multiple answers question, although most people\ngave only one preferred language.\n\n Answer | Old | Count\n ------------------------------------------------\n Afrikaans | | 1\n Belarusian | 1 |\n Bulgarian | | 1\n Castellano | | 2\n Catalan | | 1\n Chinese | 1 | 2\n Czech | 2 | 10\n Danish | 5 | 6\n Dutch | 4 | 12\n English | 71 | 416\n Estonian | 1 |\n Finnish | 4 | 16\n French | 5 | 33\n Galician | | 1\n German | 12 | 58\n Greek | | 2\n Hebrew | | 1\n HibernoEnglish | | 1\n Hungarian | | 3\n Italian | 3 | 9\n Japanese | 1 | 1\n LSF (French sign language) | | 1\n Norwegian | | 4\n Polish | 3 | 5\n Portuguese | | 10\n Romanian | | 1\n Russian | 4 | 13\n Serbian | | 1\n Slovenian | | 2\n Spanish | 4 | 13\n Swedish | 5 | 13\n Swiss | | 1\n Ukrainian | | 1\n Vietnamese | 1 | 1\n\n invalid (computer language) | ? | 37\n not understood | ? | 4\n ------------------------------------------------\n Base | | 662 / 683\n Total (sum) | 127 | 683\n\nMost of git users prefer English language, at least for dealing with\ncomputers, the same as before.\n\nIn previous survey summary the invalid responses were not enumerated.\n\nThe question itself is not well formulated, as one can see from the\nnumber of answers with computer language, and \"not understood\"\nanswers. I am not native English speaker, but there were suggestions\nto use \"natural language\" instead of \"non-programming language\". The\nquestion is formulated the same as in previous survey.\n\n\n03. How old are you?\n04. Which programming languages you are proficient with?\n\nThose are new questions, which were not present in previous survey.\n\n\nGetting started with GIT\n~~~~~~~~~~~~~~~~~~~~~~~~\n\n05. How did you hear about GIT?\n\nThis question is present in both current (2007) and previous (2006)\nsurvey. But because it uses free-form answer, and tabularization\n(dividing data in categories and generating histogram / counting\noccurences) is quite different it is not easy to compare results.\n\nIn Git User's Survey 2006 (prevuius survey) the dominant source was\nLKML (Linux Kernel Mailing List) with 74 / 115 responses, with LWN\nleading the rest with 11 / 115 count.\n\nAlthough LKML still dominates the table for current survey, it is not\nby such a wide margin (109/658 = 17% as compared to 74/115 = 64% in\nprevious one). Many people heard about git because Linux kernel uses\nit; Linus Torvalds presentation (talk) at Google (Google Video and\nYouTube) got also high count, bit higher than LWN.\n\nBut there was no \"I wrote it\" response in current survey, though...\n\n\n06. Did you find GIT easy to learn?\n\n Answer | Old | Count\n ------------------------------------------------\n very easy | 6 | 38\n easy | 21 | 136\n reasonably | 64 | 318\n hard | 23 | 131\n very hard | 3 | 33\n ------------------------------------------------\n Base | | 656 / 683\n Total (sum) | 117 | 656\n\nNice gaussian curve both for current survey, and previous year survey\ndata. Most users find GIT reasonably easy to use.\n\n\n07. What helped you most in learning to use it?\n\nTO DO. This again is a free-form question, present in both surveys.\nFor current (2007) survey data is not even tabularized yet, although\n(some) of responses got listed (without a count). I don't think\ncomparison with previous result is any interesting, but previous year\ndata can provide at least some hint on how to divide answers into\ncategories.\n\n\n08. What did you find hardest?\n\nTO DO. Yet another free-form question, present in both surveys.\nTabularized for 2006 survey, listed without count for 2007 survey, as\nfor previous question.\n\n\n09. When did you start using git? From which version?\n\nTO DO. Again free-form question. Note that in previous survey this\nquestion consisted only of the first part, and read \"When did you\nstart using git?\", so it is date that got tabularized, not git version\n(mostly). Analysis of current survey data is unfinished, and only\nanswers in which there was given git version explicitely got\ntabularized. For comparison we would need analysis of answers giving\ndate; and even then I don't think it would be especially useful,\nunless very coarse-grained (which year for example).\n\n\nOther SCMs\n~~~~~~~~~~\n\n10. What other SCMs did/do you use?\n11. Why did you choose GIT?\n12. Why did you choose other SCMs?