Voting system

[Tibor Zimányi]

Hi everyone,

from the start I started stewarding the community recently, I see there is no clear workflow as far as decision making about PRs etc. go. I think it would be a good idea to incorporate some kind of voting system to unblock some things, like old PRs.

Reasoning:

• If there is agreement about topics, we can move forward.
  
• There may be disagreements about some topics - and that is actually a good thing as that can raise concerns and help make the TCK better. However this often leads to the discussion getting stuck and not proceeding because we don't know how to continue with the topic.
  
• There may be situations when there is no feedback for a longer time. 

To avoid these things, I propose to define a voting system, where some defined majority of votes would be needed to make a decision about a PR or other topics.

If you agree, we could start defining the system. What do you think about it please? We could do something similar as it is done in the OMG group. That at the end of each month, there will be a ballot prepared from the actual open topics that were discussed. That will leave some space for discussions on GitHub or on the monthly meeting and at the same time the decision making will be timeboxed.

What do you think please?

Best regards,

Tibor

[Octavian Patrascoiu]

Hi Tibor,

The pattern followed so far in TCK when making changes (DMN proposal or TCK tests) is the following:

               All agree => the change goes in 

               Disagreement about tests => issue with OMG RTF to clarify the semantics

Kind regards,

Octavian

--
You received this message because you are subscribed to the Google Groups "DMN" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dmn-communit...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/fdd64b44-031f-4e84-a975-5d9414c2cc70n%40googlegroups.com.

[Greg McCreath]

Heya Tibor,

Waiting on the RTF is surely very painful, but, I think it is likely the best policy in this case for 'unknowns'.  We should include tests for things we're very sure about and when we're not - seek clarification (IMO).  In some cases there is a 'by convention' angle but there have been very few of those as I recall.

It does make it difficult when the spec is clearly just silly in some places and everybody ignores it - so knowing what is right and wrong is a bit grey.  Here are some examples that we have ignored for tests, and vendors have ignored in their runtimes/compilers.  Sadly ...

* If a decision does not require any business knowledge models, its value expression must be a literal expression or decision table that specifies the entire decision logic for deriving the output from the inputs.

... sooooo, if a decision does not use BKMs then the decision expression can only be literal or a decision table and cannot be a context or invocation or relation or list or other stuff ...

and ..

* If a decision element requires more than one business knowledge element, its value expression must be a literal expression that specifies how the business knowledge model elements are invoked and how their results are combined into the decision's outcome.

sooo ... if a decision uses more than one BKM the only type of expression it can have is a literal expression ... what?!

These are totally ignored - and yes, there was a JIRA raised for it which went in the bin or something.

I have this image of the DMN / RTF 'committee' as like all covered in dust sitting about on old leather chairs hardly moving at all and avoiding sunlight ... :-) :-). heh heh, and like the key outcome of a meeting is when to have the next meeting.  :-). <smiling>

I guess we could still use the vote thing and unless in agreement we can seek clarification  ... perhaps some way can be found to speed up the RTF clarification process?

Greg.

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CADRK%3DXNk4GXyc5yWzV_TJ%3D_N%3DT2Wt0S2%2B_cnU%2BxPtWzZb0wA_g%40mail.gmail.com.

[Tibor Zimányi]

Ok, thanks for the feedback. I think it makes sense what Octavian and Greg wrote. So let's keep it that way.

T.

Dátum: streda 30. novembra 2022, čas: 13:33:13 UTC+1, odosielateľ: gr...@montera.com.au

[Denis Gagne]

Hi Greg,

From one of those useless RTF members that you refer to, thank you for bringing these issues to my attention.

I will raise these issues to the RTF and see that they are corrected.

BTW, anyone can report issues to the RTF using this form: https://issues.omg.org/issues/create-new-issue

 

If you have such better working habits than the current RTF committee, why don’t you just join us and show us how it is done.

We can always use one more person that thinks they know more then anyone else <<smiling>>

 

Cheers!

Denis

 

From: dmn-co...@googlegroups.com <dmn-co...@googlegroups.com> On Behalf Of Greg McCreath
Sent: November 30, 2022 12:36 AM
To: dmn-co...@googlegroups.com
Subject: Re: [DMN] Voting system

 

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.

 

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAE5s5WJsnoj91ANmJD6iPdNzGN_ZRCP8PxX2zC%3Dc7b%2BEN9wsbg%40mail.gmail.com.

[Greg McCreath]

Hi Denis,

My apologies.  Sincerely and honestly, it was meant to be humorous - and that sometimes falls flat on its face in text - looks like it fell way short here.  Apologies again. 

