2024-09-05 Meeting notes
- Kyle McCormick
- Feanil Patel
All public Working Group meetings follow the Recording Policy for Open edX Meetings
Date
Aug 29, 2024
Participants
Previous TODOs
Discussion topics
Item | Presenter | Notes |
---|
Item | Presenter | Notes |
---|---|---|
Previous TODOs |
|
|
DEPR tickets for upcoming upgrades | @Kyle McCormick | is it appropriate at this point to make DEPR tickets for version of Node and Python that we’ll be moving on from soon?
We’ll make the DEPRs once there is action that operators can take. I.e. There is a new version for them to move to that is tested and supported. |
Roles and Permissions | @Feanil Patel |
|
py 3.13 instead of 3.12? |
| 3.13 release is Oct 2024
|
edx-platform specific maintenance | ||
studio DEPRs |
|
|
|
|
|
Action items
Recording and Transcript
Recording:
Maintenance Working Group Meeting – 2024/09/05 08:59 EDT – Transcript
Attendees
Feanil Patel, Feanil Patel's Presentation, Jeremy Ristau, Juan Carlos Iasenza, Kyle McCormick, Leonel Katsikaris, Maksim Sokolskiy, Michelle Philbrick, Robert Raposa, Sarina Canelake
Transcript
Feanil Patel: Little.
Robert Raposa: Good morning.
Michelle Philbrick: I say I have a quick question that this group could probably just answer I noticed that on the repo sheet, there's now a maintainer or at least an owner of the EDX platform repo. so, I've been currently just putting needs reviewer assigned and leaving it to the CC's and everyone to take edex platform things. But should I be tagging that group instead the one that's listed on the sheet from now on. Okay.
Feanil Patel: Ixplatform is big enough that I think you can keep doing what you're doing with it. At some point, we can sort of talk about. Different plan for it,…
Michelle Philbrick: Sounds good.
Feanil Patel: but not yet.
Michelle Philbrick: Yeah I didn't want to start doing that without checking first just in case.
Feanil Patel: Yeah, thank you.
Michelle Philbrick: All right, cool.
Jeremy Ristau: Whatever.
Feanil Patel: Morning.
Feanil Patel: Give people a couple more minutes and then we'll get going. meeting notes are
Feanil Patel: Jeremy Screen.
Feanil Patel: You guys read that, okay?
Feanil Patel: Awesome.
Feanil Patel: Welcome everybody to the Maintenance Working Group meeting.
Feanil Patel: We usually start off by God's been previous to do, so we'll just go through that real quick and then we'll talk about the agendas at hand today. I see. A couple of new faces. So I'm gonna here's the link to the meeting notes. If you have specific things, you guys want to talk about. Please feel free to add them to the Discussion Topics section.
Feanil Patel: All right.
Feanil Patel: so, previously I had this To do for writing a debit for the related repositories and code that I did about a week ago. So that is anything related to the commer repo which is commerce worker and a couple of front ends related to that. So I put a link to that here.
Feanil Patel: end up payment and front end up commerce as well as ecommerce scripts. so far, I think we've only received for one major piece feedback on that, which was what happens with How eCommerce is currently integrated with the Enterprise repositories? But since we're not removing code from existing code bases, the archive repos will continue to work until sort of a future date so we can defer that decision for a little bit.
00:05:00
Feanil Patel: So, I think that will likely get accepted. Tomorrow.
Feanil Patel: I need to link to.
Feanil Patel: elect the discussion post on here but it should be tomorrow.
Feanil Patel: Yeah, I added a note to the discussion for him but I'll add I think that there is actually already a note on here, which is
Feanil Patel: Yeah. The code removal from other systems, we'll be done in a future release and we'll have its own depot announcements.
Feanil Patel: But I thought I added a link to the official dip but,
Feanil Patel: Yeah.
Feanil Patel: I'm just gonna add that here.
Feanil Patel: September 9th. So technically Monday not tomorrow, but yeah.
Feanil Patel: No major issues as far as I know. So, if you guys have concerns, please chime in on that discussion post or on the Decker ticket itself. Either is fine.
Feanil Patel: The other. So that I added this roles and perms to the agenda. And then this stepper for Ubuntu 2004 to Ubuntu, latest is not written yet, but that I'm planning on writing today or tomorrow.
Feanil Patel: The tldr on that is in a previous meeting but just real quick is rather than going to a specific new version of Ubuntu. We want to move to the latest pointer that Github actions runners have so that we don't have to do this maintenance over and over again because for 95% of our repos. It does not matter which version of Ubuntu you're running on as long as it is sufficiently modern which is, I think anything above 1604 and you can get the version of Python that we actually care about. So since most people are running the tutor, I think it's fairly abstracted away from people, and if you're not running tutor, you can still run it on any modern Python Ubuntu version, that has the correct Python versions.
