Jakarta EE TCK, 4 May 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: TCK call #10

Description

May 4, 2022 Jakarta EE Platform TCK call #10. Minutes can be viewed via https://docs.google.com/document/d/1V1dDLJkd14EDRMPeuI0VzPtU4Lbli8FFBd1pLDLlOrY/edit#heading=h.4e96clexayoy

A

A

All right um so we're one minute in, but I think we should get started and others can join in as we go I wanted to. uh Oh, let me um share my screen. First, let's see.

A

A

A

Great okay, um I'm I'm very interested in uh hi amelia. Welcome.

B

Good morning, everyone.

A

um Very interested in um um what yeah, if anyone has any thoughts on when glass fish seven, it will pass all you know the full platform, the web profile, the um stand-alone.

A

uh You know, all the you know the new one and the existing standalone gck tests, and it's um something that um um yeah yeah that yeah I'm trying to understand better. And if anyone has um um you know, yeah has a guess as to a date um go. You know, you know yeah, no, you know it it. It doesn't hurt to make a guess, that's right or wrong. um um I I'd be most interested in any guesses. If anyone has any this year um so.

B

uh Scott, have we asked we can send a message via them via the forum to arjun. He does a lot of work on glass fish. So if anyone were to know something, it's likely him.

A

um Yeah, that's yeah that yeah that's welcome, as you know as well, um but you know just yeah it's sort of um um in the unknown really, um but if anyone has input whether they're, not you know here or or are, and um but yeah, we can certainly ask on any any form or mailing list, and um you know that that that's fine fine, it's hard to guess it's hard to to know and I'm trying to follow the trend based on test failures and um speaking of which, let me um let me just uh post the link to the wiki page in the platform tck that has the um the test results.

A

As you know, um as we go so, let's see that's yeah. We try to update that. You know.

B

A

And um yeah yeah as we go and um so yeah um if there yeah no guesses, that's fine! I I was just interested if anyone has any to share um and of course we're interested in when other other implementations will pass as well.

A

Let's see I don't I'm going to switch to the results for a second. Can you all see the results? I don't know if I shared my whole browser or just the agenda.

B

um It's got it's okay, but uh it's too small. If you can enlarge the page, that would be great. Okay,.

A

Let me do that.

A

So um I'm just going to scroll down to last week's which.

A

Let's see: okay, we don't have a lot of detail, but uh jdk 11, 17, okay, 17, okay, yep, that's that would be these and it's going to open up. Last week's jdk 17 results.

A

Oh actually, sorry! It's right there! Actually, I see it now, um yeah, so 5 000 failures with jdk 17., so I just wanted to cut to bring note to the fact that um running glass fish under jdk 17 yeah at the moment um we get more failures um um and um that you know that is, um let's see due.

B

A

Let's see, I will get the issue and get that up on the screen. There we go okay. So let me make sure this issue goes into the agenda too.

A

um Okay, um so it's so you know due to um like uh open glass, fish issue um that um is impacting our um jdk 17 testing, so um kind of an open question. You know we'll you know, will when you know, when will glass fish support um running on jdk 17? Isn't you know, is another issue really kind of related to? When will glass fish run?

A

You know pass all the platform tck tests, even just on jdk 11., but 17 seems to be its own separate question just because yeah, it's unknown, you know, will glass fish have to change a lot of components to fix issues like this one?

A

I don't know, but you know we're waiting for updates on on that issue. So it's it's it's possible that glassfish will run on 17.

A

And it's also possible that um sorry restate that it's possible it might take last fish team longer to run on 17.. um I don't. I don't really understand the impact of that yet on the on the um ee10 schedule, but the impact they have might be.

A

Perhaps other implementations might cover the jdk 17 testing in some fashion, yet to be determined what that means. I don't know what that means yet and that's something that would probably be discussed more on it on the platform um calls and it has been discussed a little bit, but um if other implementations are needed to run on 17 for the ee10 release before glass fish finishes their 17 support, um we'll have to figure out what that means, and so just kind of wanted to share those general.

A

You know, general, you know thoughts on it and just that it. There is work to be done and discussions to be had um about what that would mean, but um so that could be interesting for you know for um the other implementations to be able to consider but um yeah. I know it's a lot of work for everybody, so.

