{"thread":{"id":"20625","subject":"Git User's Survey 2009 partial summary, part 2 - from first 10","startedAt":"2009-08-17T10:24:43Z","lastAt":"2009-08-18T19:55:50Z","messageCount":6,"participants":["Jakub Narebski","Johan Herland","Nicolas Sebrecht"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"120859","messageId":"200908171224.44686.jnareb@gmail.com","threadId":"20625","inReplyTo":null,"subject":"Git User's Survey 2009 partial summary, part 2 - from first 10","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2009-08-17T10:24:43Z","receivedAt":"2009-08-17T10:24:43Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"Git User's Survey 2009 partial summary, part 2 - git difficulty,\nproficiency, uses, install, OS, editors.\n\n\nIt more than a month since Git User's Suvey 2009 was started (it was\nstarted on 15 July 2009), which is a half of planned duration time of\nthe survey.\n\nSo I think this is time for second part partial summary of Git User's\nSurvey 2009.  The previous part can be found at\n  http://thread.gmane.org/gmane.comp.version-control.git/124599\n  Subject: Git User's Survey 2009 partial summary, part 1 -\n           - announcing survey, participation\n  Message-Id: <200907312246.12134.jnareb@gmail.com>\n\n\nYou can see summary of Git User's Survey 2009 responses (and make your\nown analysis) at the following URL:\n\n  http://www.survs.com/shareResults?survey=2PIMZGU0&rndm=678J66QRA2\n  http://tinyurl.com/GitSurvey2009Analyze\n\nAfter the survey ends (or earlier, if it is requested) the raw data\nwould be published on GitSurvey2009 page on Git Wiki in CSV and XLS\nformats (like for GitSurvey2008).\n\n\nWe have currently (at the time I have checked this) around 3089\nresponses, as compared to 3236 individual responses (including 21\nresponses in 'test' channel) for survey in 2008, 683 individual\nresponses in 2007, and around 117 responses in 2006.\n\n\nBelow there is analysis of few questions from this survey; note that\nresults might do not add to 100% because of replies during using live\nanalysis tool, and the rounding.\n\n\n3) Have you found Git easy to learn?\n4) Have you found Git easy to use?\n   (Choice - Single answer)\n\n================================================\nAnswer            | to learn [%] |   to use [%]\n------------------------------------------------\nVery easy         |          4%  |          9%\nEasy              |         20%  |         36%\nReasonably easy   |         55%  |         45%\nHard              |         19%  |          8%\nVery hard         |          2%  |          1%\n------------------------------------------------\nTotal respondents |        2942  |        2959\n================================================\n\nAs you can see the myth that Git is hard to learn was busted ;-) Not.\n\nWhen analyzing this data you should take into account that somebody\nconsidering or finding Git to hard to learn wouldn't be probably Git\nuser, and thus wouldn't fill this survey.  So because it is \n*Git User's* Survey we should consider that results can be skewed\ntowards lower value (easier).\n\nWhat's interesting is comparing (percentage) results for questions\n3. and 4.; how hard is git to learn versus how hard is to use.  It\nseems like Git is reasonably easy to learn, and reasonably easy to\neasy to use.  So it looks like Git just have somewhat steep learning\ncurve, and the difficulty to learn pays in being more powerful to\nuse.\n\n\n6) Rate your own proficiency with Git:\n   (Choice - Single answer)\n\nYou can think of it as 1-5 numerical grade of your proficiency in Git.\n\n================================================\nProficiency               | resp [%] | resp [n]\n------------------------------------------------\n1. novice                 |       4% |      114\n2. casual, needs advice   |      17% |      520\n3. everyday use           |      38% |     1138\n4. can offer advice       |      34% |     1020\n5. know it very well      |       6% |      192\n--------------------------+---------------------\nTotal respondents         |                2984\nSkipped this question     |                 105\n================================================\n\nAs you can see most responders know Git enough for everyday use, or\ncan even offer some Git advice.  If we treat possible answers as a\nproficiency grade, the average proficiency is around 3.2.\n\nEither Git users don't stay novices long, or survey announcement(s)\ndidn't reach novice users.