Feanil Patel: There's a set of system packages that are needed, which is why we're still tied to Ubuntu versus any Linux system In theory, you can run it on any Linux system, but we don't know what the names of packages are in the various different distributions. So people wanted to document those equivalent packages. Then they could run it on whatever system they wanted.
Feanil Patel: so, hence, the sort of shift because with the github actions for all of our tests, There's automation to give us the Python the right node version and everything else. So we don't really need to do a lot of the same setup that we needed to do previously specifically.
Feanil Patel: And either you are good here Okay, this Is still, I think underdevelopment.
Feanil Patel: Jeremy is the dapper tickets for the front ends that are part of the new course. Authoring out. Here are those created yet?
Jeremy Ristau: No, we haven't been looking at this really. So …
Feanil Patel: Okay.
Jeremy Ristau: We're probably just gonna write one for simplicity sick.
Feanil Patel: Yeah, that sounds good. Do you have a sense of when you guys are going to start looking at this? Just so, As we figure out planning for Sue Mcintee. I want to make sure we sort of put it in the right place or start talking about it.
00:10:00
Jeremy Ristau: No, not really, but I mean if it's a desire and a timing thing, someone else could write it ever to get to, if
Feanil Patel: yeah, it's not so much The desire is always there but I think it's more understanding timing rather than because I think that is not scheduled at all, is never going to happen. So if it's not scheduled at…
Jeremy Ristau: Yep.
Feanil Patel: I would love her to get scheduled, even if it's six months from now.
Jeremy Ristau: Right. Yeah.
Feanil Patel: So, if you don't have a schedule for it, I don't care when it is as long as it's actually scheduled. So it's gonna get done. but, And if other people free up but I think there's enough old frontends that need removal that dippers working. The Deborah group is working on that I don't know if we'll get to that. Soon. But it'd be great, if we soon, this will become ancient information instead of information that's in the heads of the engineers that worked on it. And that's what I'm mostly worried about.
Feanil Patel: Gentin isn't here but I believe he was on vacation and also had a bunch of other work stuff but I did check in with him so I think he'll be able to get to this at some point and then Brian and Adolfo did write there is a new dumper that I did want to point people to which is the replacement of footers via forks, so historically When you needed to update the header and footer in your sort of variation, your deployment of open EDX, you would fork those two repose. And then you would make your modifications and you would essentially install those portions of the header and footer, the NPM package lock aliasing.
Feanil Patel: that is not going to be supported anymore because of the way that those interfaces are changing. So if you are forking the footer or have a custom header footer and you aren't aware of this, you need to take a look at this deprecation and see how you can move towards the new frontend, plugin framework.
Feanil Patel: we've got Kyle. Debra tickets for upcoming upgrades.
Kyle McCormick: Yeah. What it says on the 10? whoever is doing those upgrades.
Feanil Patel: Yeah, I wrote one for EDX platform…
Kyle McCormick: Can we get those different tickets written?
Feanil Patel: which is running both. I think for some of these other ones we had talked about And maybe this is the thing we can sort of discuss this. but I think previously, we talked about not making those tickets until the work is done, so that people can take action.
Feanil Patel: But I wonder if we can make the tickets but not move them into announce until you can actually move on to other versions because for example, the Depper of Python 311 we can do but there are repos that are still not testing on 312. So people can't move to 312 necessarily because we don't know if that'll work via CI. So there's a timing question in my head of do we make the dappers after The transition is available or before the transition is available, but we don't announce until the transition is available or something else.
Robert Raposa: And I think, at one point we talked about the last one being possible that, you could announce and say that the replacement is not yet ready and have a date for that. And then a six months out from that date for the final.
Feanil Patel: The actual.
Robert Raposa: As a possibility. but we could also just create the decker and…
Feanil Patel: Yeah.
Robert Raposa: then once we have it discuss whether or not to announce it, Yeah.
Feanil Patel: All right, when is the time to announce it?
Feanil Patel: Other thoughts. What do you think? I'll
Feanil Patel: all right.
Feanil Patel: yeah, I mean It looks I don't know how many different operators we have on here but yeah I'd love other people's thoughts on this when would it be useful to get and notification that we're dropping support for Node 18 or Ubuntu Would it be when you can actually take action or would be as soon as possible even When we're just about starting the work and, perhaps we need more than one notification of Hey, we're starting this work. This will become deprecated. And then another one saying, the transition is now available.
00:15:00
Jeremy Ristau: I am listening and don't have a strong opinion. But I thought I would not ignore the question.
Feanil Patel: Okay.
