GitLab Plan Stage, 1 May 2019

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Plan group weekly meeting

Description

Plan group weekly meeting for 2019-05-01.

A

Yeah so, as I mentioned, I wasn't able to make it here last week because of illness and where we talked about 12 data, so I wanted to bring that up again. We've got you know we're supposed to have a general direction of things. We want to do in twelve dot, oh by the first, today's the first, but it doesn't necessarily have to be finalized until the until the kickoff meeting on the eighth on the eighth. So I wanted to quickly look through the board, and so the board is linked there in the agenda item.

A

It feels like there's a lot on here and unfortunately this is my first time doing, planning without a picture and so Shawn. Thanks for your comment, so it looks like we're a little bit over subscribed, and so we need to. We need to cut some things. I don't know if, if there's things that are easy targets to cut but I wanted to at least have a discussion on it, because this is my first time and you all have much more in store for context than I do yeah.

B

So last time I looked at the list, but what I missed was there? So I was searching by feature and why missed was that a few of the features didn't have the label feature included, included some big ones, so I think that's that's probably where we should start looking at. So we've got a value stream analytics item which is potentially moving to manage. There's a mode request about that. But I don't think we should make that affect this. This planning too much. There is manual backlog, grooming and prioritization, and those are the two big ones.

B

I would say: I would say all of those two. We can probably only do one like regardless of flake. You know the other other capacity issue that we have so I would love to do manual, backlog, grooming and prioritization, because it would be super useful for us. Yeah I wasn't actually aware it was ready, but it does say UX ready. So you know. If it is we should we should go for it. Yeah.

A

So VSM that the SMS issue did not have the feature label I noticed that as I was watching the recording last week and so I added it. So it now shows up on that when you, when you filter by feature I, do think it's something I do think it's something with the proposal to move DSM to manage and to consolidate our analytics story. I think that we should definitely prioritize the manual backlog grooming over this over this issue, and so what do you think makes most sense? Shawn?

A

Is it to move this to 12.1 ORS is to drop a milestone, I move it to backlog.

A

B

Preference, just because I'm lazy is that if we think we want to talk about this again in future, it's easier to move it to the next milestone. I, don't think you know the problem with bumping things along milestones is sort of the public image it sends, but I think there's not a huge amount of participants in that issue.

B

So it's probably pretty clear late. You know if we just say like we're, moving this to talk that long as we can't go in 12, oh, but we don't want to get about it. That's that's the easier for me personally, like that's how I manage like some technical debt issues as well, like I, will put them on a future milestone and then, if it turns out, we don't have capacity in that milestone, I'll bump them out, but to a few another future milestones, we don't forget them entirely and then also we get the you know.

B

If we just bumped it five milestones in a row saying in the motor in the issue, then that's probably a sign that we don't actually care about it that much so that's my general preference yeah. Okay. So.

A

We'll push that one out I just changed the milestone to top that one and then for me, an annual backlog grooming, which I think is useful for a bunch of people and we have the internal customer label on that one as well. We'll keep that one for 12.0.

A

C

A

On just moving that one out does that give this to a happy spot, or do we need to trim more that.

B

Was weight 5, so we need to trim more, but I think we can trim holistically. Like you know, we could probably trim some bugs the ones that are lower priority on there. We could probably trim some technical debt. That's lower priority, some features that are lower priority stuff, like that, so I think the trimming in general, like I, don't want to cut too much off the top I would rather cut from like the bottom of, like you know, what's the lowest priority bug that we're doing? Can we move that?

B

How was the lot of courage to take that issue that we're doing? Can we move out, because that would be that she goes quite well with the issue we're just talking about with the manual backlog, grooming and prioritization. If you just have a single list and you can just cut the line right exactly.

A

Exactly and it's funny because I'm asking to like add something in that you know Eric Johnson has asked for which is the board list capacity line which, if anyone went to the engineering group conversation you'd, know that that's kind of how the engineering leaders are managing their own work in product we're doing the same thing. We have a board for all of our product leadership items and we're just drawing a manual line. So this would be really useful, and so we can talk about that later.

A

But just so, and just so I'm clear, it sounds like we have a capacity of 75 weights for 12, that our bows are wait for 1101.

B

That so, okay, so just just first of all, that's for back end! Second of all for 11.11 it was I, don't remember the exact number I had it written down earlier. It was a hundred and something like five hundred and twenty or something like that. But we've got like a big impact from contribute. We've got a big impact from the and not working on plan and Mario, not working on plan. So that's the main issue.

