MicroProfile Mission: Is the goal of MicroProfile still "standardization"?

[John Clingan]

This question originated out of the technical whitepaper draft, and I subsequently brought up at the last Live Hangout.   We agreed to post the topic to the discussion forum. Here is the microprofile mission statement:

“An open forum to optimize Enterprise Java for a microservices architecture

by innovating across multiple implementations and collaborating

on common areas of interest with a goal of standardization.”

The question is whether or not our goal is still "standardization". When we first launched MicroProfile, a goal was to submit to a standards body, perhaps the JCP as we did with the Config specification. With Java EE moving to Eclipse Jakarta EE, is "standardization" still a goal? When I originally read the statement in the whitepaper, I read the word "standardization" literally with the target being a formal standards organization. I don't consider the Eclipse Foundation to be a standards organization, although a specification process is being defined under Jakarta EE (which blurs the waters a bit).

A comment made during the Live Hangout was that perhaps the word "standardization" doesn't have to be taken so literally. Ex: "defacto standard" is common language, but not a formal standard. So, we thought we would get community feedback to see if "with a goal of standardization" should still be a part of our mission statement.

Thoughts?

[m.reza.rahman]

I don't think it is so complex. The goal is not open source alone but some kind of standardization in the same vein as the JCP.

Sent via the Samsung Galaxy S7, an AT&T 4G LTE smartphone

--
You received this message because you are subscribed to the Google Groups "Eclipse MicroProfile" group.
To unsubscribe from this group and stop receiving emails from it, send an email to microprofile...@googlegroups.com.
To post to this group, send email to microp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/529d4e01-07f0-4f88-921d-96575afaed62%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[James Roper]

I don't have such a strict view on the definition of standard.

When two or more technologies agree to implement the same specification, they are standardising on the same specification, which implies that that specification is a standard for those two technologies. The process by which that specification is developed is irrelevant, whether it's a formal specification process from a standards organisation like ISO or the JCP, or whether it's just one person from each technology sitting down in private and agreeing on a specification, or whether it's one technology creating the specification and everyone else following, those things might impact whether it's considered for example, an open standard, formal standard, or a closed standard, but not whether it's a standard.

A defacto standard in software is a different thing. I have heard Kafka described as a defacto standard in messaging for microservices. It's not a standard proper because there's only one implementation, no one has agreed to implement a single specification, there's no standardisation to speak of. However, it is considered a defacto standard because everyone uses it - when a single technology has the vast majority of the market share, using that technology brings most of the same advantages as using a standard, your software is portable to every major technology in the market by virtue of there being only one major technology in the market, you don't need to skill up new hires in the technology because they will most likely have already used it because it's what everyone uses, etc. This is not MicroProfile, MicroProfile cannot be described as a defacto standard.

I would describe MicroProfile as an open standard. It might not be a formal standard, but it is nevertheless an open standard, since multiple vendors have standardised on a single set of specifications for their products, and the process by which these specifications are created is open.

To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/5bbec28f.1c69fb81.f84d1.e2ce%40mx.google.com.

For more options, visit https://groups.google.com/d/optout.

--

James Roper
Senior Developer, Office of the CTO

Lightbend – Build reactive apps!
Twitter: @jroper

[John Clingan]

Thanks Reza and James. Reza, I'm not sure the goal is "in the same vein as the JCP" anymore. I'm basing this both on the Live Hangout feedback and the way we have been executing over the last 18 months or so.  The way we have been operating seems to be more in the vein of what James has layed out - we just define specs and move forward. I find it kind of interesting that we have never considered ourselves a standard, yet I've heard some (that are not active within the community) refer to MicroProfile as a standard.

The Jakarta EE working group is working on a spec process, and we can see whether or not it can be applied to MicroProfile, much like the JCP was a "stamp of approval". It may be that the Eclipse Foundation becomes that "stamp of approval".

Anyhooo, thanks to both of you for the feedback.

On Wednesday, October 10, 2018 at 10:10:46 PM UTC-7, James Roper wrote:

I don't have such a strict view on the definition of standard.

