This is a bug fix release, with a few minor enhancements included in the process.
P180 – Fix minor UI bug in TextBroker service
When the number of digits differed in the min length and max length fields, the form failed. The workaround was to add a leading zero to the minimum length field. This is now fixed.
P192 – Add custom MTurk qualification types as option in HIT definition
This is more an enhancement than it is a bug fix. We now support both system qualifications (locale, approval rate, etc) as well as custom qualifications – a topic to be covered during Inner Circle.
P200 – Make Clone change the dates of clones to the current date
Clone was preserving the change date of the input document on all clones. This is at least confusing and incompatible with the rest of the system, so it is now changed to date the clones with the date and time they were made.
P205 – Improve feed reading
We have had significant error rates on reading of RSS and Atom feeds. We’ve updated some core software to newest versions and modified properties on the HTTP interface. In testing, this has greatly improved our error statistics.
P216 – Fix the way https feed items are shown in UI
In the document list view, any document with a name that looks like a URL is made clickable. This was true of http, but not https. That was simply a mistake that has now been corrected.
P217 – Enable sequences of MTurk workflows
There was a bug in the MTurk processing for all three MTurk workflows that prevented correct operation when the input to the workflow was itself a document that was created in a previous MTurk workflow. This has now been fixed so you can chain together sequences of MTurk flows.
P218 – Make MTurk-T2 workflow function properly in production
There was a typo in the T2 workflow that caused it to fail in production. This is now fixed.
P219 – Change “Copy of” rename behavior
When a file is copied into a folder and the file name is already taken, the system created a unique name by prepending ‘Copy of’ to the name. What a workflow caused this to be repeated the system applied the rule recursively (Copy of Copy of Copy of <nam>, etc) until a unique name was found, with the obvious limitation that this can no go on forever!
The algorithm is now changed to create a unique name that does not growth with each iteration.
P221 – Improvement in script execution
Several small changes have been made to the script workflow. Most significantly, we’ve doubled the allowed script length to 3800 characters. We’ve also reduced the max runtime to 15 seconds and added debugging messages to the system logs so that we can more easily diagnose user scripting problems.
Leave a Reply