Feanil Patel: Thank let's try this. I think we had ly not creating the ticket until we had the transition available, but maybe we should create the ticket and we won't put it in announce, but we'll have it. maybe we'll announce it and then we'll just announce it again, when there's a transition available, what those are the kind of two options in my head and I kind of leaned towards Announced twice. But I also want to be mindful of noise.
Feanil Patel: When people prefer. that I say we're gonna get rid of Python 311 soon and then, we're gonna actually get rid of it, Six months from today. Or are those two messages? Is that first message? Not really useful because you can't really do anything about it. Even if you know that it's happening as a maintainer, you could, but
Feanil Patel: Yeah.
Feanil Patel: I mean, I think this is a thing, we've talked about We're getting to the point where it's probably time to start thinking about what we want to do for teak but on the Platform Roadmap, We have specific pieces of maintenance that are going to land. For each open source release. So the Ubuntu upgrade in the node, upgrade are kind of the two critical ones. that need to land. And then There's the just upgrade in the enzyme upgrade which are less critical, but we're planting on landing those.
Feanil Patel: There's that sort of the maintenance to be able to move to the new versions and then there's the mirror of that, which is the Deprecation of the old versions. I kind of given that we have other ways of informing people that the transition is happening. I'm now kind of leaning towards leaving it so that we don't make the dipper until you can do something about it.
Feanil Patel: Right, because it's on the platform roadmap as these maintenance tasks are occurring. And then when the dipper is posted, as an operator you can act on things, it's kind of a nice. Feature.
Feanil Patel: No, No. It's the open EDX roadmap.
Feanil Patel: Currently. No because we haven't decided what will go into T? But if you click on the sumac, Tab, You will see what we've sort of already promised for sumac.
Feanil Patel: So we do.
Feanil Patel: Yeah, exactly.
Feanil Patel: so, if we go under sumac, there's a swim lane for maintenance And you'll see these are the upgrades that are in being developed and in the backlog. This is also now being developed people have started making those changes which is awesome. And so, yeah, I think we to sort of now is the time to think about what we want for teak. In terms of maintenance and what that looks like.
00:20:00
Feanil Patel: But first sumac we had sort of decided to focus on these. I think for teak that might be a good time to start thinking about dropping, there's sort of two big things on my mind. One is dropping. 3/8 support is going to drop for EDX platform and I think a bunch of things in December. We have a debit for that already, but we could. Try to move everything to 312 for teak or make three twelve available for teak and then we could definitely three 11 everywhere.
Feanil Patel: That I think would be a really good thing to target for the teak timeline. And then, the other one is Django. Five, Zero and five, one are already out and so we can start fixing. Those backwards combat managing those backwards and compatible changes. Now, before Django 5-2 comes out, and we're in a rush to fix everything.
Kyle McCormick: Finial, just to clarify. You said we would drop. 311 Support and simultaneously AD retail support antique.
Feanil Patel: No. No, I think we need to add three 12 support 14. and then whenever through 12 support actually lands,…
Kyle McCormick: Yep.
Feanil Patel: whatever month that is, We make the deprecation of the 311 support at that point.
Kyle McCormick: I mean, from the perspective of an operator running off of releases does Have any simultaneously teachers support, but
Feanil Patel: He should support both. yeah, yeah, because if we add three 12 support sometime in the six months leading up to take, Then we based on our own ruleset. Can't remove three 11 support until six months after whenever we land that So it will definitely be after the release is cut.
Kyle McCormick: yeah.
Kyle McCormick: Maybe this is a topic for another time, but I'm starting to think that a six month. Simultaneous support window. Is not.
Kyle McCormick: Prudent for certain things like Python versions.
Feanil Patel: Yeah.
Kyle McCormick: I don't want to drill this topic so we can maybe discuss that in a future meeting.
Feanil Patel: Yeah, I can I'll add it here as a next time thing.
Feanil Patel: I don't know. Simultaneous
Feanil Patel: yeah, let's talk about that next time. Yeah, I think that's a bigger topic because we did arrive at the six month window when talking about the Python 38 to 11 upgrade for operators running on master as a long enough that they have plenty of time but short enough that we can do something about it but yeah for testing that's very expensive to be Doubles of all of those tests. So maybe we need a strategy for how we validate support?
Kyle McCormick: Yeah and I want to give time for real discussion on that,…
Feanil Patel: Yeah. Yeah.
Kyle McCormick: so let's definitely talk next time. Good to you folks a good chance to think about it and talk about it.
Feanil Patel: All right.
Feanil Patel: Okay, so here we said
Feanil Patel: I can't say.
Feanil Patel: Sorry.
Feanil Patel: And then the final thing, we've got five more minutes which is the roles and permissions work, which we had a discussion about a couple weeks ago.
