Is there a web app for Gospel Living?

[russellhltn]

I think they tried open-source a long time ago. I get the impression that standing up a sandbox (so they weren't working with actual member data) was more work then the benefit they were getting from development.

[scgallafent]

I think they tried open-source a long time ago. I get the impression that standing up a sandbox (so they weren't working with actual member data) was more work then the benefit they were getting from development.

We have bounced around what would be required to successfully create a sandbox with fake data. The cost is astronomical. You look at it and think "how difficult could it be to set up some ward and stake data?" but the Church stretches much further than wards and stakes.

[russellhltn]

but the Church stretches much further than wards and stakes.

Well, there's branches and missions, as well as branches in stakes. You've got bishops out-of-unit bishops, and acting bishops. You need to also have a member in the data that fit into every category - single, married, widowed, widowed with temple sealing, married for time after becoming a widow and still sealed to prior spouse, etc. Not to mention the child that was baptized two days before his 8th birthday and had to have the ordnance ratified by the First Presidency. Just coming up with the list of corner cases could be a fun shooting-the-breeze exercise.

GL would limit how much depth that would have, but still, callings and parental connection still play a big part in how the software works. (Let's see, there's children living with parents, children living with a parent and step parent, children living with non-parents, children who are the only member of household ...)

[scgallafent]

Don't forget branches in districts and branches that aren't in missions, districts, or stakes. Plus you have to update the dates or else you end up with no Primary, but then some of your test cases will break. You got out-of-unit bishops but missed out-of-unit members. Now you're getting warmed up.

For Gospel Living, we might want to customize content based on the Church area that you live in or the country you live in, so there needs to be a decent set of test cases that cover both. You also need to handle all of the organizations that exists in wards and stakes so that the appropriate circles are created. That means we probably want a separate sandbox of the infrastructure for managing circle membership and a copy of the infrastructure for managing the in-app messaging.

If you give a mouse a cookie....

[idefaye]

I understand how complex this can get, but at the same time, dev/QA MUST have that fake data already to perform their test.
I CANNOT believe they are working with real data, that would be a legal issue in so many countries, so the fake data MUST already be available. Or at least with scrubbed data...

[idefaye]

On the other hand, the data could be generated on a per need basis for a given project by the members working on it.

[eblood66]

If you give a mouse a cookie....

Indeed. Back when the church was trying to use volunteer developers, we started to create a system to allow developers to create fake data for testing and development. I was the volunteer project manager and the main volunteer developer. But it was such a big undertaking that it never really got past just handling the basics much less these many edge cases. Plus we never were able to recruit any other volunteer developers on that project (at least any that really did anything) and I could only devote 2-3 hours a week. And the church developers were trying to actually get the (at the time) new directory and calendar apps developed and didn't have time to work on this system. So it eventually got scraped.

[scgallafent]

The reality is the data can't be sufficiently scrubbed so that it is completely anonymous without building an entire fake church. The project eblood66 described did a reasonable job of creating fake member data, but it wouldn't have been able to support the needs of today's applications.

I'm not going to convince you that we have looked at the problem and determined there isn't a feasible way right now to set up a set of systems that would sufficiently support community development efforts. You're not going to convince me that it's easy to support and we should get it working so that community developers can get involved. We'll just agree to disagree.

[sbradshaw]

Plus you have to update the dates or else you end up with no Primary

I suppose when a pseudo-member reaches the "age of a tree," they could be recycled as a new Primary child. Then add some random generators to add ordinances to certain percentages of the ward and switch callings around every so often…

[russellhltn]

Plus you have to update the dates or else you end up with no Primary

I suppose when a pseudo-member reaches the "age of a tree," they could be recycled as a new Primary child. Then add some random generators to add ordinances to certain percentages of the ward and switch callings around every so often…

I'd think a "Groundhog Day" method would be easier. At some point between sandbox testing, you'd just increment everyone's birthdate. Poor Cody would be stuck in Deacon's for a decade.