Technical Oversight Committee, September 15 2026

Good boy. >> Hello. >> My cat My cat is very excited that I'm up early and on a meeting at this hour. cat thinks this is great development. >> I get them working on some uh RFC reviews. >> Definitely bad as well. >> Okay, we are complete. Oh, I I don't see Chris. Sorry. Did he mention that he he's not available this week? Yeah, just pink Chris. I had some issues to edit the word uh yeah Google Docs today somehow. now I will suggest that we should start and then uh Chris will join most probably. Okay. Hello to TC meeting 15 of September again uh our usual agenda without any uh working group updates. So we have next week uh the CF submit and [clears throat] we decide to not have updates working group updates until CF submit and next week we will skip also the meeting uh because of the submit. Uh so that's mean we have the review action action item from last week go over the dashboard and any other business. So action item from last week we didn't have any. So we can jump to the dashboard. Okay. on the dashboard. The first one in progress archiving having kurifi. >> Yeah, I have >> I have addressed the comment from Stefan [clears throat] >> and uh I guess the PR is ready to be merged. [clears throat] >> Okay. Yeah, Stefan approved already. So yeah when this is merched I guess that automation would automatically have and uh next steps would be to retire the GCP and AWS projects. But I don't think that I can do them myself. >> Uh you mean to delete the projects? Um >> yeah, I don't know what the exact term is, but yeah, I guess delete. >> You already uh uh destroyed automations from your site or >> Yeah. >> Mhm. [clears throat] So I think that the project should not be generating any costs unless I have missed something. Okay, I can check. I can check. And um regarding the GCP and AWS, you said um >> yeah, >> I can [clears throat] uh maybe you can put uh um I think Chris can manage those. Uh just provide the project names. to Chris. >> All right, cool. I will >> and we had some uh also on the uh cloud foundry orc site some references. I think those uh RAM most probably you can update or Chris >> Yeah. Yeah. Uh we are talking about Kurifi archiving Kurifi and there are also some references on the cloudfoundry.org or web page uh for Kurifi. We will need to update those also. >> Yeah, I can uh handle that. There's a larger effort that Chris and I have been talking about where um we're going to change most of the things that you see on the website right now. So um this cut will definitely uh be part of that larger effort. Yeah, we can discuss uh to how to u position the other approach we have now for deploying CF on Kubernetes. But uh now we will see some presentation during CF submit and then we can discuss. Okay. Then uh thank you Daniel uh for the contributions uh [clears throat] uh for the project and yeah um unfortunately um [clears throat] this yeah it's been a great project but [clears throat] everything has a beginning and an end, right? >> Yeah, I think it was a good project to learn quite a lot of stuff. >> Yeah. >> Okay. Then we can go to the next one. Inactive users. Um we have this was last week so we need to wait another week uh here until we uh approve. I just ping uh uh some of the working group leads to look into uh the report they have uh so and open PRs if needed. So here is another one uh about uh >> inactive users. I noticed Paul Paul was listed there and he commented that he's still running the vulnerability working working group. >> Yeah, I discussed I discussed with Paul. Uh so they are um so Paul uh they don't have much GitHub activities in the vulnerability management working group and it's fine. Um yeah, maybe I can look to [gasps] uh take out uh uh so I have the wrong one. I can look into the automation maybe to take that working group out. Also the uh pet working group is not accurate um because it's not working in the petro organization. So maybe that one. So we know about some edge cases here. It would be able to >> Okay, we can ask [clears throat] Paul to create an issue or something too. >> Yes. Yeah. >> Cool. All right. Just wanted to make sure we were tracking that. >> Yeah. Yeah. An issue would be great. >> Okay. uh asked to create a new repository for the uh new AI working group so to manage their skills. >> I think my only when I looked at this what the this seems like this is this is a generic like skills repo. So it's sort of a top level idea of skills. So I was I'm expecting that they're expanding this into skills that they're providing for like for people who for end users of Cloud Foundry. Is that reasonable to assume that? I guess the the sample skill was seemed like much more of a development skill than like but it was a generically useful development skill for all teams or any teams working like or multiple teams I guess. I don't know if that does that does that make sense as I thought as I was reviewing that. I'm not sure. >> Maybe who is the target audience of this repo. >> Right. >> So the CF developers or um they wanted to put also ask you how uh to set up CF locally easily. [gasps] So basically CF developers and users all the working groups not only for the agentic working group. >> Yeah. I guess it's like you could see that there's potentially one marketable set of skills for people consuming Cloud Foundry and like as a user and then there's another set of skills for people who are developing Cloud Foundry and