\n\nThis data could be useful later to be able to check how it differs\ngit usage for novices and for gurus (what commands they run, what\nfeatures they use at different levels of proficiency).\n\n\n7) I use Git for (check all that apply):\n   (Choice - Multiple answers)\n\nNote that above choices are neither orthogonal nor exclusive.  One\nmight want to check multiple answers even for a single repository.\n\n============================================\nAnswer                | resp [%] | resp [n]\n--------------------------------------------\nwork projects         |      79% |     2366\nunpaid projects       |      80% |     2391\n............................................\nwork and unpaid       |      66% |     1960\nonly work             |      14% |      407\nonly unpaid           |      15% |      432\nnot answered          |       6% |      190\n--------------------------------------------\nproprietary projects  |      37% |     1108\nOSS development       |      69% |     2054\nprivate code          |      77% |     2315\n--------------------------------------------\ncode (programming)    |      87% |     2591\nweb app               |      43% |     1274\nconfiguration files   |      34% |     1010\npersonal data         |      33% |      972\ndocuments             |      32% |      956\nstatic website        |      29% |      881\n--------------------------------------------\nfrontend to other SCM |      32% |      971\nsharing data or sync  |      22% |      645\nbackup                |      24% |      722\nbackend for web app   |       9% |      263\n............................................\n--------------------------------------------\nother                 |       3% |       84\n============================================\nTotal respondents     |                2989\nSkipped this question |                 104\n============================================\n\nThe answers in the table above are in slightly different order than in\noriginal, i.e. in the survey.  Analysis of correlations was done here\nonly for work / unpaid part of answer (see table above).\n\nThis question joins together (to reduce number of questions in the\nsurvey) few separate issues:\n\n * Whether one use Git for work, or for unpaid projects: this survey\n   results shows that the amount of paid and unpaid projects are about\n   the same, and that most of people (most of responders) us Git for\n   both work (paid) and free (unpaid) projects.\n\n * The question of openess / license: whether one uses Git for\n   proprietary projects or for FLOSS (Free/Libre/Open Source\n   Software), or perhaps for private code (not covered by license).\n   Here OSS development and private code/data dominates over using Git\n   for proprietary code, with around twice as much answers.  This may\n   be caused by the fact that Git is / feels more suited to opensource\n   (like) development (proprietary code can use proprietary SCM), but\n   it might be that announcement didn't make it to people using Git\n   for proprietary works.\n\n * Whether one uses Git for code (for something it was created for),\n   or for other things like documents etc.  Here code dominates,\n   having around twice as many replies as the next in turn, 'web app'\n   (which is also code, but a special case, and not only code).\n\n * next is UGFWIINI part: using Git for what it was not intended.\n   Well, except perhaps using Git as frontend or fat-client for other\n   SCM (e.g. git-svn for Subversion).  Using Git for other SCM\n   dominates with 32%, next are using Git for sharing, and for backup\n   (both with around 23% of responses).\n\n * The \"other (please specify)\" is not yet analyzed (nor even began\n   analysis so far).\n\n\n8) How do/did you obtain Git (install and/or upgrade)? \n  (Choice - Multiple answers)\n\nExplanation: binary package covers pre-compiled binary (e.g. from rpm\nor deb binary packages); source package covers things like deb-src and\nSRPMS/*.src.rpm; source script is meant to cover installation in\nsource-based distributions, like 'emerge' in Gentoo.\n\nAutomatic update (apt, yum, etc.) in most cases means binary package\ninstall; unless one uses source-based distribution like Gentoo, CRUX,\nor SourceMage, where automatic update means using source package (or\nsource script).\n\nThe option named \"preinstalled / sysadmin job\" means that either you\ndidn't need to install git because it was preinstalled (and you didn't\nupgrade); or that you have to ask system administrator to have git\ninstalled or upgraded.\n\n\nNote that this question is multiple choices question because one can\ninstall Git in different ways on different machines or on different\noperating systems.