Feanil Patel: This is still in super early stages, and
Feanil Patel: Why it's completely about that? Once there is a proposal, it will go through the standard product review process. So if people are curious about potential changes coming down the line for roles and permissions work in EDX platform or in the open.x ecosystem, generally Be sure to monitor the product review process, and I've linked to how that process works there.
00:25:00
Feanil Patel: All That's a thought I'm concerned about the libraries we run, having added support for it fast enough for us to move to it.
Kyle McCormick: That makes sense.
Feanil Patel: But yeah, there's each new Python version now has such so many goodies in terms of performance improvements. That. It is tempting.
Kyle McCormick: Food for thought, we have four minutes. And Jeremy also has a question about the last one. So, maybe we also put this till next time.
Feanil Patel: Is this similar to the work? I know that the work that to you started is Being taken as input, Jeremy, but I don't know. If it is that same sort of underlying architecture or I think this is sort of more being handled from a user experience and point of how do people manage these things and what should it look like from the end user's perspective?
Feanil Patel: And I would not be surprised that the underlying architectural decisions that were made for the two. You work would come back because those are good decisions in my mind.
Feanil Patel: Yeah, I think we should look at what 313 has dropped support for Kyle. And if it's not like a lot of major stuff then, perhaps we can just move to 313. Which would be nice.
Kyle McCormick: I will try to check out the change like a bit and get a sense for weather packages are already releasing. 313 versions yet.
Feanil Patel: Yeah.
Feanil Patel: Okay.
Feanil Patel: Okay, I think that concludes the Maintenance Working Group meeting for today. Unless anybody has any other topics to discuss.
Feanil Patel: I'm also going to add for Next time is going to be a packed meeting because we should also Take maintenance.
Feanil Patel: Those.
Feanil Patel: As. Yeah. think about things that you think makes sense.
Feanil Patel: We can take a look at the
Feanil Patel: What is it called? There are the release versions. Excel spreadsheet or whatever it is.
Feanil Patel: I'll link it in the notes. But the name is escaping at the moment. agendas, maybe
Kyle McCormick: Support Windows.
Feanil Patel: Support Windows.
Kyle McCormick: There you go.
Feanil Patel: Yeah, support. And that's Those.
Kyle McCormick: Hey, before we break, I'll leave a shout out to Ahmad who, as of today, just passed this core contributor and maintenance, and maintainer nominations. So once all the paperwork goes through, you will be maintaining four repositories. So that's for if you are unmaintained, repos and one nuclear contributor,
Feanil Patel: Yeah, awesome. Yeah. My mom's been doing amazing work across the whole system, so it's really exciting to see him take on some specific repositories as well.
Feanil Patel: All Cool.
Feanil Patel: Yeah, and it's platform. So if you guys want to stick around for EDX platform related discussion, stick around otherwise feel free to drop off. We're gonna discuss maintenance specifics to EDX platform, which I think we have a next time one.
Feanil Patel: Do we have discussion, okay?
Feanil Patel: that's,
00:30:00
Feanil Patel: Yeah, I start us off Kyle. I think you've got the only topic.
Kyle McCormick: It's alright. Yeah, yeah. when we were talking about for Jeremy's action item to write the Debra tickets, but I was on a train. So I didn't feel like I could really talk so to be completely honest, I could write the dapper in about 15 minutes. but if I wrote Deckers for everybody, I would spend a lot of time doing that, so, Jeremy, I'm just wondering you guys put in a lot of amazing work to get those front ends rewritten, and this is just a bit of paperwork at the end to kick off the removal of the old stuff. So Is there anything in the process that could be easier? Or is it just Overly bureaucratic, you're not seeming very valuable to rate that dipper. What do you feel like you're going to be stuck with work, if you write it?
Jeremy Ristau: Yeah, I mean it's not writing the departs. What you're signing up for by writing the The actual removal of code, we have to roll these pages out to edge environment, which is not rolled out yet. there's a lot more than, just the 15 minutes and writing it and also signing up to do that. Within the next six months is something that requires planning and discussion across different people. So I mean, Yeah, if you want to write the dapper and sign up for the work that's associated with it, I'm 100%. For that.
Jeremy Ristau: But that's why it hasn't been written. Yet we're running,…
Feanil Patel: Yeah.
Jeremy Ristau: we're running on one and a half people dedicated to authoring. So we do not have any plans to do any work there. And so the request to Deb are all those pages is beyond what we had planned to do. So
Feanil Patel: If other people take on the work of removing, this old code that is going to impact. It sounds like you guys. in your edge environment,…
Jeremy Ristau: yeah.
Feanil Patel: is that? Is it okay if somebody else goes in and deletes these old pages? Yeah, a dipper.
Jeremy Ristau: I would figure out how to work it into the next window of time somewhere. But being proactive about it and…
Feanil Patel: Okay.
