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

Re: RFE: "git bisect reverse"

From
HAH. Peter Anvin <hpa@zytor.com>
Date
May 28, 2009, 21:54 UTC
Message-ID
<4A1F07FA.6040008@zytor.com>
In-Reply-To
<efe2b6d70905281407x56bb788aq3dba4b27eb91d7a6@mail.gmail.com>
Ealdwulf Wuffinga wrote:
Show 7 quoted lines
> 
> It would fit a lot better into this algorithm to have some model of
> the probability that a commit will cause a skip. It doesn't actually
> have to be a very good one, because if it's poor it will only make the
> search slightly less efficient, not affect the reliability of the
> final result.
>

How about simply modelling it linearly, with 100% probability for known skip point, 0% for a known good/bad point, and a linear gradient in between? It's probably a good enough model. In practice, it will vastly overestimate the probability of a skip, so if a linear model turns out to be too conservative, I would probably just try to model it as a higher-order power function.

	-hpa
Previous: Ealdwulf WuffingaNext: Ealdwulf Wuffinga
Message 14 of 18 in “RFE: "git bisect reverse"”
  1. H. Peter AnvinMay 26, 2009
  2. Sam VilainMay 27, 2009
  3. H. Peter AnvinMay 27, 2009
  4. Christian CouderMay 27, 2009
  5. Ealdwulf WuffingaMay 27, 2009
  6. Clemens BuchacherMay 27, 2009
  7. Ealdwulf WuffingaMay 27, 2009
  8. Sam VilainMay 27, 2009
  9. Ealdwulf WuffingaMay 28, 2009
  10. Sam VilainMay 29, 2009
  11. Ealdwulf WuffingaMay 31, 2009
  12. H. Peter AnvinMay 28, 2009
  13. Ealdwulf WuffingaMay 28, 2009
  14. H. Peter AnvinMay 28, 2009
  15. Ealdwulf WuffingaMay 31, 2009
  16. Christian CouderMay 27, 2009
  17. Nanako ShiraishiMay 27, 2009
  18. Matthieu MoyMay 27, 2009

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.