\n\n==================================================\nAnswer                      | resp [%] | resp [n]\n--------------------------------------------------\nbinary package              |      71% |     2118\nsource package or script    |      24% |      704\nsource tarball              |      20% |      588\npull from repository        |      18% |      527\npreinstalled / sysadmin job |       6% |      185\nother - please specify      |       4% |      123\n----------------------------+---------------------\nTotal respondents           |                2975\nSkipped this question       |                 118\n==================================================\n\nMost people (71%) use ready binary packages, which was kind of\nexpected; that is the easiest way.\n\n\n9) On which operating system(s) do you use Git?\n   (Choice - Multiple answers)\n\nOn Unix-based operating system you can get the name of operation\nsystem by running 'uname'.\n\n============================================================\nOperating system                      | resp [%] | resp [n]\n------------------------------------------------------------\nLinux                                 |      88% |     2623\nMacOS X (Darwin)                      |      44% |     1304\nMS Windows/Cygwin                     |       9% |      279\nMS Windows/msysGit (MINGW)            |      21% |      625\n*BSD (FreeBSD, OpenBSD, NetBSD, etc.) |       7% |      213\nOpenSolaris                           |       3% |      101\nother Unix                            |       2% |       60\nOther, please specify                 |       2% |       46\n......................................|..........|..........\nMS Windows (any)                      |      27% |      820\n--------------------------------------+---------------------\nTotal respondents                     |                2991\nRespondents who skipped this question |                 102\n============================================================\n\nMost common used operating system is Linux, next is MacOS X, and then\nMS Windows (on MS Window dominates msysGit version).  This is quite\nunderstandable, as Git was created on Linux and for Linux, and it\nworks best there.\n\nThe \"other\" answer include Solaris, AIX, HP-UX (other Unix), GNU/Linux\n(Linux), Debian, NixOS (Linux distributions), none (?); iPod, OpenVMS,\nQNX, Hurd and \"my toaster\" (?).  And some commentaries and\nclarifications.\n\n\n10) What do you use to edit contents under version control with Git?\n    What kind of editor, IDE or RAD you use working with Git?\n    (Choice - Multiple answers)\n\n* \"simple text editor\" option includes editors such as pico, nano,\n  joe, Notepad,\n\n* \"programmers editor\" option includes editors such as Emacs/XEmacs,\n  Vim, TextMate, SciTE (syntax highlighting, autoindentation,\n  integration with other programmers tools, etc.)\n\n* \"IDE (Integrated Development Environment) and RAD (Rapid Application\n  Development)\" option includes tools such as Eclipse, NetBeans IDE,\n  IntelliJ IDE, MS Visual Studio, KDevelop, Anjuta, Xcode,\n  Code::Blocks but also tools such as Quanta+, BlueFish or Screem (for\n  editing HTML, CSS, PHP etc.), and Kile or LEd for LaTeX.\n\n* \"WYSIWYG tools\" option includes word processors such as MS Office or\n  OpenOffice.org, but also tools such as Adobe Acrobat (for PDF) or\n  GIMP (for images), or WYSIWYG DTP tools such as QuarkXPress,\n  PageMaker or Scribus, or WYSIWYG HTML editors such as FrontPage,\n  Dreamweaver or KompoZer.\n\n============================================================\nAnswer                                | resp [%] | resp [n]\n------------------------------------------------------------\nsimple text editor                    |      25% |      752\nprogrammers editor                    |      88% |     2641\nIDE or RAD                            |      32% |      960\nWYSIWYG tool                          |       5% |      159\nother kind - please specify           |       3% |       95\n--------------------------------------+---------------------\nTotal respondents                     |                2985\nRespondents who skipped this question |                 108\n============================================================\n\nMost popular kind of tool is programmer's editor with very strong 88%\nlead.  Next in turn are IDE or RAD with 32%, followed by simple text\neditor, with 25%.  