\n13. What would you require from GIT to enable you to change,\n if you use other SCM for your project?\n14. Did you import your repository from foreign SCM? What SCM?\n15. What tool did you use for import?\n16. Do your GIT repository interact with other SCM? Which SCM?\n17. What tool did/do you use to interact?\n\nThis whole section is new, and was not present in previous survey.\n\n\nHow you use GIT\n~~~~~~~~~~~~~~~\n\n18. Do you use GIT for work, unpaid projects, or both?\n\nThis question had slightly different wording in 2006 survey:\n\"How you use GIT? Do you use GIT for ...\".\n\n Answer / Purpose | Old | Count\n ------------------------------------------------\n work | 14 | 56\n unpaid projects | 50 | 212\n both | 53 | 377\n ------------------------------------------------\n Base | | 645 / 683\n Total (sum) | 117 | 645\n\nIn the previous survey unpaid projects (only unpaid) were slightly\nless than half of answers (43 %), while current survey shows that\n[only] around a third of git users use it only for unpaid projects.\nSo we see more git used for work.\n\n\n19. How do you obtain GIT?\n\n Answer | Old | Count\n ------------------------------------------------\n binary package | 31 | 283\n source tarball | 33 | 210\n pull from main repository | 53 | 153\n ------------------------------------------------\n Base | | 646 / 683\n Total (sum) | 117 | 646\n\nAs one can see a year ago pull from main repository dominated, while\nnow binary package dominates, and pull is least used. At least among\npeople who answered Git User's Survey; the difference might be caused\nby the fact that previous survey was distributed mainly among readers\nof git mailing list, who run latest 'master' or even 'next' version,\nand often contribute to git (see also question 30, \"Does git.git\nrepository include code produced by you?\").\n\nOn the other hand the difference might be caused by the fact that more\ndistributions have got git, or that git is more mature and there is no\nneed to run development version and/or the fact that x.y.z is now\ndevelopment version following 'master' and not 'maint', and includes\nnew features.\n\n\n20. What hardware platforms do you use GIT on?\n\nThis is free-form question, tabularized for both current and previous\nsurvey, but tabularized (slightly) differently. It is not suprising,\nas extracting architecture and dividing into categories was doneby two\ndifferent people: Paolo Ciarrocchi (if I remember correctly) for 2006\nsurvey, and Jakub Narebski (me) for 2007.\n\nMoreover I guess that the _comparison_ is not very interesting; the\ncurrent data is.\n\nHardware platfrom tables should relly be generated by someone better\nversed in computer architecture.\n\n\n21. What OS (please include the version) do you use GIT on?\n\nThis question is also free-form. For 2006 survey there is generic\ncount, and two tables: for Linux (distribution or kernel), and for\nother operating systems. The table for 2007 survey is slightly less\ngeneric than the 2006 count, but there are no OS version tables.\nThere is much more data, and more data without version number.\n\nPerhaps we should have provided examples of answers to this\nquestion...\n\n Answer | Old | Count\n ---------------------------------------------\n AIX | 1 | 1\n FreeBSD | 4 | 16\n OpenBSD | - | 3\n NetBSD | 1 | -\n HP-UX | 1 | 1\n IRIX | 1 | -\n Linux | 167? | 582\n MS Windows (Cygwin) | 14 | 22\n MS Windows (other) | | 36\n MacOS X / Darwin | 11 | 94\n Solaris | 3 | 11\n SunOS | - | 5\n ---------------------------------------------\n\nNote that one person can use git on more than one operating system.\nThe number for Linux from 2006 survey is taken as total for Linux\ntable. \"MS Windows (other)\" means msys (native) version, or\nunspecified whether Cygwin or MinGW/MSys. If I remember correctly\nthere were no native Windows version (which is currently under\ndevelopment) during previous survey.\n\nMost is Linux, as before. MacOS X is quite a percentage in current\nsurvey, and dominates among non-Linux OS, coming before MS Windows.\n\n\n22. What projects do you track (or download) using GIT\n (or git web interface)?\n\nThis question was not present in previous survey.\n\n\n23. How many people do you collaborate with using GIT?