When two or more technologies agree to implement the same specification, they are standardising on the same specification, which implies that that specification is a standard for those two technologies. The process by which that specification is developed is irrelevant, whether it's a formal specification process from a standards organisation like ISO or the JCP, or whether it's just one person from each technology sitting down in private and agreeing on a specification, or whether it's one technology creating the specification and everyone else following, those things might impact whether it's considered for example, an open standard, formal standard, or a closed standard, but not whether it's a standard.

A defacto standard in software is a different thing. I have heard Kafka described as a defacto standard in messaging for microservices. It's not a standard proper because there's only one implementation, no one has agreed to implement a single specification, there's no standardisation to speak of. However, it is considered a defacto standard because everyone uses it - when a single technology has the vast majority of the market share, using that technology brings most of the same advantages as using a standard, your software is portable to every major technology in the market by virtue of there being only one major technology in the market, you don't need to skill up new hires in the technology because they will most likely have already used it because it's what everyone uses, etc. This is not MicroProfile, MicroProfile cannot be described as a defacto standard.

I would describe MicroProfile as an open standard. It might not be a formal standard, but it is nevertheless an open standard, since multiple vendors have standardised on a single set of specifications for their products, and the process by which these specifications are created is open.

On Thu, 11 Oct 2018 at 14:25, m.reza.rahman <m.reza...@gmail.com> wrote:

I don't think it is so complex. The goal is not open source alone but some kind of standardization in the same vein as the JCP.

Sent via the Samsung Galaxy S7, an AT&T 4G LTE smartphone

-------- Original message --------

From: John Clingan <jcli...@redhat.com>

Date: 10/10/18 11:01 PM (GMT-05:00)

To: Eclipse MicroProfile <microp...@googlegroups.com>

Subject: [microprofile] MicroProfile Mission: Is the goal of MicroProfile still "standardization"?

This question originated out of the technical whitepaper draft, and I subsequently brought up at the last Live Hangout.   We agreed to post the topic to the discussion forum. Here is the microprofile mission statement:

“An open forum to optimize Enterprise Java for a microservices architecture

by innovating across multiple implementations and collaborating

on common areas of interest with a goal of standardization.”

The question is whether or not our goal is still "standardization". When we first launched MicroProfile, a goal was to submit to a standards body, perhaps the JCP as we did with the Config specification. With Java EE moving to Eclipse Jakarta EE, is "standardization" still a goal? When I originally read the statement in the whitepaper, I read the word "standardization" literally with the target being a formal standards organization. I don't consider the Eclipse Foundation to be a standards organization, although a specification process is being defined under Jakarta EE (which blurs the waters a bit).

A comment made during the Live Hangout was that perhaps the word "standardization" doesn't have to be taken so literally. Ex: "defacto standard" is common language, but not a formal standard. So, we thought we would get community feedback to see if "with a goal of standardization" should still be a part of our mission statement.

Thoughts?

--
You received this message because you are subscribed to the Google Groups "Eclipse MicroProfile" group.

To unsubscribe from this group and stop receiving emails from it, send an email to microprofile+unsubscribe@googlegroups.com.

To post to this group, send email to microp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/529d4e01-07f0-4f88-921d-96575afaed62%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Eclipse MicroProfile" group.

To unsubscribe from this group and stop receiving emails from it, send an email to microprofile+unsubscribe@googlegroups.com.

To post to this group, send email to microp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/5bbec28f.1c69fb81.f84d1.e2ce%40mx.google.com.
For more options, visit https://groups.google.com/d/optout.

[Emily Jiang]

I think the logical route to go for standardisation is via Jakarta EE if we want to put some specs through. It will be good if we can keep the same package name org.eclipse.microprofile.

Thanks

Emily

To unsubscribe from this group and stop receiving emails from it, send an email to microprofile...@googlegroups.com.

To post to this group, send email to microp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/529d4e01-07f0-4f88-921d-96575afaed62%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Eclipse MicroProfile" group.

To unsubscribe from this group and stop receiving emails from it, send an email to microprofile...@googlegroups.com.

To post to this group, send email to microp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/5bbec28f.1c69fb81.f84d1.e2ce%40mx.google.com.
For more options, visit https://groups.google.com/d/optout.

