Permalink
Commits on Jul 8, 2016
  1. #399 finally put together a collection of tests that provide pretty c…

    …omplete cover/characterization of the BaselineOf load behavior for different combinations of package versions ... the upshot is that with standard packages, packages with the same version are not loaded (incorrectly assumed to be the same) the cypress package unconditionally loads the package regardless of version --- of course with the way Monticello packages nothing is actually done if the definition lists are the same. See #399 (comment).
    committed Jul 8, 2016
Commits on Jul 7, 2016
  1. #399, #400 add tests cases for these bugs ... with the test cases, I'…

    …ve shown that Metacello ignores the package-cache when deciding whether or not to fetch a package from a FileTree repo so at least Metacello will ignore the package-cache during Baseline loads ... did not explore Configuration loads, but they need to follow the mcz repo rules ... Consequently it isn't absolutely necessary to have any special package-cache handling for Cypress packages, but I'm going to leave the code in place
    committed Jul 7, 2016
  2. #399 fix typos

    committed Jul 7, 2016
  3. #399 for 2.4.8 need to override allTestSelectors as well ... also sta…

    …rt worrying about portability to non-GemStone platforms
    committed Jul 7, 2016
  4. #399 Metacello-Cypress dependencies not quite right and test groups f…

    …or Travis* not defined correctly either ...
    committed Jul 7, 2016
  5. #399 no ill effects from using Cypress packages for all baselines ...…

    … probably should have been done from beginning ... add class var so indvidual developers can make the descision ...
    committed Jul 7, 2016
  6. #399 ... let's find out what happens when all baselines use the cypre…

    …ss scheme ... massive test failures? Probably
    committed Jul 7, 2016
  7. #390 assertion failure during tearDown appears to be culprit for 2.4.…

    …8 failures ... take steps
    committed Jul 7, 2016
  8. trim the lineup

    committed Jul 7, 2016
  9. hpi-swa/smalltalkCI#176 try removing onWarningLog, since it can't be …

    …used with Pharo-3.0 and Pharo-4.0 and oddly enough Squeak-trunk
    committed Jul 7, 2016
  10. does order matter?

    committed Jul 7, 2016
  11. enable full lineup for grins and add the travisCI script I was using …

    …to profile the test runs
    committed Jul 7, 2016
Commits on Jul 6, 2016
  1. #390 fix two failing tests:

     Metacello.TestsMCB.MetacelloScriptingWarningTestCase
      testOnWarningLogDuringLoad
      testWarningDuringLoad
    committed Jul 6, 2016
  2. #390 cache and undo the undefinedSymbols handling while running Metac…

    …ello tests that remove classes from the system during cleanup or the tests themselves ... should see significant speedup --- for GemStone --- will have to work out platform independent scheme if this trick pays off ...
    committed Jul 6, 2016
  3. #390 take an alternate tack - leverage some newer features and bugfix…

    …es for SmalltalkCI to see if a) correct tests get run b) the tests run snappy enough (specifically bugfixes: hpi-swa/smalltalkCI#172 and hpi-swa/smalltalkCI#174)
    committed Jul 6, 2016
  4. Metacello-GemStone-TestsMCB was missing from 'Tests' group which coul…

    …d lead to slightly different results depending upon how things were loaded
    committed Jul 6, 2016
Commits on Jul 5, 2016