\n\nTO DO. This is free-form question; analysis for current survey is not\ndone yet. The answer is not always simple number, but even those\nanswers which are limited to simple number (or contain simple number)\nare not done yet.\n\nI have posted question if first post in the thread about what ranges\n(what categories) should be used here. I haven't received any answer\nyet.\n\nPrevious survey have table in two parts: first histogram of simple\nnumber of people, second with names of communities (Cairo, Linux\nkernel, U-Boot,...).\n\n24. How big are the repositories that you work on?\n\nTO DO. To be done *both* for 2007 (current) survey, and for 2006 one.\n\n\n25. How many different projects do you manage using GIT?\n\nThis question was not present in previous survey.\n\nPrevious survey has empty question in this place, between those\nquestions.\n\n\n26. Which porcelains do you use?\n27. Which git GUI do you use?\n28. Which (main) git web interface do you use for your projects?\n\nThe list of porcelains is different for previous and current\nsurvey. Git GUIs were moved to the next question in the 2007 survey,\nand web interfaces to second next.\n\nNote that there is some ambiguity concerning git-gui: it was put in\ncurrent survey in the GUI question only, while some people consider it\n[also] porcelain (and put is as 'other' in porcelain).\n\nMultiple answers (one can use more than one porcelain).\n\n Answer (multiple choice) | Old | Count\n ----------------------------------------------\n core-git | 62 | 558\n cogito | 22 | 56\n\n Patch management interface: : 13 : 57\n ..............................................\n StGIT | 11 | 41\n Guilt (formerly gq) | n/a | 13\n pg (deprecated) | 2 | 3\n ----------------------------------------------\n\nOne has to take into account the fact that neither Cogito nor pg\n(Patchy GIT) were deprecated during previous survey. Cogito got\ndeprecated because all of its functionality got moved in some way to\ncore git, while Petr 'Pasky' Baudis, main Cogito developer, didn't\nhave time for catching up to new git features.\n\nGuilt is so new that it just simply was not yet created during\nprevious survey.\n\n\n29. How do you publish/propagate your changes?\n\nThis question was not present in previous survey.\n\n\n30. Does git.git repository include code produced by you?\n\n(Previous survey had bad English here: \"Is the git.git repository\nincluding codes produced by you?\")\n\n Answer | Old | Count\n ------------------------------------------------\n Yes | 73 | 99\n No | 34 | 512\n ------------------------------------------------\n Base | | 611 / 683\n Total (sum) | 107 | 611\n\nComplete reversion: from most survey participants having their\ncode in git.git (around 70%) a year ago to around a small amount\n(around 16%) currently (!).\n\nThis might be cause by the fact that (I think) previous survey was\nknown mostly to people reading git mailing list, who often send their\nown patches to this list. Therefore there it would be nice to have\n\"How did you heard about this survey?\" question. (It would also help\nfinding where it is worth to send notice if/when we would want to make\nanother survey.)\n\nNote also the following little fact:\n\n $ git shortlog -s --all --before=\"01-08-2006\" | wc -l\n 119 (and around 117 individual responses in 2006 survey)\n $ git shortlog -s --all --before=\"01-10-2007\" | wc -l\n 215 (and 683 individual responses in 2007 survey)\n\nSo it is simply not possible to have the same percentage of git\ndevelopers among git users (or rather git survey participants): 70%\nout of 683 users is more than 465, which is more than git.git has\ndevelopers.\n\n\nInternationalization\n~~~~~~~~~~~~~~~~~~~~\n\n31. Is translating GIT required for wider adoption?\n32. What do you need translated?\n33. For what language do you need translation for?\n\nThis whole section is new, and was not present in previous survey.\n\n\nWhat you think of GIT\n~~~~~~~~~~~~~~~~~~~~~\n\n34. Overall, how happy are you with GIT?\n\n Answer | Old | Count\n ------------------------------------------------\n unhappy | 1 | 13\n not so happy | 19 | 36\n happy | 53 | 179\n very happy | 41 | 302\n completely ecstatic | 1 | 112\n ------------------------------------------------\n Base | | 642 / 683\n Total (sum) | 115 | 642\n\nDuring the year the balance shifted slightly in the positive\ndirection: from mostly happy to very happy, to mostly very happy\nleaning to happy. And we have much more percentage of completely\necstatic.