Thanks @bhaugen
Learn more about recipes?
Or just follow the Hruby book and start implementing immediately?
I am still trying to learn more abt REA and valueflows via reading and asking questions see which way suits my problem set better.
At the same time, I'm experimenting with what I have read in REA book and from valueflows in my actual client projects as a way to learn as well.
- ask questions here
- send email to hruby asking him abt linkage pattern and aspects
- experiment with a bit of linkage pattern in my client project but using a lot of postgres jsonb field to add more flexibility.. i did use your terminology in the slide where you use the term "Actual" to denote the actual resource versus resource type. I like that.
@simkimsia we have the book, you can just cite page numbers if you want.
Oh roger that!
I am still trying to learn more abt REA and valueflows via reading and asking questions see which way suits my problem set better.
In general, if your problem is for a single company or an enterprise with multiple divisions but still in a mother-children structure, the Hruby book might be better.
If your problem is for a network of multiple independent agents, VF might be better.
how abt i try to narrow down using a more specific example as a way to test my understanding of VF?
I have some ActualResources: Network, Project, JobRole, OrgUnit, Activity. These were given to me as a table in a spreadsheet by the client.
I couldn't be sure of the exact relationship across these 5 different resource types even when given actual data, but I do know they are somehow related. I can make a guess, but everytime i code the database schema with this particular client project based on my guesses they turn out to be wrong and result in a lot of painful refacotring.
Which motivated me to search for more flexible solutions hence I stumble on REA and VF
There's a biz question that needs to be answered , which is given the identifiers of JobRole, OrgUnit and Project, return the associated Network.
I am trying out Linkage pattern from the Hruby book (page 119-121) and adapted it for my use. Instead of having a parent, i simply have a ActualResourceLinkage datatable that stores all the associated identifiers across Network, Project, JobRole, OrgUnit, Activity under postgres jsonb field
In other words 1 tuple or set of associated Network, Project, JobRole, OrgUnit, Activity records is represented as a single JSON dict in the JSONB field in a single data row in the postgres datatable ActualResourceLinkage. I just tried it out and it kinda worked though I am not sure what kind of possible errors I might encounter further down the road.
And I do seem to depart from linkage pattern based on the book. I wonder if VF has a better way to solve this problem I face
In either case, it's a dependent demand explosion: http://mikorizal.org/dependent_demand.pdf
Thanks. I will read this in greater detail this weekend
Instead of having a parent, i simply have a ActualResourceLinkage datatable that stores all the associated identifiers across Network, Project, JobRole, OrgUnit, Activity under postgres jsonb field
That seems different from Hruby's Linkage pattern which is a parent-child relationship.
What is an Activity?
I suspect it's some internal labeling in fact you can think of all the 5 resources of various kinds of labeling which frankly is abstract to me and even to the end users
what they want me to do is somebody (say Alice who is an Admin) in my system will type in the full associated data which I have screencaped above
and later on in a different screen, somebody else in a different role (say Romeo in Resource Manager) will create a Quotation and in the individual Line Item key in info like Project identifier, JobRole identifier, OrgUnit identifier and the system should auto fetch the associated Network and display. Subsequently, they will save that Quotation. Of course my Quotation and LineItem need to save their chosen Project, JobRole, OrgUnit, and Network at the LineItem level amongst other information.
This client has a complex workflow which nobody to date has shown me an actual end to end workflow on paper. it's all in their heads and parts of it done manually. I am helping them to automate parts of it.
Do they have anything like a work breakdown structure?
they do have a WBS but nothing to do with generating the Quotation. Their WBS is more for implementing for their client. I know this because one of another abstraction which I have not mentioned here yet is WBS number which will come from their SAP system
For now, my job is to create this quotation system. I can do it without any regard of REA, or VF. In fact, I have already created a different set of Quotation features for a different type. That has gone live in production.
Let's call this first set Quotation Type A. In Type A, the Network is associated with completely different resources such as CommonSite, Scope, Project, and Activity.
This set of Quotation I have mentioned earlier is Type B. As you can see the association is completely different. I am unable to reuse a lot of code I have previously written.
I am wary of creating brittle systems, should I encounter change requests or new user stories that break my assumptions.
hence my exploration into REA and VF