Regarding those particular issues, I see you've raised them as separate JIRAs just now - thank you.  I did raise them here also last Feb: https://issues.omg.org/issues/DMN15-65.  In essence,  it is recommending those oddities are simply removed from the spec.  There is a little more to it than just the two examples in my email below.

I did look into OMG membership, but, if I recall, the minimum fee to join was (at the time) something like $2.5k or so and a membership with voting rights was maybe $6.5k. 

You'll see below that I am actually supporting the RTF system.  It may be a little frustrating at times, but it is currently the best way to tackle unknowns (IMO) and the proper way to channel change requests - which I do also.  I have just started on the tests for 1.4 and already see stuff that needs JIRAs - which I'll raise.

Again, apols for coming across as I have done.  I'll rope it in.

Greg.

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/DM5PR02MB33056FEAA6E0A413F40C1C3ED2159%40DM5PR02MB3305.namprd02.prod.outlook.com.

[Denis Gagne]

Hi Greg,

No arms done. I was just responding in the same sarcastic tone as your original note 😉

 

BTW I will be happy to play your surrogate in the RTF when you have issues and proposals that come along with them.

Just let me know and I will see to get these moving along.

 

Which DMN implementation are you working on and for which organization?

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAE5s5WKi5og2cp%3D1MRaTK%2Buy_JY-%2BjGvxcDkLKbb0oSqeEijAA%40mail.gmail.com.

[Greg McCreath]

Thanks Denis, "touché" as they say.   Sarcasm volume knob here turned way down.

Re the RTF surrogate role - that sounds amazing.  Though (please chime in here Tibor and Matteo) - I don't know if any members here are also are on the RTF and are already taking that defacto role. Unsure. But, however it works out, having that role within the RTF sounds fantastic.  We have a couple of PRs tagged with 'waiting on RTF' and a fair few JIRA issues raised as part of TCK discussions.  We also have a fair few tests commented out pending some of those issues being resolved at the RTF level.  Oftentimes, we have commented out the in-doubt tests, raised an issue and kept moving.  On the basis that it is better to get _some_ coverage then none on some given spec aspect.  As I'm not a OMG member, nor privy to DMN discussions at that level, I can really only approach the spec at face value and write tests that exercise it as I read it.  Sometimes, this is not the same as the 'intention' of the spec.  The discussions around these points are healthy and (IMO) quite invigorating and this is often where the JIRA issues come from.

Re DMN implementations, well, that is probably more a 'beer' topic as to what has happened over the last few years and a tale of dreams/love/family-health yada yada, but at this very moment, my involvement and contributions here are non-commercial.  I'm quite a believer in what the DMN spec seeks to achieve.  Sadly, I think the world would be a better place with a standardised decisioning mechanism.  Eg, in a previous role I was involved in a process to select such a thing and the process itself cost about $2.5M AUD and the end result was exactly zero - no selection. Why? All too hard and ultra proprietary - example = Oracle OPA.  DMN was not even on the radar, nobody knew about it.  I found out about DMN afterwards when a few of us exited that place (a bank) and set about to create our own decisioning engine - the one that would have got the business in that instance and 100 other times we'd all heard about "business rules engines".  Our goal involved a little more heuristics and handling of 'unknown' values with some business-friendly backwards chaining to get some more interactivity in finding missing model data, but, DMN was a pretty good fit for what we were looking to achieve - and standardised, so we pivoted to DMN and left 'unknowns' behind.  However ... there is only me left!

So, despite being a "card-carrying" standardised decision engine zealot techie weirdo, I also feel participating here is some small contribution back to the world.  Plus, I am a little anal about testing and quality.  So, hence ending up writing tests for DMN in my spare time.  It sure ain't a great conversation starter.  But, I know the value of a good TCK as I was involved in a small embedded JVM for payments devices back in the Sun days when Java and its TCK felt like a gift to humankind.  So, I have a benchmark.

I guess as an aside and for full disclosure, I have two impls for the TCK tests - Java and TypeScript - with the TS impl running in node/browser + some react/SVG based UI tooling for diagramming.  But, nothing commercial as yet.  I've seen your toolset on YT - so you do not have anything to fear! :-). 

Greg.

[Matteo Mortari]

Re the RTF surrogate role - that sounds amazing.  Though (please chime in here Tibor and Matteo) - I don't know if any members here are also are on the RTF and are already taking that defacto role. 

<snip>

FWIW, for as long as I was part of the OMG DMN RTF, I've tried to best advocate (there) for any issue we detected here in the TCK which also had a proposal, and making sure to follow-up (in both groups).

Falko as one of the RTF chairs, I believe appreciated and supported --as the vast majority of the group.

As Denis often advocates in many places, providing not only a report but also a proposal for resolution, has been a way more effective strategy than when we initially started the TCK back in the initial days.

