Fixed a few UI issues with creating Auto-Script expected results.
The parameter count shown in plan revisions prior to today (March 30th) will be "trailing" for plan updates that change the number of parameters in the plan (parameter additions, deletions, some bulk edits). Trailing means the revision shows the parameter count before, rather than after the update.
This issue is now resolved, and the parameter count display will no longer be trailing for newly created revisions. This issue was only with the display of the revision, not a problem with the revision itself, or with reverting back to the revision.
Mind maps are a great way to visualize the variation in your test plan. They're perfect for reviewing your plan with subject matter experts or developers that aren't very familiar with Hexawise, and they're also useful for management reports, presentations, and brainstorming sessions. Consider using a mind map anywhere you need to talk about and collaborate on your test plan.
Many of you have already been using desktop mind mapping tools and Hexawise's OPML import/export to edit and visualize your test plans as mind maps, but now it's no longer limited to people with extra tools.
Hexawise now supports interactive mind maps right in your web browser. Just click the "Mind Map" link in the center panel of the "Define Inputs" page of any plan to see your test plan as an interactive mind map. Click the black errors to make edits to your plan from within the mind map, and click the "Save" button when you are finished.
Have fun!
Two separate issues, one with canceling a computation invoked from the analysis view, and another with changing strengths from the scorecard tab. Both are resolved.
Copy revisions checkbox more clearly associated with the label.
This is no longer the case.
We are proud to launch today the biggest, and most often requested single change in Hexawise since it was created, test plan versioning!
With Hexawise's new versioning, every change you (or your colleagues) make to a test plan is tracked and versioned. Click on the "Revisions" link at the top of any application page to see the revision history of the current test plan: what changed, when, and by whom. You can revert a test plan to any prior revision with a single click.
Due to the nature of how plan versioning works, versions are only available for new changes to your test plans going forward.
We're not yet done with this feature, and will soon be launching version labeling and branching (creating a copy from a prior version).
We'd love to hear your feedback on Hexawise versioning.
The performance for exports with lots of test cases has been dramatically improved.
Fixed some edge cases around moving the last step down or first step up. Also fixed the scrolling and focus behavior after clicking to add an expected result to a step.
In most cases Hexawise provides some details about the cause when preventing the creation of a value pair that would cause a "no possible value" parameter value in a test case. However there were some cases where no details were provided, and these have been eliminated.
The Hexawise Browser and Network Test (available at /network-test) has proven very useful in supporting users dealing with browser or network configuration issues so we improved the styling and usability considerably and linked to it from the login screen (where browser and network issues are usually first surfaced).
Mixed-strength was not an option for very large test plans that only support generating 2-way tests. Unfortunately this also meant the option to select 1-way strength for some parameters was missing. This has been addressed, so now you can select mixed-strength and 1-way parameters for very large test plans.
There is extensive logic in Hexawise to prevent many of the simpler cases of overly constrained test plans that would otherwise result in tests with a "no possible value" for a parameter value. This logic in some cases generates implied constraints that are needed, and in other cases disallows the creation of the paired value constraint that would cause the "no possible value".
There was a defect in this logic that could cause the order the constraints were created in to matter, and that could cause the type of constraint to matter (so a married pair and its logically equivalent set of invalid pairs weren't treated the same).
These issues have been addressed.
There is an auto-completion feature to the parameter name drop down combo box that lets you type the name of a parameter used in another plan, select it, and reuse the parameters and values in your new plan. Unfortunately this only worked if you started typing the parameter name from the beginning. Now you it will match anywhere in the parameter name, e.g. typing "master" will match both "Master Record", "Record Master", and "Platinum mastercard".
We now hold onto the cached generated tests in more circumstances where the changes being made don't affect the tests. This should speed along refining your test plan in some cases. Zoom! Zoom!
Thanks to Katie for the suggestion.
In some cases the plan notes dialog could appear each time you access "Define Inputs" rather than just when the plan is opened.
The share a project level up achievement now has additional information about how to share a project.
Hexawise detects when you're using an old browser that's not supported. This could be confusing if you're using a modern version of Internet Explorer in compatibility mode (which causes IE to report itself as, and to act as IE 7). There's now explanatory information in the old browser warning and the FAQ about not using compatibility mode (or incompatibility mode in this case!) in Internet Explorer.
Normally commas separate parameter values during bulk edit. Commas can be still included in the parameter value if they are escaped with a back slash . The warning about having more than 8 or more parameter values would appear when you had less than 8 parameter values if the values had escaped commas in them.
This was a regression from the large framework upgrade.
This was a regression from the large framework upgrade.
The keystrokes to copy the text were treated as a change. They aren't any longer.
Thanks to Katie for the suggestion!
This was a regression from the large framework upgrade.
Addressed an error in a small number of cases.
Should result in some minor performance improvements.