Skip to content

Either drop file/size from add/autoadd OR add md5 and sha to it. #262

Closed

Description

@aelsabbahy

Need to make a decision on which one of these attributes are included in add and which are not:

  • Size
  • md5
  • sha256

Activity

  1. pysysops commented on Aug 6, 2017

    @pysysops
    Contributor

    SHA-256 may be the most sensible and current attribute to default to? Especially with the recent (last year or so) posts about SHA-1 / MD5 collisions.

  2. aelsabbahy commented on Aug 6, 2017

    @aelsabbahy
    MemberAuthor

    If one of them were to stay sha256 makes the most sense.

    I'm just not sure if sha256 by default is a useful addition or an annoyance (generated test too rigid).

    I can go either way on this one. Drop them all, add them all, or only keep sha256.

  3. elliotweiser commented on Aug 6, 2017

    @elliotweiser
    Contributor

    When deciding on defaults, it's often best to go the "simpler" route.

    IMO, it's best to use none of those checks and to let the user decide what he/she wants to add.

    Perhaps this opens the door on supporting additional flags to goss add for the file test, e.g. --checksum. Thoughts?

  4. aelsabbahy commented on Aug 6, 2017

    @aelsabbahy
    MemberAuthor

    So the opposite exists today in Goss with --exclude-attr. There could be an --include-xtra-attr global flag that does this.

    That said, to your point about "simpler." It really comes down to what would the majority of users expect from a goss add file. My gut tells me @pysysops is right with just sha256, but no size.

  5. elliotweiser commented on Aug 6, 2017

    @elliotweiser
    Contributor

    I think it comes down to whether or not a file is expected to change very often. While I enjoy a special love for the sha256 test, I cannot claim one way or the other that static files are the target of the majority use case.

  6. aelsabbahy commented on Aug 7, 2017

    @aelsabbahy
    MemberAuthor

    Yeah, that's what bother me about size. It's such a flakey test, but with even less value than sha256.

    Maybe less is more. Keep all the "extra" tests as manual.

  7. elliotweiser commented on Aug 7, 2017

    @elliotweiser
    Contributor

    FWIW, I wouldn't be opposed to switching out size with sha256 to see what kind of feedback it generates.

  8. aelsabbahy commented on Dec 14, 2019

    @aelsabbahy
    MemberAuthor

    Deciding to drop size

  9. added 2 commits that reference this issue on Dec 27, 2020
    9bd080e
    427348a
  10. added a commit that references this issue on Jan 27, 2024
    2e5720b
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions

      ⌂ 换网址