\n\nGood work!\n\n\n35. How does GIT compare to other SCM tools you have used?\n\n Answer | Old | Count\n ------------------------------------------------\n Better | 80 | 505\n Equal (comparable) | 20 | 96\n Worse | 8 | 30\n ------------------------------------------------\n Base | | 631 / 683\n Total (sum) | 108 | 631\n\nHere not much changed. The shape of this histogram is almost the same,\nwith slight shift towards \"better than other SCMs used\".\n\nNote that 2006 survey has \"Equal (or uncomparable)\" instead of\n\"Equal (comparable)\" as it is now.\n\n\n36. What do you like about using GIT?\n\nTO DO. Yet another free-form question. Tabularized for 2006 survey,\nmost encountered answers listed without count for 2007 survey.\n\n\n37. What would you most like to see improved about GIT?\n (features, bugs, plug-ins, documentation, ...)\n\nTO DO. As above: tabularized somewhat for previous survey, with the\nlist of ideas (divided broadly into categories) below the table; with\nsome ideas dropped. For current survey there is only list of most\ncommonly encountered, and most interesting ideas.\n\nSome of the 2006 ideas got implemented, or are being implemented, like\nbetter documentation (Git User's Manual, documenting options), shallow\nclones, win32 native binaries (via MinGS, in development), subproject\nsupport (plumbing and beginnings of porcelain), libification and\nbuiltinification (GSoC projects), increased verbosity when needed and\nmaking error messages more helpful, graphical merges (git-mergetool:\ninterface to file-level graphical mergers like Meld, KDiff3 etc.), per\nuser configuration (~/.gitconfig rather that ~/.gitrc, and there is\neven system wide configuration). Some of ideas are repeated (like\n\"The Git Book\" idea, although \"Git User's Manual\" fills it somewhat),\nsome ideas provided too hard (e.g. lazy clone aka remote alternates)\nor without someone to implement them. Some got abandoned, some will\nprobably never get implemented.\n\nEven some suggestions in 2007 survey are actually implemented already,\nfor example git development Changelog (present in the form of\nRelNotes), shallow clone support, submodules support. This means that\nnew features are not very well announced (which was also one of\ncomments in current survey).\n\n\n38. If you want to see GIT more widely used, what do you think\n we could do to make this happen?\n\nTO DO. List of suggestions for 2006 survey; for current 2007 survey\nonly two most striking.\n\n\nChanges in GIT (since year ago, or since you started using it)\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\n39. Did you participate in previous Git User's Survey?\n40. What improvements you wanted got implemented?\n41. What improvements you wanted didn't get implemented?\n42. How do you compare current version with version from year ago?\n43. Which of the new features do you use?\n\nThis whole section is of course new (Git User's Survey 2006 was first\ngit survey ever; git is not very much older than that), and was not\npresent in previous survey.\n\n\nDocumentation\n~~~~~~~~~~~~~\n\n44. Do you use the GIT wiki?\n45. Do you find GIT wiki useful?\n\nPrevious survey has two questions in one here: \"Do you use the GIT\nwiki? If yes, do you find it useful?\". This made it hard to\ndistinguish if \"no\" means \"no I don't use GIT wiki\" or \"no it is not\nuseful\".\n\nThis was one of the improvements over previous version of survey.\n\n\n46. Do you contribute to GIT wiki?\n\nThis question was not present in previous survey.\n\n\n47. Do you find GIT's on-line help (homepage, documentation) useful?\n\nIn 2007 survey there is additional \"somewhat useful\" answer, which was\nnot present in the 2006 survey.\n\n Answer | Old | Count\n --------------------------------------------\n Yes | 88 | 377\n No | 20 | 28\n Somewhat | - | 172\n ------------------------------------------\n Base | | 577 / 683\n Total (sum) | 108 | 577\n\nThe results are similar: most users find online help useful.\n\n\n47b. What is your favourite user documentation for any software\n projects or products you have used?