For the OMG jiras I've raised personally I've made sure to mention credit where it was due, something like ~"this issue was detected by Greg in the DMN TCK group", I am very positive there were quite a few, whose initial jira's description had that footnote marking.

I am very conscientious about attribution/credit.

I believe we were successful in having quite some of them clarified not only in 1.4, and the same for previous DMN releases.

I sympathise with preferring a speedier process, but for as long as I was stewarding, I wanted to keep the original direction for the TCK, and on other hand I need to appreciate in the RTF there is more than just execution concerns.

Ciao,

MM

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAE5s5W%2BM00GDM0YG7pTkpmAZ7%3D%2BV1nWjCg0YO-RK5v2ZHzmrvA%40mail.gmail.com.

--

Matteo Mortari

linkedin.com/in/matteomortari

youtube.com/MatteoMortari

[Greg McCreath]

Hi Matteo,

I wasn't sure of exactly how the roles were played out - but, I really (certainly) assumed you were in there batting for the TCK - hence the prompt to chime in!  I have also seen the "detected in the DMN TCK group" as well in your raised issues, so not missed at all, and certainly appreciated.  

As for your stewardship, you have alway been level, passionate (overused, but true here) and forthright, and as you say, aware that DMN is more than just about execution - which, in the TCK world, is a particular focus!  So, the balance is welcome.

May I ask if you still have a role in the DMN/RTF?  If that is not up for disclosure, or in possible transition then np - all good.  No answer required and "no comment" is okay.

Greg.

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAHOoiEqEx9o0v9RU_1-6jKHmzhO3%2B9yWX-_ttUbJtsv_FTJqVA%40mail.gmail.com.

[Matteo Mortari]

May I ask if you still have a role in the DMN/RTF? 

<snip>

Unfortunately I am personally no longer part of the OMG DMN RTF.

On the other hand, it is my understanding that others from IBM have joined the RTF.

You can find more information here, here and here(or here).

Ciao,

MM

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAE5s5WKy_Xd8_La6ut6iG-%3D0KW%3DXxbOMjqxyp5Gxvqqah8vgpA%40mail.gmail.com.

[Tibor Zimányi]

Hi everyone,

Matteo you can freely name me :) Greg, I recently joined the RTF representing IBM. As far as RTF members, we have multiple in our group. I want to continue what Matteo was doing with the concerns discussed in TCK. However with multiple members of the RTF group here, all of them are welcome to bring the topics from TCK to the RTF group. And vice versa. I would recommend also for the RTF members (including me) to bring topics to our TCK group if there is a question/concern how some things are implemented, or if there is a need for a broader opinion, because not all vendors are part of the RTF group.

Thank you everyone for this discussion. As I joined this community, I had the feeling it is a "bit" dead. Seeing this thread makes me think the opposite.

Best regards,

Tibor

You received this message because you are subscribed to a topic in the Google Groups "DMN" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/dmn-community/QnpAl2gVWkY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to dmn-communit...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAHOoiEr-3Z6-JW%3D4iw109PhQJ%2BJLTHAp3HEL4DPMQ1sQWr5jdQ%40mail.gmail.com.

--

Tibor Zimanyi

[Greg McCreath]

Thanks Matteo,

Good to get that squared.  Btw, I liked the goatee.  "Being back the goatee" :-)

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAHOoiEr-3Z6-JW%3D4iw109PhQJ%2BJLTHAp3HEL4DPMQ1sQWr5jdQ%40mail.gmail.com.

[Greg McCreath]

"Matteo you can freely name me" made me laugh out loud.  Righto.  You're on the spot now. ;-). It is _excellent_ to see that this project has such representation at the RTF level - that is a very heartening thought.  

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CADBd1DR8h-a_B2KQ-%3D7RXU_oGMi7Dg3p_Y%3Dx74Hs%2BPXPSLkJww%40mail.gmail.com.

[Denis Gagne]

Excellent.

Matteo did an amazing job both as TCK chair and RTF member. We all ow him a lot of gratitude and recognition for the professional manner in which here carried out his duties.

I am certain Tibor will do a great job as well.

The work of the TCK is very important to ensure a balance between the work of the RTF and what the market requires.

 

As mentioned by Matteo, the best kind of raised issues are those that include a detail proposal on how to solve them.

 

@Greg McCreath here there are other such issues with detailed proposal that are lurking in the RTF please make sure to let me know (or anyone else on the RTF) and we will try to streamline them through the RTF process.

 

Thank you all for your work.  We are making a difference in the market.

To view this discussion on the web, visit https://groups.google.com/d/msgid/dmn-community/CAE5s5W%2B3G99T%3DY2z8XkkG_%3Dkpt7SHmRR03SjKvMB1tmxc%3D2NGw%40mail.gmail.com.