The monthly SAS IT 3WM outage takes place this weekend (August 19-21). Therefore, JMP DevOps is planning not to start Stringbot runs on the new tracks till next weekend.
Welcome to the JMP Development Update for August 19, 2022!
As some JMP teams focus on new features for JMP 18 and bug fixes for 17.1, others continue their work to wrap up JMP 17.0. Doc teams are creating and checking final PDFs and Help and have handed off book source files to the L10N team for delivery to translators. JMP QA teams in Cary and Beijing are preparing for final installer candidate creation (planned for 8/23) and validation on a matrix of different JMP and OS combinations.
We didn’t meet as a dev group this week, but I do have an empty dev meeting slot next Thursday 8/25 if anyone is interested in presenting. Hope you can join on 9/1 to see a preview of John’s plenary talk for Discovery! And on 9/8, we will meet for a demo of the new Marker Simulation platform by developer Luciano Silva.
Please don’t forget that the SAS level 1 security training is required for JMP R&D staff and others who are involved in product development. The training is due to be completed by the end of August. I know many of you have already started or completed the three modules of training, but others have yet to enroll. You can register and begin taking the training in SABA here. I’m on my last module myself - let’s all try to finish that up in advance of the deadline. Thanks very much!
General Announcements
- If you’re willing to be a featured employee on Inside JMP, please download a copy of the Get to Know form. Simply fill out the questions and when you’re done, upload it, along with 3 photos of you to the folder. Thank you for helping to make our JMP family a little more familiar!
Important Upcoming Dates
- First JMP 17 Final Installer Candidate – 8/23/22
- Please visit Confluence for a full list of JMP 17 dates.
JMP Development Updates
- Please visit this page to view bug verification lists for JMP 17.0.
Statistics
- Elizabeth: Fixed an issue where via scripting a nominal Y could be used with the Poisson distribution in GLMM (JMP-13807); that might be ok if the underlying data is numeric but not otherwise, so it’s caught now (it was always disallowed interactively). Fang found an issue with a small fraction of rows’ conditional CIs in the random grouped regressor case. We were being too strict with a check in the degrees of freedom calculation, so relaxed the criterion a bit (JMP-13827). Investigating estimation issues for an example from Ono-san with the Poisson distribution where block-to-block variability is large sending our results off towards the infinite (JMP-13766); it appears to be a starting value issue.
- John Sall: Fixed crash in Explore Outliers for K Nearest Neighbors if number of observation was exactly 8 (JMP-13789). Fixed crash for Explore Outliers when Exclude button clicked and auto-recalc was on; the excludes triggered deleting the object that it was inside the code of. There were four separate methods to fix. The key breakthrough was found by Dave Barbour using stack traces from the crash reporting system (JMP-13830). Sped up Response Screening by multi-threading robust scaling; one case with many rows became 4 times faster. The bottleneck was not noticed earlier because most examples of very wide data did not have a lot of rows (JMP-13781). Attended JMP18 planning meetings.
- Jianfeng: Started to go through my feature and address older items. Here are some features that I have completed lately: JMP-7967, JMP-13214 and JMP-7994. Doing research on equivalence test for paired data and nonparametric equivalence test as well.
- Laura C-S: Fixed a couple bugs with the starting values of loadings in SEM. Started researching and implementing the instrumental variable method with 2SLS for latent variable regressions so we can use that for improving those starting values next. Participated in strengths workshop. Answered a couple questions on SEM-MGA from Yusuke Ono.
Stat Documentation
- Michael C: Reviewed stat areas of online help and PDF help for 17.0 (with Colleen). Completed security training modules. Discussed doc branching plans with Qiana, updated doc reports to show 17.1 work, and started triaging backlog of process-related doc issues in Jira.
Stat Testing
- Mark: Focus on UI/UX workstream activity. Conducted second usability test of Easy DOE with Sarah Sylvestre. Worked with UX Development Process team towards Usability Lab at JMP Discovery Summit next month. Converting JMP journal version of usability test script to scripted application for moderated and self-study usability tests of Easy DOE. Attending Product Security Training.
- Jacob: Cleaning Easy DOE unit tests. Participating in the UI/UX workstream. Verifying Jira tickets. Writing new tests for Explore Outliers, focusing on cases with excluded rows.
- Tonya: Testing the new ZI SHASH distribution. Working on blog posts for new features in 17. Bug verification.
Product Management
- Bryan F:
- Easy DOE Usability Testing
- Coordination with Kaylee and Jacob Rhyne about usability test
- Processed usability test results from Marilyn Wheatley
- Began processing EasyDOE usability results with UX Development Process Workstream members
- Discovery Summit Usability Logistics
- Communicated with Jessica about InsideJMP article requesting volunteers
- Provided training for Discovery Summit Usability lab volunteers
- Created SignUpGenius page for volunteers to choose suitable time slots
- Communicated with Tanya about the image for the testing machines
- Communicated with Robin and Holly about providing swag to usability participants
- Further communication with legal (Jeffrey Smith and Wanda Allen) about NDA
- Communicated with Evan, Aurora, Erik, and Aaron about usability scripts
- Working on logistics for external usability test of Bivariate platform
- Investigated using userinterviews.com to recruit participants
- Communication with Alex and Dieter about creating a virtual JMP environment for usability testing
- Roundtable/mini CAB for "Looks are Everything"
- Provided feedback on JMP Live publish dialog
Technical Support
- Duane: I am slowly getting back into the swing of things but haven’t updated my scripts for the TS JMP Live reports. I’ll be working on that soon! We had an interesting morning, Tuesday, where 1,000s of duplicate cases were created in Serviced Cloud due to an infinite email loop and out-of-office messages. Candice was able to end the cycle and delete the dups. “Real” cases continue at a steady pace, as shown in the plot below.