\n\nThis question is present in 2006 survey, and was removed in current\none. The idea behind question was I guess to have the results in hand\nif git ever was to change documentation format. Most likely current\nformat of documentation (AsciiDoc), or at least idea behind it\n(it should be possible to read sources, editing sources without\nspecialized editor support should be easy even for people who don't\nknow the format) is here to stay.\n\n\n48. Do you find help distributed with GIT useful\n (manpages, manual, tutorial, HOWTO, release notes)?\n49. Did/Do you contribute to GIT documentation?\n\nThose questions were not present in previous survey.\n\n\n50. What could be improved on the GIT homepage?\n\nTO DO. List of suggestions for 2006 survey, nothing yet for 2007\nsurvey.\n\n\n51. What topics would you like to have on GIT wiki?\n52. What could be improved in GIT documentation?\n\nThose questions were not present in previous survey.\n\n\nGetting help, staying in touch\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\n53. Have you tried to get GIT help from other people?\n\n Answer | Old | Count\n ------------------------------------------------\n Yes | 68 | 357\n No | 45 | 261\n ------------------------------------------------\n Base | | 618 / 683\n Total (sum) | 113 | 618\n\nAround 60% of people tried to get GIT help from other people,\nfor both current (2007) and previous (2006) survey.\n\n\n54. If yes, did you get these problems resolved quickly\n and to your liking?\n\n Answer | Old | Count\n --------------------------------------------\n Yes | 68 | 326\n No | 45 | 53\n ------------------------------------------\n Base | | 379 / 683\n Total (sum) | 113 | 379\n\nThe precentage was a bit worse during earlier survey (around 60%) than\nfor corrent one (around 86%).\n\nThis might be caused by the fact that git is now more userfriendly,\nand has more features, and problems are easier to resolve.\n\n\n55. Would commerical (paid) support from a support vendor\n be of interest to you/your organization?\n\nThis question was not present in previous survey. It was requested\nduring an RFC for this year (2007) survey.\n\n\n56. Do you read the mailing list?\n\n Answer | Old | Count\n ------------------------------------------------\n Yes | 67 | 204\n No | 50 | 406\n ------------------------------------------------\n Base | | 610 / 683\n Total (sum) | 117 | 610\n\nNote that 2006 version had \"subscribe\" instead of \"read\" in this\nquestion. Even despite that the number of people reading git mailing\nlist decreased from more than half (around 57%) to around third\n(around 33%). This might be caused by the fact that notice / info\nabout Git User's Survey 2006 was distributed mainly among git mailing\nlist and among mailing lists for projects which use git (see also\ncommets to questions 19 and 30).\n\nThe fact that only one third of git users are reading mailing list\n(which is nevertheless quite large percentage) means that features\n_have_ to be documented somewhere besides mailing list archive and\nlogs (commit messages).\n\n\n57. If yes, do you find the mailing list useful?\n58. Do you find traffic levels on GIT mailing list OK?\n\nPrevious survey has two questions in one here: \"If yes, do you find it\nuseful, and traffic levels OK?\". They were split in current survey.\n\n\n59. Do you use the IRC channel (#git on irc.freenode.net)?\n\n Answer | Old | Count\n ------------------------------------------------\n Yes | 23 | 182\n No | 93 | 376\n ------------------------------------------------\n Base | | 558 / 683\n Total (sum) | 116 | 558\n\nMore people use #git channel now (33% as compared to 20% before).\nNevertheless relatively few people use this form of comminication and\ngetting help.\n\n\n60. If yes, do you find IRC channel useful?\n\nThis question was not present in previous survey. It nicely follows in\nthe series of questions about git wiki and git mailing list (do you\nuse? is it useful?).\n\n\n61. Did you have problems getting GIT help on mailing list\n or on IRC channel? What were it? What could be improved?\n\nThis question was not present in previous survey.\n\n\nOpen forum\n~~~~~~~~~~\n\n62. What other comments or suggestions do you have\n that are not covered by the questions above?\n\nTO DO. There is list of responses for 2006 survey; the 141 responses\nin current (2007) survey are not yet analysed.\n\n-- \nJakub Narebski\n(currently away from net)\n"}]}