B

We do have someone else starting next week on the back end team, but obviously we'll take them awhile to get up speed, especially given that some their first weeks at contribute, so yeah, yeah, sure, Donald, I, don't know if you've had a chance to look at front-end capacity. If it's a similar story or if it's different I.

D

Was going through the switch rate, as you said, oh.

B

You know just asking about front-end capacity if you've had a chance to look yet or if you have a sort of idea of like. Are we way over and $12 at the moment? Are we about right? What's.

D

You something yeah.

B

I could see you having airport issues I have the gate.

D

Honey would do here myself.

D

I I still need to wait a couple of the the front end issues in 12, oh, but as far as Purvis we calculated how many we did 11 11. So far, it looks like so far. We were around on the front end. Only at 31 we typically get around if 11 10 we got 50 done with contribute with three funded engineers going.

D

Wait should be I'm guessing around already 45 foretold. That up like I, said I still need to wait out a couple of the ones in 12.0, but looking at what was weighted just front end we're at 43 right now. So there's probably a few we're gonna have to take out on the front end. Only issues awesome.

B

Okay and on this one we were talking about which sounds like it would be kind of a headline feature about the manual backlog, grooming and prioritization I. Think that's more front-end than it is back in because or back-end. We already have the ability to order issues on a board, and this is just basically ordering issues not on the board. So I think there is some backend work there, but I think the majority of it is front-end so I'll.

B

Let you sort of sort of go into that in detail after this call and figure out like if we should be able to do that or not.

D

Perfect, that sounds good.

A

Okay, so that was all I wanted to talk about 412 that oh I can work asynchronously after this meeting to kind of continue to and Shawn and Donnell. Oh all, I see need your input there, but unless there's more to do synchronously with the team I think we can move on 412 letter. Does it make sense yep.

B

Just I know we're gonna go on to this next, but if we want to add stuff after I think it's probably best to like, do it and trim and then potentially add these things that we're talking about, but do those are swapped, so you know, we've got the capacity sorted and then we exchange things. I think is the easiest sort of model there yeah great okay,.

A

Let's move on I had the next agenda item too then. So, let's move on. There's there's three things that have popped up that you know since I've, taken over for Viktor there's three things that have had popped up in the past week and I just wanted to get the team's thoughts on them. The first one is instance, level epochs, which is directly from a large customer. It's it's basically asking for what I would say is is almost like program management view of various departments within their gitlab instance.

A

This is definitely by no means I almost put like whip-like on the front of this issue. If that was a thing like this would be like a whip issue, because I'm not sure if this is a good idea, I think there's other ways to potentially solve for this, but I think that it's interesting and we also have this issue called group level. Things like the issue is called group level things and I almost thought of.

A

Maybe there needs to be an issue called instance level things where we expose functionality features at the instance level, and so I just do this here. We don't necessarily need to discuss it synchronously on the call, but I wanted to make everyone aware of of this issue that I created yesterday yeah.

B

I'm not gonna, read my whole comment here, but I'm gonna, sort of just outlined sort of how I was thinking about this. So one way you could model this in get live right. Now is the exact situation they have at the moment, but with every group they have per department in the company inside one top-level group, so you've got your instance. You've got a single top-level group for the entire company and then everything inside that and then epics can be created at the top level. They will work, etc, etc.

B

You know group boards will work there, etc, etc. I don't think asking people to do that, isn't necessarily a great fit because it sort of defeats the point of having top-level groups. If our recommendation is only have one per instance. Also, it doesn't work for us at Ghaleb, because we have a bunch of top-level groups most notably get lab or get up, and also the charts repository. We have four helm, charts and stuff. So.

B

The idea I had, which needs more thought is that you could have a way to create a group. So actually let me rewind a bit so the way it works right now is for a label, a group label, a group milestone an epoch or a group board things flow down the groups, so you can have a parent group. You can have childhood subgroups and subgroups can use labels from parent groups. Ethics can contain issues from child groups. Group boards show you issues from subgroups as well.

B

So the problem here is that top-level groups don't have a parent, but I thought. Maybe it's possible to create another group that could have top-level groups as its children, but not have regular children and not do the membership access inheritance flowing down thing that happens right now, so, for instance, a gay lab. We have a group that we have some groups that are at the top level, because we don't want everybody at gate, lab to be able to access that group and if you have access to get lab comm.

