GitLab Design Reviews, 12 Jul 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Dev Manage:Plan:Ecosystem (UX Design Review) — 2021-07-12

Description

No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).

A

So let me try this screen again. The sigma screen this one okay. So I'm trying to do like a test of.

A

Of these two types of inputs with either like an icon or like an actual input field and I'm having trouble with the the way that the overlay works so I'll show you what it looks.

A

A

So the idea is that you can hover this behavior and you can do an input and it'll switch the date here.

A

The problem is this: input fields is different for some reason in terms of the functionality or the mechanics behind the the prototype layout. So.

B

A

Kind of works, so let's change the date, um but there's some weirdness.

A

Let's try it again.

A

B

A

You doing this all at hover right now. uh No, this is on click, so I'm clicking, and I think somebody fixed it. This is weird I really would like drop off weird stuff and move the this frame out of section when I clicked on it. In other places,.

B

Let me try it again.

B

Oh, so it would just push all the content down.

A

A

Show you what was happening.

A

I'm trying to find the screen grab that holly had yeah. Let me open.

B

Up that file too okay, I see here.

B

Let's see if I have that problem.

A

Expired so holly took a screenshot of this, which is the.

A

Problem and you'll see how it kind of moved all this hover over the side and that's this particular frame that I'm trying to play with.

A

But it looks like somebody's gone in here and done some changing.

A

No, but that's a different one. I don't know.

B

This is something.

A

Else so in the prototype that was happening where it would fly over this, the flyover would be weird, it wouldn't work the way I thought it would so. I was trying to figure.

B

A

There was a way to fix this um interactive behavior. Let's see.

B

Oh interesting, okay, I'm kind of cycling through right now.

A

The idea is that, so, if you hover it, it will assume if you click on it, it will open the drop down and then, if you click on that again, it will swap the instance. That's here as an overlay, but the problem is you have to click on the hover, not the overlay.

A

So if you were to just that's where I think the the confusion of this weird image came from, is that because it's actually swapping this.

A

This container, actually his is the better the correct one. Has this empty space on the side, swapping this container, it's trying.

B

A

Do something different with this container, which should be wait a minute? Is he oh? I see so somebody did change this for me. Oh.

B

A

Yeah because it used to be this calendar picker too, oh, so they added someone added that white space too yeah. No, no! I added this one because I was trying to make it work and it didn't and it looks like they fixed it entirely.

A

So I'm gonna ask or say thank you to whoever fixed it. I wear some leave a note.

B

I'm kind of confused at like so I'm just playing with the prototype right now and it looks like only uh one one of the date pickers in each column works. Is that correct? So we'll do this.

A

Yeah, so if you look at it says no expiration set versus the other ones which have an expiration and what I'm trying to do is I want to ask the participant to go to the this person teresa and.

B

A

A date and define a date as a behavior, so I'm trying to see which is the which is a nicer appearance, because we want to change the um the input field because we want to add this color variation to show expiration or an easier way to show expiration, rather than we have here right. So the hover, the popover tooltip, is not even a thing.

A

I've done that as like an intermediary, but the problem is like. If you look through these, you can't really tell with a first glance. You know how soon they're going to expire versus here or has expired right. It's like! Oh, if you're an admin I want to know which ones are going to be expiring soon, so I'd be able to fix them really quickly or I can visually see that, and so I I don't know if I could do this inside this text or if I should or if I should do this new proposal.

A

I probably should have asked that if I can actually just drop in the colored text in here as an mvc instead of having to re-engineer this input, behavior yeah.

B

That might be better yeah and then this is this: what you were going to actually use for, like a user testing session.

A

Yeah exactly I wrote a script to go with it to ask you know: please click on this person's, uh then change.

B

B

uh Got it and then yeah.

A

I think that's a good idea.