B

All right, scott, let me interject here: maybe it's not uh the right um window to ask, but for 10 it's okay, we're using glass fish, but we haven't. We have talked enough about neutrality and especially for testing. Is there any plan in the future for the releases that are coming up after 10 to use something else? Aside from glass fish? We are thankful to the glass fish right since 2005 when it came to life, but do we want it to continue to support it as we do now without diversity right?

B

There are very few implementations that use something else. You know we all need to push towards and move forward with glass fish, but what does it mean for the ecosystem? The lack of absolute um innovation on that side of the fence? We put all our efforts into last fish and there is no diversity.

B

Where can we do something like it doesn't mean that, like most most of the work needs to go into diverse vendors to say, hey, we will use this. This is a lot of work.

B

I'm not sure how others feel, but are these conversations be intact in the public, because I know we have them in the private communities a lot for many years, but I rarely see them coming into public and I don't see neutrality here at all since they called us recorded. I think it's a nice thing for us to to start being more comfortable discussing hey. Can we do this.

B

It's not about switch just did say, switch, I'm talking about adding other server that will help us with the tck. So switch is maybe um what is it called? Not the right word. I want to adding something else that could be worked towards to when we work with the tck. We change it a little bit here and there we're talking about um simplifying the tck, adding new tests and things like that. But if we continue to use only glass fish, what does it say about our technology and its future?

B

There's no diversity we need. We have to you know it's not only about features right.

A

um I I I I I I get the um yeah, I certainly get to you know, get the idea.

A

The um I mean the reason for 10.

A

You know you know to use glass fish as a primary target is because it implements all the optional technologies. So it's the only it's the only implementation that can pass all of the um you know the different tcks.

A

um um So I I I think, the the um the fallback on that even for 10, if we, if, if we couldn't pass with glass fish, all the technologies, I think the fallback would have been to remove some technologies so that other implementations that include the provided a subset of those technologies could you could have been the first implementation. So it's so it's really! I guess it's. You know, um you know yeah, so they're.

B

A

um Sorry we failed.

B

No go ahead. I'm sorry! I stopped you there. Oh okay,.

A

Yeah yeah, um so so the um so you know so so so the.

B

A

The um first implementation yeah needs needs to pass. All tcks is the current rule um and but yeah but anyways the fault. The fallback was that we could have. You know we would have to remove things that the um um the set of all whatever first implementations that would have been used. Instead, the set of all technologies would have been. That would have to be the defining set of what's in the release. We'd have to remove everything else. That's not in those compatible implementations as a superset.

A

You know, as they combine as a combined set.

A

So I mean that's that was sort of a fallback, so it's kind of I mean that's kind of a pro. You know a question going forward as well. You know if we can figure.

B

A

How to do that then.

B

Yes, and- and my point is- I don't see the conversations coming into the threads- um I mean the email forums, it is continues to be private, and you know everyone here hopefully understand how I feel about privacy and open source with with relations to technology creation, and it just seems to me that, though we know is there, we have seen the elephant we're not discussing it.

B

We think the people and those private calls are the only ones that matter and care for this, because we continue to keep it private, but there is no reason for us not to discuss this via the mailing list, but it would reach you know if we were to use. You know there are melonies. There are like close to a hundred other ones like 50, but we could have you know it's like it doesn't really even need to be only the melonists.

B

We could be more social about this, it's just but scott. I don't want to multiply this. I'm just saying I'm glad that everyone here is aware, and thanks for writing in the minutes. You know what the status. How do we do this? You know it's not for this call to discuss, but definitely can we have it like as a standard every time we have a call at least five minutes dedicated to you see a status.

B

The status my recommendation is the status is: where are we on this, so that is recorded, and at least we are aware of there's no progress and why is that and etc, etc, etc? Right.

A

Yeah, I I don't remember um it may have been you know it may have been in private discussions. um um In my opinion, the only reason why it didn't come up in detail in the recent platform calls is just because we haven't had time to you know to to kind of okay. If we did that, how would we do it? You know, but we didn't, and in the last few weeks, though we did it, it did come up. It's just there's just been there's just been very little time to discuss.

