It's all about the answers!

Ask a question

Builds and Build Definitions without Source Code?

Glyn Costello (14046) | asked Sep 21 '21, 3:15 a.m.


Can build definitions be utilised in CLM (especially RQM) when not working on a Software project? i.e. when working on a physical product, perhaps to manage modification status of the equipment?

Has anyone got some good examples of how this could work?



Accepted answer

permanent link
Davyd Norris (2.4k217) | answered Oct 11 '21, 6:24 p.m.
Yes absolutely but I think, as Ralph points out, you may be mixing your terminology...

A Build Definition is used almost exclusively in EWM to set up an automated build process of some sort. This is where it's most useful. In ETM you can create them but they are really just placeholders - if you've created them in EWM then you'll be able to see them in ETM and that's quite useful, but when you create a Build Definition in ETM it's really only used as a grouping mechanism for Build Records.

Build Records are the really useful thing in ETM and I have several clients that use these in non-software projects to show a BoM and configuration for the system under test:
 - one client uses these to describe the mechanical and hardware/electronics assembly that was put together for the testing to be run on. The Build Record includes rev numbers for the key assemblies
 - one client uses this to include the provisioning record for the equipment that went into a substation. This also include firmware revision info for the PLCs and other controlling systems
 - one client uses it to describe the environment that has been constructed for a particular V&V phase of a rail system, in particular for things that are not present or that have been stubbed out at that time
Glyn Costello selected this answer as the correct answer

Glyn Costello commented Oct 12 '21, 7:26 a.m.

Do you have a high level approach you can share regarding that first client example?

Davyd Norris commented Oct 12 '21, 7:50 p.m.
It's pretty simple, but we had to spend some time to figure out the best way to represent everything because these guys make a complex product with lots of parts and sometimes they test sub assemblies in an already assembled system.

The key here was that the build record should only include the parts you're actually testing, and the test cell should include everything else. It gets even more complex when the parts under test in one phase then get put into service in a test cell to test other bits in the next phase!

Basically all they did was to create a build record, name it according to what the assembly was and an ID that was stuck on it at the end (eg. Power Supply Unit #0012). Then in the description they listed all the main parts with any rev number and part number:
Power Transformer 12AU S/N:12445
Main Supply Board rev B P/N:23456B
Monitor Board rev A P/N:89065

Finally they assigned someone to assemble it and they marked the record as complete when done

One other answer

permanent link
Ralph Schoon (63.2k33646) | answered Sep 21 '21, 4:08 a.m.

 As far as I can tell, at least for EWM, the build definition itself does not provide a lot of value. The Build definition is only referenced within the build loop in EWM. You need a build result that has been created using the build definition to achieve anything in EWM.

At least within EWM, Build Engines and build definitions and build results can be used to automate anything that can be executed and keep an auditable track of the result. 

I do not know where build definitions show up in RQM and what they are used for. 

Glyn Costello commented Oct 11 '21, 7:16 a.m.

But in general can builds be used for non-software items? Like documentation, drawings, spreadsheets, etc. stored in RTC SCM?

Ralph Schoon commented Oct 11 '21, 9:07 a.m.

Builds execute one or typically more steps to produce an outcome. You can use them for whatever you want, with or without accessing the SCM system. If there is automation available, it can potentially be included. E.g. generation of documentation (e.g. JavaDoc), running some tool to extract and compute information, routing, maybe pulling in mechanics data.

Think of build being a chain of automation that is executed and creates a set of result files from some input files/data (and often automatically checking for quality). 

The build result is something that usually ties the input and output together, to make it reproduceable. It is also possible to store the total build input and result data. Usually this would be done in a binary repo such as Artifactory and not EWM SCM. 


Ralph Schoon commented Oct 11 '21, 9:08 a.m.

For storing the cumulated build data, see some thought here:

The build engine and build tool is just an example.  

Your answer

Register or to post your answer.

Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.