The WYSIWYG tools are last, with very small 5% of\nreplies; also people who use WYSIWYG tools also use other kind of\ntools to edit (presumably to edit commit message at least).\n\n>From browsing through \"other tool\" responses it seem like there should\nbe included 'web interface' among possible tools (written as: wiki\nfrontend, GitHub editor, <textarea>).  Other responses include Xournal\n(note taking application for graphic tablet, using stylus), Tinderbox\n(also note taking application), Git GUI tool: GitX (?), custom\ngit-aware CMS (probably also 'web interface'), scripts and oneliners,\ncommand line and 'On occasion, Unix text manipulation tools'.  Also\nthere is \"audio editors\" answer (which probably should fall under\n\"WYSIWYG tool)...\n\nMany answers in 'other tool' falls in one of ready categories; perhaps\nthey are clarification?\n\n-- \nJakub Narebski\nGit User's Survey 2009: http://tinyurl.com/GitSurvey2009\n"},{"id":"120862","messageId":"200908171244.21693.johan@herland.net","threadId":"20625","inReplyTo":"200908171224.44686.jnareb@gmail.com","subject":"Re: Git User's Survey 2009 partial summary, part 2 - from first 10","fromName":"Johan Herland","fromEmail":"johan@herland.net","sentAt":"2009-08-17T10:44:21Z","receivedAt":"2009-08-17T10:44:21Z","isPatch":false,"sender":{"key":"johan@herland.net","avatar":"https://avatars.githubusercontent.com/u/547031?v=4"},"body":"On Monday 17 August 2009, Jakub Narebski wrote:\n> 9) On which operating system(s) do you use Git?\n>    (Choice - Multiple answers)\n>\n> On Unix-based operating system you can get the name of operation\n> system by running 'uname'.\n\nI find it interesting to compare the answers to this question against \nprevious years' surveys:\n\n=============================================================\nOperating system            | 2009 [%] | 2008 [%] | 2007 [%]\n-------------------------------------------------------------\nLinux                       |      88% |      86% |      90%\nMacOS X (Darwin)            |      44% |      47% |      15%\nMS Windows/Cygwin           |       9% |      10% |      >3%?\nMS Windows/msysGit (MINGW)  |      21% |      16% |      >0%?\n............................|..........|..........|..........\nMS Windows (any)            |      27% |     <26%?|       9%\n\nAlthough several of these numbers are probably inaccurate, and\nassuming that the responses on this question is representative,\nwe see a stabilizing trend in Git's user base, at least when\nlooking at Linux vs. MacOS. Windows seems to be growing steadily,\nand it's good to see that the msysGit portion is outgrowing the\nCygwin portion.\n\nThanks to Dscho and all the others that make msysGit better\nevery day!\n\n\nHave fun! :)\n\n...Johan\n\n-- \nJohan Herland, <johan@herland.net>\nwww.herland.net\n"},{"id":"120997","messageId":"200908172339.08767.jnareb@gmail.com","threadId":"20625","inReplyTo":"200908171244.21693.johan@herland.net","subject":"Re: Git User's Survey 2009 partial summary, part 2 - from first 10","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2009-08-17T21:39:06Z","receivedAt":"2009-08-17T21:39:06Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"\nOn Mon, 17 Aug 2009, Johan Herland wrote:\n> On Monday 17 August 2009, Jakub Narebski wrote:\n\n> > 9) On which operating system(s) do you use Git?\n> >    (Choice - Multiple answers)\n> >\n> > On Unix-based operating system you can get the name of operation\n> > system by running 'uname'.\n> \n> I find it interesting to compare the answers to this question against \n> previous years' surveys:\n> \n> =============================================================\n> Operating system            | 2009 [%] | 2008 [%] | 2007 [%]\n> -------------------------------------------------------------\n> Linux                       |      88% |      86% |      90%\n> MacOS X (Darwin)            |      44% |      47% |      15%\n> MS Windows/Cygwin           |       9% |      10% |      >3%?\n> MS Windows/msysGit (MINGW)  |      21% |      16% |      >0%?\n> ............................|..........|..........|..........\n> MS Windows (any)            |      27% |     <26%?|       9%\n\nWhet might be also interesting is to have (also) bare numbers of\nresponses.  Percentage alone wouldn't tell full story.  If you have\n(for example) 1000 Linux users, and there gets released native MacOS X\nbinary installer, and here comes 500 MacOS X users, the percentage\nwould decrease, because there are now more users, some of which are\nnot Linux users.