Jeremy Ristau: having to compare it to other roadmap items. there is no like To you value. And that's the hard argument for prioritizing it over other things. So if we're going to incur, Production issues. Because of work that is happening. Externally, we have to prepare for that. But, choosing to do it at a certain time,…
Feanil Patel: Yeah. Yeah. Okay.
Jeremy Ristau: is it?
Jeremy Ristau: But you could force our hand. and then we can get it done like that.
Feanil Patel: Right, I don't want to do that…
Feanil Patel: but
Jeremy Ristau: yeah, the way of getting it done that is a way of …
Jeremy Ristau: making sure it gets into a roadmap somewhere but it is, more stick than carrot. Yeah.
Feanil Patel: Yeah. Yeah.
Jeremy Ristau: Really exactly. Yeah.
Feanil Patel: Right. Yeah.
Feanil Patel: Yeah, I mean I think I stand by my statement which is that this to me is completion of the work that you guys did and I would love her to get scheduled. I understand that it won't necessarily be tomorrow but not being scheduled at all. Is not useful if that makes sense.
Jeremy Ristau: That's right. I understand that and it still hasn't been scheduled yet because of the fact that it just doesn't make it up the list of when we have product like discussions of priority,…
Feanil Patel: Cool. Yeah.
Jeremy Ristau: there's no,…
Feanil Patel: Okay.
Jeremy Ristau: there's no argument for this work.
Feanil Patel: I think In the Future for any replacement sort of work that to you embark on in the core product. This sort of removal should Accounted in the initial work. I think that would be the expectation moving forward.
Jeremy Ristau: Yeah, so I think we planned for that at the end of sort of the replacement of all studio pages. And then we didn't like…
Feanil Patel: Yeah.
Jeremy Ristau: then we couldn't get to all studio pages and then that, some pages got picked by other groups and it kind of became more of a larger thing or…
Feanil Patel: Right. Right.
Jeremy Ristau: are more spiders out things. So yeah, I mean that's certainly a good. Reflection point, but I don't think we're looking at it page by page at the time. Yeah.
Feanil Patel: So Yeah, I think it's a good reflection for next time, is that when we're replacing a page, the whole life cycle of that replacement should be considered per page. Which is.
00:35:00
Jeremy Ristau: Yeah.
Feanil Patel: Especially within the core product. As more things move to plugins and as we get a route plugins for front ends, hopefully this will become less of an issue and we can sort of work on the car product. Independently of specific replacements people have
Jeremy Ristau: there's also the fact that Feedback exists about the new products. That I don't actually know when.
Feanil Patel: Yeah.
Jeremy Ristau: When it's to you Acceptable, And so we've just been like,…
Feanil Patel: Yeah.
Jeremy Ristau: delivering as we can on the things that we know about but getting through, that backlog is also not something easily,…
Feanil Patel: I mean. No, that's totally fair.
Jeremy Ristau: prioritize all the time. So
Feanil Patel: And I think it's really useful that that's one of the reasons I think the Decker is actually very useful, is that on the dapper, we're able to collect that feedback as a community, and then it would not necessarily be Up to you to sort of build every feature the community wants, but it would be useful for a community perspective to enumerate. The differences and Figure out if we want the new thing or the old thing is the default.
Feanil Patel: and this is, I think A growing pain of shifting the way that we decided, what goes into the core product as interfaces. And this just happened to be kind of some of the most recent work that followed the old paradigm.
Jeremy Ristau: Yeah, I would agree.
Feanil Patel: So I think we'll have to do better moving forward but I would really love to get this. Planned in prioritized. so that we can actually get the old code out and stop maintaining it which we're all paying the tax for
Feanil Patel: all right.
Jeremy Ristau: So I guess from that perspective, feel free to write. if it takes 15 minutes and what you're looking for is feedback, not To deprecate the pages. Because it's the signing for a time by which we will do some work that we have difficulty prioritizing. That is what's keeping us from writing the Decker. But I'm hearing something different which is let's write the dapper and see if people are okay with getting rid of these pages or dead feedback. Which is different.
Kyle McCormick: It's going to be both.
Jeremy Ristau: Right. Exactly.
Feanil Patel: Yeah.
Kyle McCormick: I think if I write the ticket, I'm gonna start pushing pretty hard on getting rid of the stuff that it has been replaced.
Jeremy Ristau: Then maybe we do need a separate debit per page. Because it's about whether or not each page meets the needs of all I think it will become too big of a decker if there's just one and we're trying to look for specific issues with every page that we wanted deprecate. That might be unwieldy.
Feanil Patel: Maybe, okay, to create the one. And then, if there are issues, spawn off some tickets to collect those and Update the core one with what's left.
Jeremy Ristau: But we already have the one for the editors, right? And that already feedback.
Feanil Patel: Right. Right.