B

I know it's cats, let me cut you there. The excuse that we put ourselves through is always we don't have time for this, we're always strapped for time. We need to stop apologizing for not having the time and prioritize what we need to do with our excuses. These excuses have been going for so many years since we moved like java ee to the eclipse foundation, and I swear to you that I see this continue to be a problem for the next.

B

Maybe three years we will be arriving 2017 and still have not solved this, because we will always have less time. It is about prioritization, and if we talk about prioritization, we don't have it as a priority. So we need to be honest with ourselves and say: hey. We are not prioritizing it because we have something that works.

B

Let's move forward, and I think that would be more honest. We can always smile and say we don't have the time, but that is not the truth, at least in my view,.

A

Yep and then just uh yeah not to drill too far into it, but um yeah we just we. We we have been running our time in the calls so but yeah, okay.

A

Let's see all right um so so yeah, so we covered, we covered that as much as we can. I think uh so um anything any any updates on the faces. Tck. I think I think the last I read were you know there. There were some. You know, tests that aren't ready that just are going to get you know developed in the future in a future release is that is that right allen is that uh yeah.

C

uh Yeah so for facebook like, uh but just to give a background, we have. We had migrated all the test sources from the jakarta tck to this new repository of api, so uh it actually uh it's in a new folder, along with the new tests that are added in the uh faces before so uh so. Yes, uh I can see that from the email that the ballot for uh faces that they have started.

C

Before so they have, they have created this new tck that contains new tests and the old pck, so the bundle is generated and available in the tck location for jakarta, idck itself.

C

Sources from the karthik with a lot of help from community, uh in fact, one specific person called world ordinary uh usa. We were able to resolve like all the failures except uh 25 of them, so for those failing tests, we have uh added them to the exclude list for now, and those 25 will be addressed in the next release.

C

Also to note that we have around 5400 tests in the old tck, along with that, the new tests are about 240, I believe so uh yeah we have just excluded 25 tests out of those which will be addressed in the next release.

C

Yeah most of the work was done by the faces, spec team and someone who had volunteered to follow the pages.

A

Great great, terrific, all right, okay, so so that so so that that means we're um we're in good shape there uh and uh and and did, and that I I think I saw you ask, but you know the so the face the faces ci is going to you know is, is building and running.

A

C

uh Jakarta it is building the tck. uh I think the execution of the test is done by the vci.

C

Okay, uh I don't know if it is a temporary thing, but for now for the ballot they have used, the bundle that is built from the cathode is okay, so, ideally they would move the building also into the faces thinking instance, probably soon.

A

Okay, um and and do you know, which are they just they're running.

C

uh Yes, but uh yeah. Ideally, uh uh the implementation is moira, but uh it's like a glass fish is used as a vehicle.

A

A

Nice, um so any any other, uh what's, oh any any other, um any other um standalone tcks that are, you know the new ones that aren't ready yet are aren't done yet. I think.

C

No, uh uh the other tck that was uh recently separated was security tck. uh So that was already. I think the security uh ballot have already started.

A

Faces bail at third this morning as well by the way, oh sorry, facebook.

B

A

It did start this morning, okay,.

C

A

C

A

In the state that allows for them to go to ballot nice.

C

Yeah this was the last one to start.

B

I think it just went out right this morning.

A

Yes, that's what I said. Oh that's, plastic.

B

Yes, I'm sorry, I was like muggling here it was like yeah. That's amazing.

A

A

I I guess I should share that um um the um plot you know the platform: tck user userdoc.

A

It's still in need of updates for um the new, the new tcks and um other than json b and jsonp they've been updated to mention json, b and jsonp, but um we need to add the other. um You know tcks are needed to be run by the different implementations.

A

um I I made a first pass this morning at um just you know, mostly update updating the version numbers and a few other updates. Just in a draft. You know pr just to kind of get that going and let me just I'll link in the uh add the link to that.

A

Started updating platform oops tck.

A

Yeah oops uh yeah, you know just the user guide, it didn't didn't, um do the install guide, but you know that's, you know you know. That'll that'll be next, but also you know it will add. We'll add um you know the new tcks.

A

Need to be added still um so typically, we wait till the very end to update the dock, so just trying to get that out of the way.

