1. 16 Apr, 2021 4 commits
  2. 15 Apr, 2021 2 commits
  3. 14 Apr, 2021 12 commits
  4. 12 Apr, 2021 1 commit
  5. 09 Apr, 2021 1 commit
  6. 08 Apr, 2021 5 commits
  7. 01 Apr, 2021 2 commits
    • Harald Sitter's avatar
      add missing includes for freebsd · 42aca174
      Harald Sitter authored
      every time I use errno I forget that it requires an include lest it
      fails on freebsd.
      
      also add string.h for strerror for good measure
      42aca174
    • Harald Sitter's avatar
      refactor deleted file detection · 1de2c484
      Harald Sitter authored
      instead of using procfs' /map_files use /maps. this will allow us to
      also process data after the process had gone away (e.g. when processing
      data from coredumpd) and makes no difference for the running process.
      
      the parser is now in its own file and has a unit test
      
      there's also new tech in place that additionally adds stat support. this
      is also to aid in future development for coredumpd support. if a library
      has changed in the time between the core dumping and drkonqi starting,
      then we need a way to figure out if the file has changed even when we
      can't rely on the ' (deleted)' marker in the /maps data.
      the easiest way to achieve this is to compare the inodes. while that
      isn't always accurate it's probably accurate enough for our purposes.
      1de2c484
  8. 30 Mar, 2021 1 commit
  9. 26 Mar, 2021 1 commit
  10. 08 Mar, 2021 2 commits
  11. 07 Mar, 2021 1 commit
  12. 05 Mar, 2021 1 commit
    • Harald Sitter's avatar
      unbreak button states · e929d4c1
      Harald Sitter authored
      versionpage used KAssitantDialog's setAppropriate but that triggers
      internal re-evaluation of button states as per the appropriateness of
      adjacent pages of the current page.
      this then would short circuit when a premature conclusion page was shown
      (because the user provided no or insufficient information to file a
      report) and result in the back button getting enabled even when that
      wasn't reasonable (i.e. finished was emitted with showBack=false).
      
      notably this broke for none-kde addresses that aren't routed into
      bugzilla iff bugzilla responses are so slow that they'd arrive after the
      user reached the conclusion page
      
      the fix for this is super awkward. instead of using the existing
      appropriateness system, use a higher level one on our end. this also
      requires us to be able to iterate the page sequence. to achieve that we
      now put all pages in a vector in the sequence that they are in the
      pagedialog. on page changes we then establish whether we were moving
      forward or backward and this way skip over inappropriate pages.
      
      (could have used the hash, turn it into a map and use that instead, of a
      new vector but honestly I want the hash to go away and we'll manage the
      hand full of extra bytes until then)
      e929d4c1
  13. 04 Mar, 2021 7 commits
    • Harald Sitter's avatar
      map kdeinit5 · 4a5270c7
      Harald Sitter authored
      for the rare occasions that kdeinit5 crashes, rather than a thing loaded
      by it
      
      BUG: 433483
      FIXED-IN: 5.12.3
      4a5270c7
    • Harald Sitter's avatar
      wire up productInfoError to fall back to 'kde' product · 6360b56b
      Harald Sitter authored
      when a product fails to resolve in bugzilla we'll want to fall back on
      the generic 'kde' product so the user can still conveniently report the
      bug and get it triaged.
      
      this most notably deals with scenarios where a kaboutdata product/binary
      does not have its own bugzilla product and also has no mapping data in
      our file (e.g. kdeinit5 is part of the frameworks-kinit product, without
      mapping data we'll not able to file it there and the submission fails)
      
      when product mapping now fails it will signal an error, the error is
      connected to the fallback logic, which simply switches the product to
      'kde' and retries the resolution
      
      there's also a hint about why this bug was filed against 'kde' being
      added to the report summary so people don't get confused and also to
      perhaps prompt them to fix the mappings file
      6360b56b
    • Harald Sitter's avatar
      don't assert that there is 1 product · c3f6cd23
      Harald Sitter authored
      ...there may be none!
      
      we'll throw this as an exception to trigger error signaling in
      bugzillalib.cpp. Alas, as it turns out this isn't actually wired up to
      anything. so this effectively just turns the failed assert (which may
      crash) into a runtimexception (which never crashes).
      
      to reasonably deal with this the wizard will need some improvements as
      well
      c3f6cd23
    • Nicolas Fella's avatar
    • Harald Sitter's avatar
      move UI setup logic out of checkTexts · 38582b8b
      Harald Sitter authored
      it has no business being there, it is... not checking text.
      
      specifically this resolves a problem whereby the new initial focus
      setting was switching focus while the user types.
      
        "thou shall not covet thy neighbor's focus"
      
      CHANGELOG: input focus is no longer dancing tango with two text fields
      BUG: 433480
      FIXED-IN: 5.21.3
      38582b8b
    • Harald Sitter's avatar
      refine warning message · 454baf7e
      Harald Sitter authored
      it's a bit daft to log not valid but then print isValid -> remove the
      negation from the stringy portion
      454baf7e
    • Harald Sitter's avatar
      log why debugger entered failure state · 3b8d672c
      Harald Sitter authored
      we conveniently overload FailedToStart with also meaning IsNotInstalled
      and IsNotValid making it impossible to tell what's wrong from just
      looking at the GUI
      3b8d672c