B

If you create a sub group in there, anybody who has X you get the highest of whatever your access is on all these groups right and if you've got AI access on the parent group, you will have it all the way down. So I think that would also potentially help with that problem.

B

Now that's obviously got a bunch of issues. I've listed some of them there. One I just thought about right now was on get lab. Comm specifically, would we ask people to pay for tuned groups to his feature because, like you, paper user in your group right, so that's a problem as well.

B

But that's that's the way I was thinking about it because I think instance, level works fine for single tenant instances, but it doesn't really work for multi tenant like it lab comm, and it would be great if we could solve this for ourselves at the same time, because we have this issue where we have our website repository in the get lab, comm group and application source code in the gate, lab or group, for instance, and if people are working on both of those things, they can't show up on the same issue board because they're not in the same parent group right now.

B

So that's that's what I was thinking. There are other ways we could potentially solve it, but I think that one sort of probably worked best for all the plans- stuff at least but I, don't know about other stage groups and whether that makes sense at all and we definitely need some feedback from manage which a store as well. Yes, although.

A

Yeah one of the things I want to make sure we don't get into is having our end users basically having to get a PhD in how get lab groups and subgroups and projects work hierarchically. Just in order to get the thing to work, how they needed to work from a permissions and visibility perspective. And so that's you know the one of the biggest complaints that other project management tools have have out.

A

There is like you have to like literally understand a second query language or like a second workflow engine, just to understand and get your thing to work within the tool, and so it.

B

Was tell me you used earlier sorry like when you were introducing this issue initially, when you were saying like we need a sort of, like instance, level, whatever section I can go back and watch the recording later, but I'm trying to remember what the specific term you used was. There's something management, I think Oh program management, program management yeah. So we could even call these groups like program management groups or something like that. Maybe to make it clearer, I, don't know I'm still thinking about it. Yeah.

E

But but typically program management is a middle level in the hierarchy, so group today Maps almost perfectly to what and I'm writing the IDC response right now and I'm. Calling the group's program management, okay, Leo portfolio, would be the thing at the top of it, making best of decisions as to what things were going to work on later on so hierarchically, it would be perfectly a program project and so I, don't know. I'm I'm scratching my head on this one, because I at least shown my thoughts on your answer.

E

That says, if that's what you want then put everything underneath the top-level group and you have all the functionality you want it's there, I mean I, don't understand in the description. I haven't read through your comments and details. You've looked a lot, but it seems to me that we're just labeling it something different at the top level in since level and move on. You know, I mean I, don't know. I've got to read through in detail. Yeah yeah.

F

Right there and.

E

I agree with Eric: you don't want people to have to become. Did you have a PhD and get lab to use your lab? Yes,.

A

So I would recommend just in the essence of time, and we have a bunch of other agenda items. Let's just communicate asynchronously on this one I'm not looking to schedule this. You know this this release, so it's okay for it to be a little bit fuzzy right now, but I welcome your comments there. Let's move on the second one, I have in the list, so this one I would like to activate on sooner rather than later so board list capacity line.

A

This is a this is kind of like and at what I would call an MVC of what was originally written down is like managing board capacity, and so there's a there's another issue which outlines like a number of various proposals on how to manage board capacity. This is an incredibly pared down version that just says like just give me a line or the ability to make like a UI element where issues can be above and below a line that can be dragged and dropped and Shawn you had a question: I responded to it.

A

The answer is 100% option. Two in your in your issue. Comment there I don't want this line to have like hidden functionality with respect to capacity management. I just want the user to be able to drag and drop things above and below it and for them to manage it, and so the ask would be you know is this: is this in a place where we could potentially get the scheduled for 12:00 do as a swamp, or do we need more UX work to run a sprint ahead and then schedule a search one?

A

So I wanted to bring that up, since this is my first time going through this process with this team I think we need ordering done prior to.

D

This right, what.

B

Do you think so this this one's about boards and the other ones about, so you can already order on boards and the other ones about ordering on a ListView. So they are not it wouldn't. Listen. People wouldn't necessarily be stepping on each other's toes. If they were working on these two together, I think is that right, Eric, sorry, the line goes in the board because I think it.

G

Would make mind goes in them, it.

B

Would also make sense for the line to go in the backlog, for you we're talking about in the other issue, but I guess we can add that, like.

A

Yes, that's not what this is for, though this is specifically for boards. Okay,.

F

So somehow we need we've been, will we we will need some sort of you expert to like a button that says add a line and then you drag it around. I guess correct.