Kyle McCormick: Yep.
Jeremy Ristau: Yeah, so we couldn't deliver on some of the new pages without the editors I think so. Is there a chaining mechanism in the deprecation process or
Feanil Patel: Yeah. I mean you can have task lists and dependencies
Kyle McCormick: Yeah. I think the chaining mechanism is you Write descriptions that describe the training.
Jeremy Ristau: Just write it. Okay.
Feanil Patel: Yeah. Yeah.
Kyle McCormick: Yeah, so there's a unit deprecation and it's like we have to do the editored deprecations first and has all been listed. So,
Jeremy Ristau: Yeah.
Feanil Patel: But there's a lot of other smaller pages that I think don't have any issues and it'd be good to just drop having two versions of
Jeremy Ristau: I mean,…
Kyle McCormick: Yeah.
Jeremy Ristau: we're suffering from having to bounce in between subdomains and having performance issues and the call out of not being able to land to not be able to have the editor over the block that it's editing.
Feanil Patel: Yeah.
Jeremy Ristau: That's because we don't subdomain to sub domain and you get a full paper refresh.
Feanil Patel: Right.
Jeremy Ristau: And so we can't land in a particular place. so some of those are only going to appear…
Feanil Patel: Right.
Jeremy Ristau: if we choose to deprecate and roll out specific pages,
Jeremy Ristau: I guess we put the dapper out and see if people are okay with that or do we expect people to infer that those kinds of things will happen?
00:40:00
Kyle McCormick: think what we can do is we can put peppers out and explain what if we're aware of stuff like that ahead of time, to put it in the Dipper ticket. The one you just called out. I hadn't realized that the full page.
Kyle McCormick: Takeover was a matter of. the unit editor being legacy and the editor itself being In the MFA. So call outs like that are like and…
Jeremy Ristau: Yeah, right.
Kyle McCormick: cause like that and the feedback we got from MIT or Why I like having the dipper tickets early?
Kyle McCormick: and we can put dates up and if people say, hey We're having problem with this particular mfe page. We want to keep the old one, but there's another one that no one has any problems with and it unlocks some performance improvement if we get rid of the legacy version, Then there we can use that feedback to prioritize that second stepper.
Feanil Patel: Right.
Feanil Patel: Yeah.
Jeremy Ristau: yeah, I think that's where I'm not looking to assert an opinion. I think other people would like me to assert an opinion. And I'm saying that opinion comes with baggage and I can't afford that baggage right now.
Jeremy Ristau: and that's why it doesn't exist.
Kyle McCormick: Better understand.
Jeremy Ristau: And I also understand the maintenance cost as well it's not something I'm enjoying not proposing, but
Feanil Patel: Right? The reality of the situation is…
Jeremy Ristau: Yeah. Yeah.
Feanil Patel: what it is. No, I think it's everybody's doing the best they can. We'll get to it when we get to it. I think that, as long as you're attempting to try and schedule it and you're making good decisions, then We'll either wait for that or some other resource will free up and we'll get to it that way.
Jeremy Ristau: Okay.
Feanil Patel: But we don't want to lose it because I think that's where we start actually paying a bigger tax on maintenance. Is when we lose that, we need to do it for a long enough.
Jeremy Ristau: Yeah. Yeah, I'm also not totally sure that the people who built the replacement pages, Didn't do a lot with the old front end in order to do that so it's not a context that they're swimming in right now and losing and…
Feanil Patel: where, Okay.
Jeremy Ristau: and the maintenance of the course authoring repo is still ongoing. So we're benefiting from that knowledge.
Feanil Patel: Got it, okay. But that's a good thing to point.
Jeremy Ristau: So yeah. I mean
Feanil Patel: Yeah, that makes sense. because the people who created the old pages are long gone. So nobody…
Jeremy Ristau: Yeah.
Feanil Patel: who worked on the new Things about the old one that we would benefit from during the removal process.
Jeremy Ristau: Yeah, it wasn't like a refactor. it's much more hardcore.
Feanil Patel: Right. Yeah.
Jeremy Ristau: Brain killed replacement. Yeah.
Feanil Patel: Yeah. That's That's a good point.
Robert Raposa: And wondering if even just creating a wiki page or something. several good points have come up, that will probably vanish what's real dispersed that he come back and discuss this and then the video. So and it sounds like typically a Debra might be the place to do that. And we've got many times about The Denver is not happening right now. But maybe just a place to temporarily collect good points.
00:45:00
Feanil Patel: Right.
Robert Raposa: So that. Yeah, no. a place to collect these notes so that when someone decides that the dipper is the right next action, and they're good.
Robert Raposa: Have that as data and…
Feanil Patel: All right.