[Werner Keil]

I did have a conversation about that and Config in particular with Emily yesterday at JCON.

Based on impressions from the Jakarta EE Spec Committee, the situation does not seem entirely clear for MicroProfile yet. As another Eclipse Technology project (JNoSQL) is on good track and should contribute to a new Jakarta EE specification soon, in general it should be possible for some parts of MicroProfile, too. Maybe not for those that interact with other external specs like OpenTracing which are specified in different places like the Cloud Native Computing Foundation or similar, but for others that are not a wrapper around existing specs a standardization via Jakarta EE would make sense.

Some like the JAX-RS Client feel like in its case offering extensions and improvents to JAX-RS directly rather than adding a separate Jakarta EE spec makes a lot more sense.

I cannot say, if existing code may keep a package name or not. Based on the Java EE JSR situation, if an API like MicroProfile Config, Health or Monitoring had some industry momentum already, allowing backward-compatible package naming would make sense, at least in my opinion, but there could be reasons not to allow that for a new Jakarta EE spec, e.g. because "org.eclipse" is not where things are standardized.

Werner

[Ondro Mihályi]

I think that our goal was to contribute MP specs to something bigger than just the MP project, hence the "goal of standardization". In the beginning, we had implicitly JCP in mind. The situation has changed, but the goal is the same. Now we target bringing some specs into JCP (Config JSR), into Jakarta EE (most of the specs), and even JDK (reactive operators).

If we agree that "standardization" means contributing a spec to a bigger body than the MicroProfile project, it's dead simple - the goal stays the same and we can move on to other topics. No need to discuss whether Jakarta EE or anything else is a proper standards body.

--Ondro

To unsubscribe from this group and stop receiving emails from it, send an email to microprofile...@googlegroups.com.

To post to this group, send email to microp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/529d4e01-07f0-4f88-921d-96575afaed62%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Eclipse MicroProfile" group.

To unsubscribe from this group and stop receiving emails from it, send an email to microprofile...@googlegroups.com.

To post to this group, send email to microp...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/5bbec28f.1c69fb81.f84d1.e2ce%40mx.google.com.
For more options, visit https://groups.google.com/d/optout.

[Ken Finnigan]

In the Hangout I pushed back against using the phrase "goal of standardization" for a few reasons.

Firstly because standardization can have many meanings, and people take it the way they're familiar with. Which is why those that have dealt with Java EE in the past think JCP or the like, where as James' first thought is it implies defining a "common ground".

Secondly, with that phrase in the mission, does it mean the mission was a failure if every specification MicroProfile creates doesn't get proposed as a standard in some standards body? To me it implies standardization is the only goal, as it's the only goal mentioned in the mission, and therefore not achieving that goal results in failure.

I would prefer the phrase is either removed completely or at least worded in such a way that states standardization is a nice benefit if it happens, but not the only goal of the project.

Ken

[Ondro Mihályi]

I read the phrase again and it basically says that we collaborate together in MicroProfile with a "goal of standardization". In that sense, "standardization" includes common MP specs which are implemented by all involved parties. Otherwise the statement would just mean we collaborate but we don't aim to establish common (standard) specs and APIs.

So I think the statement is correct. I'm OK with rewording it, e.g. to use some other word to describe creating common specs and APIs (a.k.a "standardization"). But the statement needs to explain that our goal is to create common specifications. Bringing them to a higher level or a standards body is an additonal goal, but optional. The statement doesn't separate these 2 intentions and I think it's on purpose.

So I would leave the statement as is or reword it along what I wrote above.

--Ondro

[Marcia Peterson]

Ondor, I agree with you, as a Project Manager and somewhat new to the IT field, standards are built into the project as you go along, since the IT field already has standards this provides a framework to align the project with a goal in mind. However, I do enjoy everyones comment as well. Thanks.

Marcia Peterson

To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/c85d48a6-2c6f-444c-83fe-6cbf5603a0d5%40googlegroups.com.

[Emily Jiang]

I am ok with the current statement or something similar. According to the definition of Standaridation in Wikipedia:

Standardization or standardisation is the process of implementing and developing technical standards based on the consensus of different parties....

MP community has different parties involved. All specs we defined here are de factor standards.

Emily