A

All right, okay, I.

C

A

The floor is open if it, if to.

B

A

A

Oh, let's see there is.

A

The next column for discuss, refactoring and simplify and get this.

A

Okay, so I guess that's that was from last okay, so so um I I guess from from last from last call yep. I see.

C

A

Yeah last call we talked about um I'm talking about refactoring and simplifying um one. So you know one one idea um on refactoring that I can um yeah we discussed in the mailing list too, but one one one thought is um add more. um um I should say nominate, nominate, nominate, more committers.

A

To work on new, you know test.

A

New ways of running running tck tests and um so the yeah, so the idea that the idea that they're there and what I was thinking of with nominating more committers is to um try. You know, try and get um just to increase our base of um committers on the platform tck um bring in more.

A

You know, more people that have contributed to um the um the other new new tcks, either in the ones in progress or um the ones that have been written and try and get them to help out on the uh the platform tck rewriting and um what um it's in in the um I'm just going to reorder.

A

I'm just going to move up this last item on that oops did. I just do.

A

Cut too much so try that again, okay, um so the um one you know one one one thought um about um um the um you know the you know the main brain yeah, the main branch or the master branch. um uh Whatever it was, we had decided the new name. Oh the yeah. I think it's gonna, be, I think, we're going to.

A

I think I think, we're in favor of version version based branch names but um anyways whatever the branch is um um that the the suggestion is is to start the the move, the refactoring into our main, develop development branch, so that we kind of keep the ant tck um testing um as long as it's needed um side by side with the new you know, maven and our killian or whatever, whatever you know, framework um side by side, so um so that um if we run out of time for the next release, we don't you know end up with nothing to show from the refactoring.

A

So we can kind of get make incremental progress and be able to ship that with the next release without having having to block the next ee release on the completion of the refactoring, and I would love feedback on on that idea.

A

So if so, if we do that, we'll we'll we'll um drop the refactoring branch and just have master main or eeo 10.1 or 11, whatever the you know, whatever the you know, the new branch name is and do a refactoring there. You know there directly so good. um So I guess it's not too crazy for an idea.

A

um If no one, no one is saying that's crazy.

B

No, I on the concrete, I think it's amazing. This is why I, I posted it back. It's easy to forget that we met a month ago.

A

I yeah. I really appreciate that because I did forget that until sorry, yeah.

B

No, it's already great. uh We don't approve the minutes from past calls, and you know it's just like. I remember that you wanted to talk about this and then there were two minutes left and I said, let's just end it, and then I felt just my responsibility since I caught the last one.

A

Thank you so um yeah I mean the tck mailing list. Yeah. Let's get the discussion going. Yeah we've heard that you know we know about that and um does anyone have any other topics that we didn't cover yet that should be covered.

B

um I have I have a question here. um The jakarta 10 is going to be released and within the eclipse foundation there is always the you know, the blog and all the things that um that go into the page of the release, I'm not sure if, in the pck part of the fans that so much work has been done, any of you are or have been welcome to send, quotes, or things like this clearly, I'm not behind like the managing of that.

B

But it seems to me that you should have voices and especially in a formal announcement, and if you are not, that is a mailing list for marketing where you could send your quote and hopefully say hey. This is a quote.

B

I work I have an integral contributor to this project and sometimes it's not about highlighting or putting the light on ourselves, but I think it is important to be heard, especially when some of you are extremely technical and contributing your time to the to to the project attending the calls is awesome, but the work is done and they get right when I quote that is maybe less than 60 words might be wonderful, I'm not sure how alwyn and jared and brian feel I think scott is used to this kind of like notoriety.

B

But uh um you know, would you appreciate having that opportunity? And if so, uh you know um what is stopping you from requesting that.

B

Clearly, I'm asking you everyone here, including you as cat.

A

I would I would put myself into the group of people who you know I mean I'd like to do. You know: hands down yeah heads heads down development and you know you know. I participated in all these calls and everything, but um you know I definitely you know, would fall into the group of you know: people who wouldn't just you know you know you know vulnerably volunteer that, um but I think it's a great idea. Maybe we could share some of it on the mailing list.