Robert Raposa: and maybe it's just taking. Additional Notes In these pages, it could be where we discussed it ever. But, I think there was
Robert Raposa: whatever point they just made. About knowledge, an assumptions and who has knowledge and the way earlier about the candidates.
Feanil Patel: Yeah, this might be a good place to put it is on Our pilot issue around this.
Robert Raposa: They?
Feanil Patel: What You think this is the right place? Because we've been talking about sort of what this dipper contract means and this is all actually still new. So we don't even have it sort of integrated into the debris documentation yet.
Robert Raposa: That this is a general issue. This is not studio specific.
Feanil Patel: yeah, I mean I think it's like some of the things we talked about today are this notion of Depers perform a couple of different tasks and how do we manage that or do we need to change the process to do it? That's like a learning that is coming out of us trying to use diapers for breaking changes.
Robert Raposa: yeah, so I think that the place that you just mentioned might be a great place for those types of learnings and any bullets that are Specifically around the studio. Should end up somewhere and they could be right here in this box. I don't know, but it is normally say good point,…
Feanil Patel: Yeah.
Robert Raposa: those are things that really did come back to you. And Something actually takes a story.
Feanil Patel: I wonder this has become a part of the Deborah process because the product process didn't exist at that time. And if the product process had existed, then be like, What are the capabilities of the new thing and what are the capabilities of the thing? We're replacing all of those decisions would have been made essentially upfront rather than at removal time. Which is More correct, anyway.
Robert Raposa: so, One question has that happened? Because obviously it could be late. But Couldn't that be part of a product review process now, instead of the DECKER process, and then it just goes into whatever process when it's all done and ready.
Feanil Patel: Right? That's what I'm saying is I think that for new things that go through the product review process. I think that would happen. and Serena maybe this is a place for I would love here. I think this is true but can you check me on this? Which is when we are thinking about replacing things via the product process, we thinking about the thing we're replacing and sort of like, Delta's between the new thing and the old thing.
00:50:00
Sarina Canelake: Yeah. I don't even know if it's come up at all in the current product review process. I mean what I'm seeing is people proposing, net new features and net new enhancements like graded discussions for example, it's just like a net new feature on top of something we already have.
Feanil Patel: That.
Sarina Canelake: I think if you're proposing a redo of the instructor dashboard, let's just throw that out there. I would think people would be really interested to make sure there's parity and functionality along with good UX and good placement of the capabilities etc. Right because we might imagine for that particular one that we might be talking about moving functionality into studio instead of in the LMS. And so that would sort of be a big discussion about where things live. I don't think there's any discussion about the decker of the legacy. I'm not sure…
Feanil Patel: Right. …
Sarina Canelake: if that's sort of the question you're getting at.
Feanil Patel: I guess, two different questions one. I think also, Jeremy has an example of a thing that we can think about.
Sarina Canelake: Mmm.
Feanil Patel: But the question that I had was sort of like, let's say, we wanted to replace the instructor dashboard. I would imagine that in this new world, that should go through the product review, process. If We Want, It wouldn't become a Decker…
Sarina Canelake: Yes.
Feanil Patel: until it went through the product review process.
Sarina Canelake: And I think we're so far away from thinking about the Debra process to Jeremy's example. removing the breadcrumbs was approved, as part of the product review, process I think that's a UX element. is that go through the Deborah process as well, when we decide to revamp a page and then we remove a UX element. I don't actually know the answer to that.
Feanil Patel: Right.
Feanil Patel: Yeah. Same I would think that for it's hard to know because I think a lot of the deaf or process. It's not completely operator oriented but it is often operator oriented. But then there are things where there is product impact. when we're getting rid of code, we are getting rid of functionality. And so it's often from that direction of we want to get rid of this code. Here's the impact on functionality and not from the other direction of, We want to get rid of this functionality. Here's the impact on code. So I think it might have been a miss that deprecate that didn't go through a deprecation.
Sarina Canelake: Yeah,…
Feanil Patel: And that's just a muscle.
Sarina Canelake: it's an interesting question,…
Feanil Patel: We have an exercise.
Sarina Canelake: because if we have, A group of product managers steering the direction of the platform who've all agreed on this. How do we handle disagreement in the diaper?
Feanil Patel: Right.
Feanil Patel: And that's a great question. Or is that agreement from product? Maybe a step in that should result in the deprecation being announced so that they could is a good way to collect community feedback from other. Operators about things But I know that product already has their own sort of feedback mechanisms that are different from that.
Sarina Canelake: Yeah, and
Feanil Patel: So that's perhaps,…
Sarina Canelake: Yeah.
Feanil Patel: a Disconnect that we should think about more.
Sarina Canelake: Yeah, I definitely think of the deper processes operator facing.
Feanil Patel: Except for when it isn't.