those aren't necessarily overlapping skills. >> Yes. So you mean uh they should be separated in a different repo or um [clears throat] >> I mean or I I don't necessarily know what the exact implementation looks like. It it's more I think of just a a question about like what is the organization that that we're looking for here because it could I mean it could even be one repo and just have two directories or something in it. >> Yes. something like that or you know multiple directories because you know we have multiple personas or whatever. So, do you have a like platform operator skills drawer and a app developer skills drawer >> and a cloud foundry developer or I don't know it's hard to say these words but cloud foundry contributor uh skills drawer. >> Yeah. Uh definitely the initial set are targeting the developers called foundry contributors. Um but uh I could um yeah discussion point uh when we uh uh define the structure of the repository so we can discuss this aspect. Cool. So, you're fine with this, correct? Or >> Yeah, I'm fine with it, but my GitHub's not fine with it because it's taking forever to load the review. >> Okay, no worries. Then I can go. Okay, I didn't watch much that one. Okay. Then RFC's so aric yesterday. There's uh some progress on the uh um credits. So there we will get at least credits for one year to see how this goes um to cover um and then see afterwards. Is that uh okay to go for final comment period with that? >> Yeah, I think so. It's more question to the governance board at the end of the day, right? tech point of view it looks fine and >> yeah we have to get more experience understand uh what so I was talking to Sachin he said that uh this local setup is working fine maybe we have more opportunities here to validate GitHub actions many things and not go over costly uh concourse validations uh we have to get more creative Because uh uh if we validate like we validate currently then the costs uh high. Okay. Uh it looks like people are finally starting final comment period on this. >> Yeah, sounds good. Okay. Um the next one is about Google distributed call uh code. I have kept um we had okay there is um [clears throat] so the comments are actually resolved uh on this one. So um there was one about whether people will get access to that infrastructure. I don't know whether that aspect has been updated. It's one who's maintaining this long term. >> Um so long term >> so u ownership model between Google and SAP. >> Okay. Are there any current committers who will be working on it? >> Yes, this is the list. >> This is the list they propose. >> Okay, Ned. Ned's there. Okay. And the idea is to have a separate uh uh work uh area in the foundation infrastructure working group owning the stem cell the two repositories here. I see do we know if having a separate stem cell builder is intentional? Why? I guess it's a Linux. We've every other infrastructure we've handled with a like the same Bosch Linux stem cell builder repo builds the stem cell and then they all stay pretty much in sync. So having a separate repo instead of another flavor in the current stem cell world concerns me and I don't know what the I don't think it I apologize I just skimmed this. I don't think it was covered here why we wouldn't do that and I can ask a question on it too. >> Yeah. Yes, you can ask. Um, yeah, I didn't question that part. Yeah, >> I'll put a question on it. >> But, uh, you are right. We had also many different flavors in the past. Spending a lot of time working on stem cells. Stem cell drift is not not something we want. We can avoid it. Okay then uh still some discussion I think uh final comment period is that uh um should [clears throat] we wait them for the stem cell discussion repository? >> Uh I guess we don't have TSC next week. Yes. >> Um >> I don't know what what's the what are the rules around ch updating a proposal while it's in the final comment period. Does that re that just resets the final comment period, right? Or does it nec it doesn't have to unless it's does it have to? >> I'm fine starting the clock and having a conversation while we have the clock running to get things keep things moving. I don't want to hold things up. Um >> yeah, >> the question is whether it is a blocking uh comment or not, whether we say we have to agree on that before it's accepted. >> We get to make up our own rules. So, uh >> I'm fine starting the starting starting. They keep things moving. >> Okay. Uh and I think we can see if we see that the agreement is needed to accept that one then we can postpone the uh yeah Yeah, please comment on that material so that we can >> Oh, well then we have another one, a [clears throat] new one. Um so this is basically about providing uh platform walks over hotel. Yeah. So at the moment we have only the matrix [sighs] of hotel and this is to extend that support for uh platform locks. Uh I don't know whether uh do you have any comments on this or um I I have some abstract concerns about merging all the and I think I think it's addressed down towards the bottom at least of merging all the platform logs as well as the application logs. I know we already do this with metrics. So it's already a little bit conflated. >> So this is about uh platform logs. It's not going to uh match the application logs. Uh >> well, application logs are flowing through the hotel collector already and so we'd end up with all the for logs and we'd end up with platform logs and metrics and app logs