A

I don't know, I don't know how to pull it together. um Maybe we could work, I don't know if we I don't know how to do it collaboratively. I guess would be you know you know. If we you know we could just do. We could kind of do it jointly.

A

You know if there's desire there and then you know people could contribute to that, and um you know we could add to it and and kind of you know all sign it together. It I don't know just think yeah.

B

A

B

God you nailed it um always um brian. How? What is your take of this? Like community um small, quotes, something that you know that represents you know the aggregates um work. The kit that could be used, you know, would be so valuable. We talk about refactoring and the need for new contributors.

B

How do you expect to get new contributors if you are so silent, not seen or appreciated anywhere? You know we're not talking about paycheck we're talking about bringing contributors from all over the world to say, hey. I want to have that mentor. I am studying in computers. I still graduated I'm looking for a project. I feel welcome here right.

A

Yeah yeah, I suppose, like one plug on the um on the new contributors. We can talk about the refactoring opportunity opportunity to you know to work on the refactoring, um but I think it's you know once it's refactored, then I think we will get more contributors, but.

B

It should be like, while at it I'll win you one, you want muted yourself. Do you have anything to say.

C

No, I I think I agree with scott about including the quotes like we can.

C

We can check with the mailing list and ask people to you know about their opinion. I mean uh about the major contribution that was done for jakarta. That's what about the uh quotes for list blogs right, major uh contributions that was done for the time.

C

So we were talking about the codes that need to be included in the release block for the from tc case right.

B

Not only the co, it's not the calls. It's just like your work and the tck, like you know, reviewing tests passing tests reviewing commits merchant commits um these kind of things like some of you were granted um commit access immediately through employment. Others of you through work, got to be committers, we're talking about your current work and how exemplar and simple is exemplary.

B

Is it how much fun and frustrations you have through the process of tend, which is the first release with features? Nothing is more exciting to a new, a developer than creating something new, especially features. This is our time to to share right. I think scott has a great idea of hey. Maybe something of um community quote that you know the tck ecosystem could ask the eclipse for foundation marketing team to to put into that release right. We, I I see the releases and the announcements, and my critique has always been relaxed.

B

Community voices and the level of um you know we don't hear from from from from a an api, I would call like the tck an api totally standalone right, like we call that api, just because it's lazy easy. We never hear from you know like to say this group did, you know we believe and then sign, or you can say you know the the the the platform or the tck um project right.

B

It seems to me that the tck ecosystem could lead the effort to become much more um inviting and visible on those formal communications, and even so, this project should have its own blog posted in the jakarta news, and you know you could we could use the tck github to write a blog and say this is what went behind.

B

You know the tck work, and it would be amazing because changing that blog can only take a pr and many different people, not committers can be a part and feel appreciated, because some the help are not committers and tommy trump created a what is it called a pro, an application that pulls in every committer every contributor that has done all any work from the moment.

B

9.1 was finished that pr will trigger all of it and it would be. Each contributor will be highlighted, but we're talking about what else can we do, and I think a blog together will also help right. This is just talking about. How do you communicate so that more committers more contributors can be added to the work that is currently done? We have a lot of work.

B

We continue to complain that we don't have the time, but it's because we need more help and the only help that is possible is when we speak up and say this is what we're doing join us, and this is how we do it right.

B

Simplicity, works and being humble, also works, but we need to speak up and be super proud of the work completed right and it's just so simple. It takes communication, constant communication.

A

Awesome, okay, um all right, so that sounds like yeah. That sounds like something something to you know that we'll do and anything else.

A

All right, um it looks like we have 20 minutes of our collective time left. So I wanted to you know thank everybody. This is, you know great to um see you all and talk with everyone here.

B

Well, um I tweeted, I always tweet, I love it. I love each one of you I like seeing I I love openness, so I always tweet. um So there are um the tweeters sharing the tag in the chat. um So please you know if you don't have twitter, it's fine, but definitely um you know highlight this. Is you know we invested the time to show up and discuss openly, so it's just super fun. Thanks to scott for bringing us together.

A

Awesome all right, everybody uh have a great rest of your day or night.

B

May yeah super healthy may everyone. Let's do this.

A

Just want to take a look at the chat quickly, see what oh okay! Thank you thanks. Yes, I'm gonna stop recording.