From: Greg.mpls@... Date: 2018-10-27T13:21:21+00:00 Subject: [ruby-core:89576] [Ruby trunk Misc#15239] [patch] test-spec win32ole Issue #15239 has been updated by MSP-Greg (Greg L). suke (Masaki Suketa) wrote: Thanks for looking at it. Apologies for not testing on Travis. > I'm afraid your patch would fail on environment except Windows (on Linux, etc). Sorry, I thought I had 'windows guards' on everything, I may have missed a few. Thanks for looking at it, as the Appveyor CI is having issues with the IE object. Appveyor build/test times vary a great deal, so I felt this was needed to lower the 'resource needs' of these tests. I've never had an issue with them (using IE) running locally... ---------------------------------------- Misc #15239: [patch] test-spec win32ole https://bugs.ruby-lang.org/issues/15239#change-74622 * Author: MSP-Greg (Greg L) * Status: Open * Priority: Normal * Assignee: suke (Masaki Suketa) ---------------------------------------- Some of the current Win32OLE spec tests use the InternetExplorer control. This control is not available on Azure pipelines. Attached patch changes to use the MSXML control, adds a guard for it, and also does some refactoring to lower requires when not run under Windows, etc. Also, since the MSXML object is quite a bit 'smaller' than InternetExplorer, tests run faster. Passed in my fork at https://ci.appveyor.com/project/MSP-Greg/ruby/builds/19663145 Note also that since there have been recent commits stabilizing the spec suite, the above build job changed the mswin spec tests to run parallel, and they passed. The change from serial to parallel is not included in the patches. It is a separate commit in the branch on my fork. GitHub patch is at: https://github.com/MSP-Greg/ruby/commit/d1daf9a66f491abfac5e84a50f152505baa1ccac.patch ---Files-------------------------------- spec_win32ole.patch (25.8 KB) -- https://bugs.ruby-lang.org/ Unsubscribe: