git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: git version statistics

From
Jeff King <peff@peff.net>
Date
Jun 1, 2012, 08:52 UTC
Message-ID
<20120601085211.GB32340@sigill.intra.peff.net>
In-Reply-To
<981b755b-ca86-4320-a4fc-8aa28caa099d@mail>
On Thu, May 31, 2012 at 11:20:46AM -0400, Stephen Bash wrote:
Show 14 quoted lines
> > The interesting thing to me is how spiky it is, and where the spikes
> > fall. I would expect to see a spike around the highest maint release
> > of each major version (so v1.7.8.6, for example, with many fewer
> > installs of v1.7.8.5, v1.7.8.4, and so forth). But that's not what
> > happens.  The most popular v1.7.8.x versions are .3 and .2, and hardly
> > anybody bothered to move to v1.7.8.6.
> 
> I wonder if the spikes correlate with time between releases?  For open
> source projects I tend to download the most recent when setting up a
> new machine (or after encountering a nasty bug), and then not upgrade
> for a while.  So in that mode of operation, releases that are "the
> newest" for the longest would get more users... (though at least on my
> Mac the homebrew project is helping me break the habit and stay more
> up-to-date)

Good point. If you assume that users pick a random day to upgrade or install and choose the latest version, then you will get a non-uniform distribution. Because the release dates are non-uniform, their time spent as the latest is not even. There may also be natural variations in installations over time (e.g., over holidays).

Here are the release dates for the v1.7.8.x series, as well as the adjacent master releases:

  v1.7.8   2011-12-02
  v1.7.8.1 2011-12-21
  v1.7.8.2 2011-12-28
  v1.7.8.3 2012-01-06
  v1.7.8.4 2012-01-18
  v1.7.9   2012-01-27
  v1.7.8.5 2012-02-26
  v1.7.10  2012-04-06
  v1.7.8.6 2012-04-26

So .2 and .3 were latest for 9 and 12 days, respectively. However, .4 was also the latest for 9 days (until v1.7.9 came out), but does not have as many users. So why did nobody bother upgrading to v1.7.8.4? And why wouldn't v1.7.8 have a spike, since it was at the top for 19 days?

I can see why v1.7.8.5 and v1.7.8.6 are the way they are (they were never latest, and most people would just install v1.7.9 or v1.7.10 instead).

So I think your theory probably explains some of the data, but not all (and it seems that most people don't really seem to care about old maint releases once a new master release is out).