\n\n=================================================\nOperating system            | 2009 | 2008 | 2007#\n-------------------------------------------------\nLinux                       | 2623 | 2434 |  582\nMacOS X (Darwin)            | 1304 | 1330 |   94\nMS Windows/Cygwin           |  279 |  297 |   22\nMS Windows/msysGit (MINGW)  |  625 |  452 |    1\n............................|......|......|......\nMS Windows (any)            |  820 |  749?|   58*\n----------------------------+------+------+------\nAnswers                     | 2991 | 2842 |  645       \n=================================================\n\nFootnotes:\n[#] First download on current msysGit page is from Aug 2007\n[*] Including unspecified version of Git for MS Windows.\n[?] \"Other, please specify\" not included.  It also doesn't\n    take into account overlap (one might use both msysGit and Cygwin\n    version, e.g. on different computers)\n\n-- \nJakub Narebski\n\nGit User's Survey 2009: http://tinyurl.com/GitSurvey2009\n"},{"id":"121109","messageId":"20090818155452.GA12938@vidovic","threadId":"20625","inReplyTo":"200908171224.44686.jnareb@gmail.com","subject":"Re: Git User's Survey 2009 partial summary, part 2 - from first 10","fromName":"Nicolas Sebrecht","fromEmail":"nicolas.s.dev@gmx.fr","sentAt":"2009-08-18T15:54:52Z","receivedAt":"2009-08-18T15:54:52Z","isPatch":false,"sender":{"key":"nicolas.s.dev@gmx.fr","avatar":null},"body":"The 17/08/09, Jakub Narebski wrote:\n\n> Git User's Survey 2009 partial summary, part 2 - git difficulty,\n> proficiency, uses, install, OS, editors.\n\nThanks for these partial summaries.\n\n> 3) Have you found Git easy to learn?\n> 4) Have you found Git easy to use?\n>    (Choice - Single answer)\n> \n> ================================================\n> Answer            | to learn [%] |   to use [%]\n> ------------------------------------------------\n> Very easy         |          4%  |          9%\n> Easy              |         20%  |         36%\n> Reasonably easy   |         55%  |         45%\n> Hard              |         19%  |          8%\n> Very hard         |          2%  |          1%\n> ------------------------------------------------\n> Total respondents |        2942  |        2959\n> ================================================\n\n<...>\n\n> What's interesting is comparing (percentage) results for questions\n> 3. and 4.; how hard is git to learn versus how hard is to use.  It\n> seems like Git is reasonably easy to learn, and reasonably easy to\n> easy to use.  So it looks like Git just have somewhat steep learning\n> curve, and the difficulty to learn pays in being more powerful to\n> use.\n\nI believe it would be interesting to know who (from the question 6.)\nthink what later. We may expect that people of the grade 4 and 5 (\"can\noffer advice\" and \"know it very well\") underestimate the difficulty to\nlearn Git.\n\nAlso (and as you said), this \"Git's users\" survey won't have answers\nfrom unsatisfied users who left Git. We can't rate the number of users\nwho left Git because they found it too much hard to learn.\n\n> 6) Rate your own proficiency with Git:\n>    (Choice - Single answer)\n> \n> You can think of it as 1-5 numerical grade of your proficiency in Git.\n> \n> ================================================\n> Proficiency               | resp [%] | resp [n]\n> ------------------------------------------------\n> 1. novice                 |       4% |      114\n> 2. casual, needs advice   |      17% |      520\n> 3. everyday use           |      38% |     1138\n> 4. can offer advice       |      34% |     1020\n> 5. know it very well      |       6% |      192\n> --------------------------+---------------------\n> Total respondents         |                2984\n> Skipped this question     |                 105\n> ================================================\n\n-- \nNicolas Sebrecht\n"},{"id":"121127","messageId":"200908181900.03356.jnareb@gmail.com","threadId":"20625","inReplyTo":"20090818155452.GA12938@vidovic","subject":"Re: Git User's Survey 2009 partial summary, part 2 - from first 10","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2009-08-18T17:00:00Z","receivedAt":"2009-08-18T17:00:00Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"On Tue, 18 Aug 2009, Nicolas Sebrecht wrote:\n> The 17/08/09, Jakub Narebski wrote:\n> \n> > Git User's Survey 2009 partial summary, part 2 - git difficulty,\n> > proficiency, uses, install, OS, editors.\n> \n> Thanks for these partial summaries.\n\nYou are welcome.  Next summary either around 1 September (3/4 duration),\nor after passing number of responses in previous survey; it also depends\nwhen I would have time.