Development Operations
- Jim: Setting up new replacement UT machines. Preparing for the JMP 17 SODA installer on macOS. Preparing for the new JMP 18 per file change coming next week. Perforce support.
- Shannon: Investigated mystery crashes from crash database with help from Dave B. Reviewed prevalent crashes for 16.X and 17.0. Working through security training.
Documentation
- Susan C: Working on Student Subscription doc with Ross. Closing or deferring doc JIRAs for the EG and JMP Live. Fixed issues with help keys in Genomics doc. Provided some feedback to Tom. Troubleshooting cut-off equations issue in help.
- Chris B: Finalized removal of Description markers from FrameMaker files. Checked the JSL help and PDFs for errors before submitting their final 17.0 versions. Created the JMP Documentation Library PDF. String Bot review practice. Applied our bandaid CSS fixes to 17.0 help.
System/Host Testing-Cary
- Nick H: Updated Ernest on the preferences work I have done in hopes to come up with a plan for JMP 18 to make preference names between GUI and JSL more consistent. Fixed an issue with some socket tests that weren’t converted quite correctly from the C++ unit tests stream to JSL, which was causing intermittent failures in the unit tests. Filed a few bugs: Uncovered a crash in the Window Handler unit tests on Mac in (only) JMP 18, found an issue that prevents users from scrolling a data table on Mac when hovering over a scroll locked column, and an issue with Custom Local Settings() preferences expression. Worked on a number of 18 verifications pushing for 17.1 integration adding JSL unit tests where needed.
- Audrey: Prep tasks for moving to Atlassian Cloud, organizing Atlassian Admin training and previewing one of the self-paced learning courses, working with Lisa on JSL scripts and graphs about our Jira activity, final prep for final installer validation tasks next week.
- Mandy: Verifying changes at 18.0 (Labels to codes script, Tabulate, Formula evaluation, etc) that developer wanted to consider pushing back to 17.1. Testing a new If/Else Workflow concept added to 18.0 that might be considered for 17.1. Installed 17.0 JMPPRoTrial and found a Firefox JMP Help issue that may be resolved now.
- Mike M: Image diff catch-up, updating JMP Help Checklist for upcoming installer testing, verifying defects.
System/Host Testing-Beijing
- Eric Wang: (1) Refactored end-to-end tests for control chart warning notification. Added new common functions to add subscription on report/folder/space. Added new common functions to update warning setting on group/report. Re-wrote control chart warning notification tests for new scenarios. (2) Verifying the fix of JMP pool hang/crash issue.
- Da Li: (1) Bugs verification which are related to OSIPI, MFI and Export-Excel component. Also verified JMP Live bugs. (2) Create a critical bug during UT failure investigation: JMP-13784 Google sheets import does not work (3) Provide comments/reply to Audrey about OSIPI issue from customer.
- Mike Wu: (1) Completed JMP 17 Help/Doc review. (2) Completed JMP 17.0 OLE VB Automation monthly testing. (3) Completed JMP Clinical 8.1 quarterly check. (4) JMP/JMPLive 17 bugs regression testing.
- Kimi: (1) Worked with developer to investigate [JMP-13721] JMP crash when executing "Quit ("No Save")" on Mac. (2) JMP now uses the latest Azul JVM. Tested the new changes, see [JMP-13277]Update to latest Azul JVM. (3) Fix ut failures for Formula Depot tests. (4) Sanity check for Data-Client. (5) Jenkins: Debug the failure of UT environment on Windows. Debug the failure of UT environment on Mac. Working on support python init/connect testing (via python tool).
- Nala: (1) Verify Jira tickets.
- JMP Jira: 4 Bugs / Stories created ( Link ), 20 Bugs / Stories verified ( Link)
- JMP Live Jira: 4 Bugs / Stories created ( Link ), 13 Bugs / Stories verified ( link )
Data Discovery
- Dan S: Fixed selection filter clearing other row states, fix Mosaic plot axis issue in Distribution, fixed focus issue allowing disabled buttons to trigger via keyboard input. Looking at new changes to BarSeg.
- Ernest: Pushed changes to Workflow builder code generation and journal creation to support if/else/return. Working on refactoring preferences to support import/export.
- Dave B: Looking into a couple of different anonymous crashes, fixed mfi crash with password protected excel file, getting jmprunner ready for per file change.
- Sibusiso: General refactor and understanding of the tabulate code, fixes for date increment in JSL, and implemented a Getter for Minor Tick count in JSL.
DOE
- Peng: Updated density drawing codes in Distribution platform, addressing one of the two issues reported in JMP-4273, S0640353: Density curves on a log scale are strange. Research work with Bill Meeker. Making a UI prototype for DOE. Made changes for newly reported issues in Life Distribution with Grouping.
- Caleb: Last week, I was at JSM and had a great time catching up with colleagues and listening to some good talks. This week, I’ve been fixing a few more bugs with the sample size explorer and working on paper revisions for resubmission.
HTML5
JMP Live
- Aurora: 17.1: Simplified a confusing error message. 18.0: Using database transactions more consistently. Removing support for the flagging/moderation feature. Miscellaneous: Helped a small amount with Eric Hill's Proposed Usability Study for Discovery Summit 2022. Other prep for Discovery. Updated a woefully out of date article on JMP Live Settings.
Note from Shannon
Thanks to John Sall, who passed along a picture from Brad Jones of the biggest “little red triangle” I’ve ever seen - a present given to Brad’s close collaborator and JMP superfan professor Peter Goos from his students.
Hope you all have a great weekend!
Love that biggest little red triangle!