-Peff
Previous: Stephen Bash
Message 84 of 84 in “git version statistics”
  1. Jeff KingMay 31, 2012
  2. Jeff KingMay 31, 2012
  3. Junio C HamanoMay 31, 2012
  4. Jeff KingJun 1, 2012
  5. Junio C HamanoJun 1, 2012
  6. Jeff KingJun 2, 2012
  7. Tomas CarneckyJun 2, 2012
  8. Jeff KingJun 2, 2012
  9. 1/4 move git_version_string into version.cJeff King, Jun 2, 2012
  10. 2/4 version: add git_user_agent functionJeff King, Jun 2, 2012
  11. Thomas RastJun 19, 2012
  12. Jeff KingJun 19, 2012
  13. Jeff KingJun 19, 2012
  14. 1/3 Makefile: apply dependencies consistently to sparse/asm targetsJeff King, Jun 19, 2012
  15. Junio C HamanoJun 19, 2012
  16. 2/3 Makefile: split GIT_USER_AGENT from GIT-CFLAGSJeff King, Jun 19, 2012
  17. Junio C HamanoJun 19, 2012
  18. 3/3 Makefile: split prefix flags from GIT-CFLAGSJeff King, Jun 19, 2012
  19. Junio C HamanoJun 19, 2012
  20. Jeff KingJun 19, 2012
  21. Junio C HamanoJun 19, 2012
  22. Jeff KingJun 19, 2012
  23. Junio C HamanoJun 19, 2012
  24. Jeff KingJun 19, 2012
  25. 0/8 makefile cleanupsJeff King, Jun 19, 2012
  26. 1/8 Makefile: apply dependencies consistently to sparse/asm targetsJeff King, Jun 19, 2012
  27. Jonathan NiederJun 20, 2012
  28. Jeff KingJun 20, 2012
  29. Jonathan NiederJun 20, 2012
  30. Jeff KingJun 20, 2012
  31. Jeff KingJun 20, 2012
  32. 01/11 Makefile: sort LIB_H listJeff King, Jun 20, 2012
  33. Junio C HamanoJun 20, 2012
  34. Jeff KingJun 20, 2012
  35. 02/11 Makefile: fold MISC_H into LIB_HJeff King, Jun 20, 2012
  36. Junio C HamanoJun 20, 2012
  37. Jonathan NiederJun 20, 2012
  38. Jeff KingJun 20, 2012
  39. 5/11 Makefile: fold XDIFF_H and VCSSVN_H into LIB_HJonathan Nieder, Jul 7, 2012
  40. Junio C HamanoJul 9, 2012
  41. Jonathan NiederJul 6, 2012
  42. 03/11 Makefile: do not have git.o depend on common-cmds.hJeff King, Jun 20, 2012
  43. Jonathan NiederJun 20, 2012
  44. 04/11 Makefile: apply dependencies consistently to sparse/asm targetsJeff King, Jun 20, 2012
  45. Jonathan NiederJun 20, 2012
  46. Jeff KingJun 20, 2012
  47. Makefile: document ground rules for target-specific dependenciesJonathan Nieder, Jul 7, 2012
  48. 05/11 Makefile: do not replace @@GIT_USER_AGENT@@ in scriptsJeff King, Jun 20, 2012
  49. Junio C HamanoJun 20, 2012
  50. Jeff KingJun 20, 2012
  51. 06/11 Makefile: split GIT_USER_AGENT from GIT-CFLAGSJeff King, Jun 20, 2012
  52. Jonathan NiederJun 20, 2012
  53. Jeff KingJun 20, 2012
  54. Jonathan NiederJun 20, 2012
  55. 06/11 Makefile: split GIT_USER_AGENT from GIT-CFLAGSJonathan Nieder, Jul 7, 2012
  56. 07/11 Makefile: split prefix flags from GIT-CFLAGSJeff King, Jun 20, 2012
  57. Jonathan NiederJun 20, 2012
  58. Jeff KingJun 20, 2012
  59. 08/11 Makefile: do not replace @@GIT_VERSION@@ in shell scriptsJeff King, Jun 20, 2012
  60. 09/11 Makefile: update scripts when build-time parameters changeJeff King, Jun 20, 2012
  61. 10/11 Makefile: build instaweb similar to other scriptsJeff King, Jun 20, 2012
  62. 11/11 Makefile: move GIT-VERSION-FILE dependencies closer to useJeff King, Jun 20, 2012
  63. Jonathan NiederJun 20, 2012
  64. Jonathan NiederJun 20, 2012
  65. Jeff KingJun 20, 2012
  66. Jonathan NiederJun 20, 2012
  67. Jeff KingJun 20, 2012
  68. Jonathan NiederJun 20, 2012
  69. Automatic dependency tracking in the Git build system (was: Re: [PATCHv2 1/8] Makefile: apply dependencies consistently to sparse/asm targets)Stefano Lattarini, Jun 21, 2012
  70. Junio C HamanoJun 20, 2012
  71. Thomas RastJun 20, 2012
  72. Jeff KingJun 21, 2012
  73. Junio C HamanoJun 21, 2012
  74. 2/8 Makefile: do not replace @@GIT_USER_AGENT@@ in scriptsJeff King, Jun 19, 2012
  75. 3/8 Makefile: split GIT_USER_AGENT from GIT-CFLAGSJeff King, Jun 19, 2012
  76. 4/8 Makefile: split prefix flags from GIT-CFLAGSJeff King, Jun 19, 2012
  77. 5/8 Makefile: do not replace @@GIT_VERSION@@ in shell scriptsJeff King, Jun 19, 2012
  78. 6/8 Makefile: update scripts when build-time parameters changeJeff King, Jun 19, 2012
  79. 7/8 Makefile: build instaweb similar to other scriptsJeff King, Jun 19, 2012
  80. 8/8 Makefile: move GIT-VERSION-FILE dependencies closer to useJeff King, Jun 19, 2012
  81. 3/4 http: get default user-agent from git_user_agentJeff King, Jun 2, 2012
  82. 4/4 include agent identifier in capability stringJeff King, Jun 2, 2012
  83. Stephen BashMay 31, 2012
  84. Jeff KingJun 1, 2012

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.