\n\n> > 3) Have you found Git easy to learn?\n> > 4) Have you found Git easy to use?\n> >    (Choice - Single answer)\n> > \n> > ================================================\n> > Answer            | to learn [%] |   to use [%]\n> > ------------------------------------------------\n> > Very easy         |          4%  |          9%\n> > Easy              |         20%  |         36%\n> > Reasonably easy   |         55%  |         45%\n> > Hard              |         19%  |          8%\n> > Very hard         |          2%  |          1%\n> > ------------------------------------------------\n> > Total respondents |        2942  |        2959\n> > ================================================\n> \n> <...>\n> \n> > What's interesting is comparing (percentage) results for questions\n> > 3. and 4.; how hard is git to learn versus how hard is to use.  It\n> > seems like Git is reasonably easy to learn, and reasonably easy to\n> > easy to use.  So it looks like Git just have somewhat steep learning\n> > curve, and the difficulty to learn pays in being more powerful to\n> > use.\n> \n> I believe it would be interesting to know who (from the question 6.)\n> think what later. We may expect that people of the grade 4 and 5 (\"can\n> offer advice\" and \"know it very well\") underestimate the difficulty to\n> learn Git.\n\nWell, you can do some analysis yourself, using just created filters at\nhttp://www.survs.com/shareResults?survey=2PIMZGU0&rndm=678J66QRA2\n\nFor example:\n\n================================================|\n                  |    proficiency with Git     |\nTo learn          |-----------------------------|\n                  | any*[%] | 1-3 [%] | 4-5 [%] |\n------------------------------------------------|\nVery easy         |     4%  |     3%  |     5%  |\nEasy              |    20%  |    19%  |    22%  |\nReasonably easy   |    55%  |    55%  |    56%  |\nHard              |    19%  |    20%  |    16%  |\nVery hard         |     2%  |     3%  |     1%  |\n------------------------------------------------|\nTotal respondents |   3005  |   1786  |   1212  |\n================================================|\n\nFootnote:\n^^^^^^^^^\n[*] This means no filter, which includes people who didn't answered\n    question about proficiency, but answered this one.\n \n> Also (and as you said), this \"Git's users\" survey won't have answers\n> from unsatisfied users who left Git. We can't rate the number of users\n> who left Git because they found it too much hard to learn.\n\nHere also you have to be careful about the fact that \"correlation does\nnot imply causation\" (or, if you prefer Latin version, \"cum hoc ergo\npropter hoc\").  The causation might be both ways: gurus don't find\ngit hard to learn, or don't remember that it was hard to learn; people\nwho find git hard to learn do not become git gurus.\n\n-- \nJakub Narebski\n\nhttp://git.or.cz/gitwiki/GitSurvey2009\n"},{"id":"121150","messageId":"20090818195550.GB12938@vidovic","threadId":"20625","inReplyTo":"200908181900.03356.jnareb@gmail.com","subject":"learning curve (aka Git User's Survey 2009 partial summary, part 2 - from first 10)","fromName":"Nicolas Sebrecht","fromEmail":"nicolas.s.dev@gmx.fr","sentAt":"2009-08-18T19:55:50Z","receivedAt":"2009-08-18T19:55:50Z","isPatch":false,"sender":{"key":"nicolas.s.dev@gmx.fr","avatar":null},"body":"The 18/08/09, Jakub Narebski wrote:\n\n> > I believe it would be interesting to know who (from the question 6.)\n> > think what later. We may expect that people of the grade 4 and 5 (\"can\n> > offer advice\" and \"know it very well\") underestimate the difficulty to\n> > learn Git.\n> \n> Well, you can do some analysis yourself, using just created filters at\n> http://www.survs.com/shareResults?survey=2PIMZGU0&rndm=678J66QRA2\n\nThank you very much.\n\n> For example:\n> \n> ================================================|\n>                   |    proficiency with Git     |\n> To learn          |-----------------------------|\n>                   | any*[%] | 1-3 [%] | 4-5 [%] |\n> ------------------------------------------------|\n> Very easy         |     4%  |     3%  |     5%  |\n> Easy              |    20%  |    19%  |    22%  |\n> Reasonably easy   |    55%  |    55%  |    56%  |\n> Hard              |    19%  |    20%  |    16%  |\n> Very hard         |     2%  |     3%  |     1%  |\n> ------------------------------------------------|\n> Total respondents |   3005  |   1786  |   1212  |\n> ================================================|\n\nThis looks