A

Yeah there's one thing: I have outstanding I'm, gonna post an issue comment right now and I. Guess it's more of a question which is: should we also allow for naming of the line and then subsequent actions like an edit of a name? Should we actually allow multiple lines, I would say for an NBC. We should just allow like one line per list, but we wanted to get the team's thoughts on that as well. When.

B

We first added boards, we have one board project and you couldn't name it so I think that map's pretty well to that as well. Cool.

A

All right, I'm gonna, make a decision and just say, or in the issue comment just say for this NBC we will not allow naming of the line nor multiple lines per list. I have.

F

A question I haven't actually read through the whole issue yet, but is this? How is this different than creating multiple boards for different releases like what the line signifying that, but it probably won't get done if you're below this line yeah, so would that not be on a separate issue more like if you create a 12-point ou issue board with all your issues, and then you create 12.1 with all those so.

A

The only thing there is there's a use case to where someone wants to see like what won't get done on the same list like they want to be cognizant and what's behind the capacity of the team and just be able to visually, see that on the same board,.

F

B

I think the way we can think about this is the conversation we just had before about like how much weight was on 12.0. If we imagine for a second, these worked with weight, which is a future future iteration. Like you know, we could have that that list of issues- and we could say well, aligns like you know we could say well. The line for engineering is here, so it can then see well.

B

I need to move stuff and, above and below that line and then take everything below that line, and then, when we start implementing yeah that goes into a different, different milestone. But it's useful for the point before that, where you're actually trying to figure out what what you're gonna be able to take at all in the milestone.

B

E

Makes a lot of sense if we can imagine when we talk about imagining the line where it's set, based upon the weight or the count of the things above it such that the lion is either 5 issues or a weight of 10. You know, or the line will move move around based on that. That would be where this gets really interesting. Yes,.

B

That's the second issue: I think that Eric's extracted this film to like create a material. That's right.

A

I don't know the answer to that. Alexis I do know that, like one of the the best use cases I've seen for this, one modeled is like Eric Johnston and his engineering leadership work. Where he's like literally created a dummy issue that the title of the issue is called the line and it's a bunch of dashes and he's using that issue to essentially create his own line, which I think is an interesting use case of like a small, primitive, like you're, using an issue for this. But it's also like not not super ideal.

C

A

I'll try to find that board and and if you won't go back and watch the group conversation, he actually like shares his screen in this one, because we talked about dogfooding I asked him a question about dogfooding and how he's I think the question I asked was: what are you most excited about? You know two dog food and get lad that you kind of can't today and he gives an answer and he's like. But here's like another example of something that we are dog fooding, but it's not a great experience and he shares a screen.

A

So I have to find the board ish the board, link and I'll. Do that and I'll put that in this issue. Thank.

C

A

Okay, cool, let's move on so I would like to understand. If we can do this in 12 that, oh so Annabelle I guess it sounds like you have some outstanding questions, I don't know if we can just work in the issue on this one I'm.

F

Just yeah, we can't work in the issue it just we I'd like to make it. You know a clean experience. It's not like a random button, that's like add a line or I. Don't know that's a little awkward so and with the summit coming excuse me contribute coming up. I don't want to rush it like we need.

A

To do design and development.

F

In the same milestone where it have to change it, yeah yeah.

A

I understand so I know this is a new one. I will.

A

What it's not completely clear to me right now is like the the sequencing of events for this team and once again, I apologize for kind of coming in late, but the sequencing of, like you, know, design work and then that funneling into back-end that needs to happen across two releases. If that can sometimes happen in one release, I just I'll lean on on you also, let me say: help guide me and figure that out.

D

Yeah, ideally with typically, unless it's something that requires just a small amount of UX, we like to work a release ahead, so if we can, if we can slot this one in if time line, if that's okay, if we can slot this one in for called dado on the UX side and then just plan on doing that, good development, twelve one, that would be the preferred path.

D

If that doesn't work as far as timelines, we possibly can make it work all in the same release. Yeah.

B

Also like in general, I feel like we try to just do stuff asynchronously throughout the milestone anyway. So not necessarily like you know UX, can you drop everything and look at this, but you ex? Can you look at this like by? Obviously we couldn't do that in this case, but you know what I mean like it. If we could have said to Annabelle and I like this, like? Can you look at this a couple of weeks ago?