Sarina Canelake: yeah, I know that's fair but I couldn't give you if you had asked me from the start, I would be like No this is never anything but operator facing so
Feanil Patel: Right. Right. Yeah. And then we run into things like the editors and studio pages and now we're sort of having products
Sarina Canelake: Right. Right.
Kyle McCormick: or e-commerce and xq and notifier, and
Sarina Canelake: Yeah, no,…
Feanil Patel: Right. Exactly. Yeah.
Sarina Canelake: you totally right. But those also Involved an amount of technical things like the technical thing is happening. We're removing a page because it's tech debt and…
Feanil Patel: Right. As.
Sarina Canelake: it has user input, removing a UI element feels a little bit different,
Feanil Patel: No, I think it's A
Sarina Canelake: I mean, I can watch for that and we can try going through the depot process. I think it might be something that's hard for a group of products managers to remember and our process is already eight steps each step with four or five sub bullets.
Feanil Patel: Yeah, no I don't think it'll be easy and I don't think it'll be obvious. I think this is a nuanced thing where you'll have to, sort of, think about it a little bit, but I think there will be things that are product decisions that need a dipper. I don't think all product decisions that remove stuff will need a decker.
00:55:00
Sarina Canelake: There.
Kyle McCormick: So, we should figure this out but this meeting is not the place. Where you guys want to follow up?
Feanil Patel: Right.
Feanil Patel: Is this a conversation where some of us deper folk drop in on a future product working group meeting and have this discussion?
Kyle McCormick: Sounds like a good idea.
Sarina Canelake: Yeah, probably. Jenna's out for the one next week.
Feanil Patel: Okay.
Sarina Canelake: So maybe the 24th.
Feanil Patel: Okay.
Kyle McCormick: could also just do this post sumac, I don't know if we need When our brains are all a little less full.
Feanil Patel: Not better.
Feanil Patel: Yeah, what's your thought on that Sarina? Does that make more sense? Post Cmac.
Sarina Canelake: Post-sumac cut. It becomes a really insane time for the core product working group. And we're doing all the testing.
Feanil Patel: Because they're writing all the notes and stuff.
Feanil Patel: Yeah.
Sarina Canelake: So it's either post sumac release or the 24th is what I would say.
Feanil Patel: I think we should do it on the 24th.
Kyle McCormick: Okay.
Feanil Patel: Is there an agenda for those meetings that we should put this on Sarina?
Sarina Canelake: I can create an agenda.
Feanil Patel: Yeah.
Sarina Canelake: Where's my?
Robert Raposa: I don't know if it's a part of that same conversation and I'm not sure just based on your response, the earlier, whether I was there on what others mentioned, but when it comes to this studio work in particular, you imagine if we had this different process, During project product review. And we had product review the studio work. We might,…
Feanil Patel: Right.
Robert Raposa: we might know more now about what we want to like Deborah and whether it's ready. and all I would say is it possible to actually run it? If it hasn't to run it through that process, even though it's late and things already exist and make it part of a product review process that doesn't ever so.
Feanil Patel: Gotcha.
Robert Raposa: That's not
Feanil Patel: Yeah, no, I totally misunderstood that. Thank you for repeating. I don't know if it would be useful to put it through the studio process. I think at this point, The pattern we followed with the editor it where we make the dapper and get feedback. There is probably good enough. that's my gut response to it just in terms of the amount of process effort for the results we want.
Kyle McCormick: I…
Robert Raposa: I mean.
Kyle McCormick: if it was a year ago today, I would say Yes Robert but I think we already captured the most contentious feedback The editors I don't think the other pages are going to be nearly as contentious.
Feanil Patel: Yeah, but it's a good thought. Alright, we are at 9:59.
Robert Raposa: Do that.
Feanil Patel: thank you everybody for The Rousing Discussion, I think, we're sort of figuring this out, but if we can make the product feed the product process a little bit. Better tied into some of the removal and deprecation that will be useful. Way of operating in the future.
Sarina Canelake: Yeah, I made the meeting notes and notes and tagged you in them. So Put that on your calendar.
Feanil Patel: Perfect, thank you. I put the meeting on my calendar.
Feanil Patel: If anybody else would like an invite, I mean, I'm happy to add you now or you can add yourself. It's on the Open at ex Working Groups, Calendar.
Jeremy Ristau: So, thanks everybody..
Robert Raposa: Make.
Feanil Patel: I think everyone.
Sarina Canelake: Yeah. Have a good name,…
Kyle McCormick: yeah, thank
Sarina Canelake: Dave, whatever.
Kyle McCormick: Have a good time.
Feanil Patel: Yeah.
Feanil Patel: That's the presentation.
Meeting ended after 00:59:39 👋
This editable transcript was computer generated and might contain errors. People can also change the text after it was created.