consistent. Here are some more grained summaries based on\nyour filters:\n\n\n                      =========================================================\n                      |                       TO LEARN (%)                    |\n===============================================================================\nProficiency           | very easy | easy | reasonably easy | hard | very hard |\n===============================================================================\nno filter (any)       |      4    |  20  |       55        |  19  |     2     |\n――――――――――――――――――――――|―――――――――――|――――――|―――――――――――――――――|――――――|―――――――――――|\n1. novice             |      3    |  12  |       49        |  25  |    11     |\n2. casual             |      2    |  17  |       54        |  25  |     3     |\n3. everyday use       |      4    |  21  |       56        |  18  |     2     |\n4. can offer advice   |      4    |  23  |       56        |  17  |     1     |\n5. know it very well  |     11    |  20  |       54        |  14  |     1     |\n===============================================================================\n\n                      =========================================================\n                      |                       TO USE   (%)                    |\n===============================================================================\nProficiency           | very easy | easy | reasonably easy | hard | very hard |\n===============================================================================\nno filter (any)       |      9    |  37  |        45       |   8  |     1     |\n――――――――――――――――――――――|―――――――――――|――――――|―――――――――――――――――|――――――|―――――――――――|\n1. novice             |      5    |  17  |        47       |  25  |     7     |\n2. casual             |      5    |  27  |        51       |  15  |     2     |\n3. everyday use       |      8    |  36  |        47       |   8  |     1     |\n4. can offer advice   |     11    |  44  |        40       |   4  |     0     |\n5. know it very well  |     22    |  37  |        36       |   5  |     1     |\n===============================================================================\n\nWhat appears here first is that the averages of the third category\n(\"everyday day use\") are _very_ near from the averages of the whole users.\nThis gives a good confidence in the results.\n\n\nSimplified summary:\n\n                      ===============================================\n                      |                  TO LEARN (%)               |\n=====================================================================\nProficiency           | (very) easy | reasonably easy | (very) hard |\n=====================================================================\nno filter (any)       |     24      |        55       |      21     |\n――――――――――――――――――――――|―――――――――――――|―――――――――――――――――|―――――――――――――|\n1. novice             |     15      |        49       |      36     |\n2. casual             |     19      |        54       |      28     |\n----------------------|-------------|-----------------|-------------|\n3. everyday use       |     25      |        56       |      20     |\n----------------------|-------------|-----------------|-------------|\n4. can offer advice   |     27      |        56       |      18     |\n5. know it very well  |     31      |        54       |      15     |\n=====================================================================\n\n                      ===============================================\n                      |                  TO USE   (%)               |\n=====================================================================\nProficiency           | (very) easy | reasonably easy | (very) hard |\n=====================================================================\nno filter (any)       |     36      |        45       |       9     |\n――――――――――――――――――――――|―――――――――――――|―――――――――――――――――|―――――――――――――|\n1. novice             |     22      |        47       |      32     |\n2. casual             |     32      |        51       |      17     |\n----------------------|-------------|-----------------|-------------|\n3. everyday use       |     44      |        47       |       9     |\n----------------------|-------------|-----------------|-------------|\n4. can offer advice   |     55      |        40       |       4     |\n5. know it very well  |     59      |        36       |       6     |\n=====================================================================\n\n\n-- \nNicolas Sebrecht\n"}]}