and metrics flowing through it, which could be great. Like I don't want to say it's bad. Um but it just changes my worldview a little bit of how what I expect. That's I'm just ingesting it and trying to think of consequences. >> Address application works in a second one where we need also to have some tenant concept. Interesting. I mean my recollection it's been a little while is that application logs do go to the hotel collector but they aren't so you can you can have the equivalent of a drain through the or a sis not a dra um a fire hose connection of like every you can have all the application logs in the hotel collector but there is no the tenency model is weak so replacing all of you can't replace like application specific sys log drains Correct. >> So uh you please comment but my understanding was that we are going to just uh uh get here the platform locks uh and then Bosch based works um and for the application works additional work is needed to provide the tenant concept. So how we can um make sure that you get the works only you have access to and so on. Um >> yeah there's yeah there's there's no tenency at all. >> Yeah. >> Okay. And our last one um this is basically RFC proposing to dis to discuss the ownership of the Kubernetes release assets which uh [sighs] this CFN kind RFC uh um introduced in the CF deployment working group. So um yeah people are seeing this more in the ownership of the uh working groups the components belongs to so they are more aware of any changes introduced changes and can keep those in zoom. So this is what say basically proposing that um migration towards the working groups. I think the direction is for me okay what I miss a little bit is uh some recommendations documentations on which uh tools and techniques you want to use to mirror the uh equivalents that we have in the Bosch world. So bosch links for instance the config parameters ops files and variables. These are the the terms that everybody knows. Um they have certain semantics to for instance keep the system domain across all Bosch deployments exactly the same. And it would be good if we would um give a little bit guidance how that could look like in um in the Helm charts because otherwise you have a big mess of overlay and funny named um attributes and configuration parameters and um so the car setup is a bit it works but it's a little bit unstructured I would say. I plan to talk directly to the folks that have invented that too. >> Yeah. >> To have here a bit more more guidance before we roll it out to everybody because otherwise I can slightly imagine what happens. >> Yeah. So there are also some uh talks during this year submit around this topic. you can directly talk there also. >> Okay. Then uh yeah, please put your comments in the PR [clears throat] >> and this is all about RFC's. Um so on the status side, I don't think we have any updates yet. um issues. This is still the one uh will be addressed with the RFC for GDC Google distributed code. Uh this one is still open. Uh we know uh needs to be addressed. This is a new one. Um this works to me the uh wrong repository. [laughter] I I was I was trying trying maybe somebody sourced it. Okay, I will put a comment to the meeting. You don't know need to watch me uh writing. Uh yeah, the rest uh um I think we uh need to close the old ones at some point. Um somehow to have a clean dashboard. Um yeah. Okay, I will comment on this and the other one uh are in progress. Okay, then any other business we have topics. So my last reminder to register for cloud foundry summit. Um we'll likely be at summit this time next week. So if you've not registered yet, please do. If you have teammates, friends, I don't know, enemies, [laughter] whoever you know, uh please have them register for you. So quite literally at the very last link. >> Yeah. >> Hope it goes well next week. I'm sorry. I'm not going to be there. >> [gasps] >> I think we canled the meeting already somewhere. I asked if I'm not mistaken. We can shortly check. Oh, it's all still there the meeting so I am wrong I uh mixed with the vacation time period [clears throat] could be could you uh run be able to cancel the meeting for next Tuesday then or uh is there are there people uh so I'm sorry uh maybe uh Matthew Greg and Chris you are already um a forum. Yeah. Are you available? Uh >> I I think it's probably good to defer for a week just to give folks that are attending the conference, you know. >> Mhm. >> Time to enjoy that. I don't think there's anything else that at least I'm not seeing anything that's pending that's got to get a decision made next week. No, [clears throat] it can be also discussed as if needed. >> That's fine. >> I'm canceling next week. Makes sense. [clears throat] >> Okay then. Yeah. Are you able? Could you be able? >> It should be gone. Just refresh your calendar. You cancelled already >> as you were talking. Yes. >> Okay. Thank you. >> Can you reload that page? Can we see it? Gone. >> Okay. >> Want to make sure it all went. >> Yes. >> Okay. This one is to be cancelled as well. Application runtime interfaces working group [snorts] or you want to run. [laughter] >> Got it. >> That's all right. >> Yeah. same justification >> then let's cancel this one as well. >> Okay. Any other business? Any other topic? I'm sorry. >> No. >> Okay then. Uh we are done for today. Thanks for your time. Take care. See you in two weeks then [clears throat] or during C submit. >> Okay. See you. Bye-bye. Bye-bye.