B

Maybe it doesn't need like maybe it does, but maybe it doesn't need like as much UX attention if they can answer those questions, or maybe it does, that they'll be in a better place to answer that, if we earlier, we can ask the questions, so it's not necessarily about a release ahead. It's more about just like you know, get the question out as soon as possible, and then they can sort of manage it with their other with their other tasks, cool.

A

I'm gonna go ahead and just skip the next item under and we can move on to number three.

F

All right, so it's up to me now, yeah.

A

That's you sorry, I.

F

Just wanted to let the team know that the UX is the UX team is doing this UX audit to make sure that our website won't get logged as accessible as possible, and we started with color checking the contrast and making sure that none of our elements are using only color to convey a meaning, and you know across gitlab. We do that a lot so we're trying to fix these.

F

So we've Alex, this and I have gone through and identified all the places within plan that we have these sort of UX issues and you will be working over the next few releases. I, don't know how many releases this is supposed to span, but we're gonna be adding a few issues into each milestone, and these are mostly UX issues. I believe, like you know, darkening colors and making sure contrast is where.

C

F

To be, but that will affect the UX capacity for the next few releases. This is something that we were supposed to work with Victor on, but he's not here, so we're not really sure how to plan right now. So I just wanted to give the team the heads up.

A

And about when you say, work with Victor on what does? What does that look like? What does that mean? How can I help there? We.

F

Were told to so when we create all these issues and identifying, we identified solutions for as many as possible if they were small enough, some require a lot more work and then we were just kind of told to work with our p.m. to schedule. These issues I still need to find out a little bit more about what that means, and how long?

F

When we're supposed to have everything finished, because that would help a lot and then based on the capacity that we already have for each upcoming milestone figure out which issues UX could start working on along with all of the actual features that were currently still working on so right now, I, don't think, there's anything to do. I I need to get more clarity on scheduling.

A

Okay um feel free to reach out to me if you need help there. If there's anything, I can do.

F

Onto Donald.

D

Yeah so I realized I, failed and forgot to invite him to this meeting. Sorry so listen this next week, but welcome. We have a new friend and engineer Martin. You.

B

D

It on YouTube.

B

Yeah, that's true: I thought he can record his own YouTube video, introducing himself Berger but I'll move into next week, but yeah a new friend lived in here, yeah cool Annabelle. Do you want to take Pedro sytem.

F

Yeah uh I just I, saw this yesterday, okay, Pedro has moved permanently to the create team, so we will have an opening for another plan. Ux designer I assume, because now crate has three and implant used to had three he's doing this, because I can provide the historically overbeck's background of gitlab Inc in a plan, and there isn't anyone and create for that because you get there. Two designers are very new and he's still going to help contribute to plan because they are sibling stages. I haven't heard of that. That's really that makes sense.

F

Yeah, that's bad.

F

B

So I thought this would be good point to check in with crucial and Brett so Brett's, not on the call. But I did speak to him about this bit yesterday because we decided to implement the epic tree view in epics in graphic you all from the starts from the back-end perspective. It's kind of hard to answer this question right now, because we have to do a bunch of setup work which we knew we would for this issue because we didn't have, for instance, the concept of a group or an epic ingratitude.

B

Well before this and next time we try and implement an ethics feature in graph QL. We won't have to implement those again, because the whole point is we've just done that, but it seems to have gone okay, so far, I don't think there are any major issues. Crucial I think you're still on the call I was wondering if you could speak carrot felt from the front end. Yes,.

C

So, as far as fronting implementation was concerned, it was it was rather easy to work with, especially that so most of the times when we worked in bits back in API for any front, an app should help you deal with different structure of key names or in broad JSON response that we received. That isn't the case, because, with graph QL, it is predictable how the structure of response would be so you wouldn't have to parse much.

C

The data that you want in a certain structure is what you mentioned in the graph query itself and that's what you get so it is easier to work in that way, but at the same time, what I understand? How graph you will is beneficial for any front-end development. Is that any particular API key or item that you need in response? So you just go ahead and update the query and it is readily available. I think that doesn't work directly in our case, and that is something that I discussed with Bob today afternoon as well.

C

Is because that all the data that we are querying viagra fuel requires to go through the rails controllers and helpers just to make sure that all the privileges are in place. So, for example, if I would need a certain key or a certain property in an object to be present while querying the data, if it isn't supported, then I cannot just void and update the query and expect it to be available. I would have to make sure that the backend has implemented it correctly and in case the query fails.

C

I would let Brett know that this is something that we need from the backend. So in a way it is still similar to what we had earlier with the traditional API is where any changes that we would need in the response. We would communicate it with the back-end developer instead of us as a front-end engineers directly go ahead and update the query. So that is the only missing part, but I think it is still very initial stage of graphical support within good lab I.

C

Think once we reach to the point where it is widely available across the app, it should be much easier to work on any query without having any back-end developer manually, intervene and make sure that the data is available. In.

B

The response that yeah- that's that's, pretty much matches up with what Brett was saying as well. So just just to sort of ask you a hypothetical question which I didn't prepare before if we would implement another ethics feature using this graph, QL API. Do you think that would be quicker now that we've done this work, assuming that this was merged, which is not because we're still working on? Yes,.

C

Yes, because right now, graph QL is being used only for the tree tab. So the more that you link. It currently adds pre tab to the page body of ethics and we are not using that same response for the other tabs or for the page body for that matter.

C

But if we were to move that entire graph QL querying part to the root page itself like if we were to get the data on the page road itself, then we don't have to make any additional queries for the rendering of the page, because all the data is available on the page road itself. Having said that, what we can do in the next release, or in next couple of releases without doing a major refactor, is that for both the tabs, the road map as well as speedcap.

C

We can leverage the same network because that we are making and we can reuse that response to render both the road map page as well as road map tab, as well as the tree tab. Because right now there would be two different requests to face. The data for the effects, one for the road map and one for the tree, but since this, mr is already adding lot of front-end and back-end code to just get this free time running, adding another emerges to make sure that whole map also uses the same.

C

Query would be too much for the release. So I would definitely open technical data issue just to save on one additional query, because we don't want to again query exact same data if users, which is the tab, although the request happens only once when you click for the on the tab for the first time after that, any time switching or doesn't involve any query. But since we are already querying the data, why not query it on page load and then use that response for both three, as well as room yeah.

B

I think we actually have an issue already open about that for the road map, because I think Heinrich created an issue said that said that we're just like sending way too much data for the front end for the road map like it, doesn't need it, and we said we'd just like postpone that until we have graph QL implemented here, because then the front-end could just you know, request this period. He did instead of like after yeah we to ask the backend for a generic thing, and then the backend gives it a bunch of stuff.

B

It doesn't need, but that's cool. Thank you. Thank you very much. For that summary. Crucial I have.

G

A question in regards to to what gosh I've mentioned before that, if we need something else in the front-end would need to change the structure of the graph key well query: is the backend being implemented in a way that it automatically generates the documentation of the model, the API, so that we could check that to see what is available? No.

B

We have an issue in $12.00 about that. Alright, alright, because yeah, that's a big mixing gap at the moment, so.

C

So right now what I do to test the query, whether all the values that I'm looking for are present or not? Is that I just go to graphic UL page for the instance and there whenever I type, something within the query object? It would aught of the options available and that sort of helps a lot just to see like what are the items that are available and whether it already has something that I need, or so in a way those things are already there.

C

One thing which I think is missing is that for any property type right now the data type of that property's value is defined on the back end, I'm, not sure if I'm missing something here like if there is any way for me to pass the value type within the query that I am using on the front end, then that should be much helpful because right now why you would need to cast it something else. So right now, boolean values are coming in as boolean, so that is fine but ID properties.

C

So, for example, I have a query for the tree where epics and issues list would have ID property, which needs to be numeric, but the response sends it as spring. So I looked at the Apollo graph fuel stocks, whether I can provide data type within my query and back-end would add there to that data type. I, don't think that works from the back end, because the the query type that were the Ruby files for the query, type that we have in the back and already specifies what data type this feeds to be exposed with.

C

So there we have already determined that it needs to be string. So in that case there isn't a way for me to manually override that on the front end or maybe I'm missing the implementation part of how to deal with the query and pass on the custom data yeah.

B

I, don't I, don't actually know the answer to that, but maybe either create an issue or ask in a graphical channel and slack okay, an issue and post in the graphical channel and slack sure, but yeah I'm actually quite surprised the IDS of strings but I'm sure there is a good reason for that.

B

You cool, thank you very much for the update, crucial that was all we had today. So have a great day. Everybody and I'll speak to you. Well. I actually went during this week this meeting next week, because I'll already be flying to New Orleans, but donald's can run it in my absence, Eric after we stopped recording. Have you just got like a couple of minutes to answer some question that can't be on YouTube yeah? Absolutely.