B

Like as you're talking out loud, I would probably ask maybe someone on front end engineering like what's the yeah. What's the constraint here? Is it possible just to change that text within the input field because for you know when I was looking at this at first glance, looking at on the right column there, it looks like that's kind of fixed. I guess I didn't know that I could really change it. I guess yeah the right there in my mind, like pressing that little calendar button would create like a new, a whole new date.

B

A

That was my first reason. You mean that this there needs to be a stronger separation between these two.

B

A

B

Well, maybe that too, but actually you know what I'm looking at first, for example in version two on the way right, when you know you have like the expiration there's like a date, and then you have like the button.

A

So you're thinking it would add, like another row of dates here.

B

Yeah, I just thought that that that first column is just kind of like a read only and then you ask adding like another one. I guess it's a little. Maybe the whole thing is like a little out of context yeah, so maybe maybe setting up the session more in, like uh uh like a more of a story like maybe yeah, not showing them all these variations for the user testing, but maybe showing them like an actual like putting it more on a in a flow. I guess this.

B

This might be better or maybe easier to kind of convey like a story.

A

Yeah yeah, that makes more sense, yeah.

B

And I was having trouble with.

A

How I'd want to ask the questions or how instructed to test scenarios? So I that's something the next problem, but.

B

A

The first problem I had.

B

A

Weird overlay that somebody seems to has fixed for me here. It looks like.

B

A

It got it here.

A

Well anyway, it looks like that's resolved. So then I can watch.

B

A

B

On that, on that note I was curious. I haven't used user testing, yet I was going to use it for like a quick session. uh Would you recommend it or one of one of the one of the concerns that I had was with it was? uh Can you the the people that are actually doing the the testing?

B

How how uh let's see like how knowledgeable are they about get lab or how or are.

A

B

General kind of usability issues, or can you can you get really specific with with I've.

A

Had reading you know, yeah I've had mixed responses where it just felt like they were people randomly taking a test, um like my father does, as I mentioned, we're talking about that holly seems to have gotten really good feedback um becca. Also, if you could speak with her.

B

A

Might have really good feedback, um but the the hardest part of user testing is setting up the screener to validate those users of who you want to actually get and that's really the hardest part.

A

I don't feel confident in my screeners, even though I use the ones they recommended, and so I'm not sure why my my respondents feel like they're, not as uh accurate as or not as technically prolific or versus just oh. This is my job. I take screener tests all day long and whatever one I hit. I make some money on uh yeah yeah right. Consider that my dad does for fun he's like oh yeah. I did another survey today. I got paid 50 bucks.

A

Okay, I'm sure you really helped.

B

Somebody interesting that's good to know yeah I'll, let you know if, if your dad pops into my test, yeah, okay, that's good! So is the screener also part of that uh user testing, or is it a screen or a separate like application or separate service? No.

A

No, when you set up user testing, you go through and define a screener. So um let me see if I can do one real fast oops.

B

I just started looking into that last week, a little bit I just figured I'd, give it a try. I was always doing like more traditional kind of solution, validation that took a look longer yeah and we used the qualtric screener. So I haven't tried this yet.

A

Where is there, it is login.

B

Yeah, I was also thinking about signing up as like a tester.

A

Just to see yeah that actually might be a good idea, so we can see what they're seeing and help us.

B

A

The testing process right.

B

A

uh So, let's see what my computer's also slow, actually my internet customer.

B

A

I don't know what this is go away, uh so this is one of the examples that I had done previously and let's.

A

A

Now that's a test: let's go.

A

A

Oh, it's the same thing. I had this problem before where it was hard to set up the oh here. It is edit screener right below it.

B

A

So you have to put this one, which is like a just a license: agreement: okay and then you have like this one, which is job title and you can accept.

B

A

Different roles got it okay, and then you can do this one, which are what tasks? Are you responsible for your at your place of work? So like sales design, network admins is admin, and so this one was a strong like everything's rejected, except one right um and then the next one is which tools you currently use selected, buckets circle, ci code, ship, etcetera, etcetera,.

B

A

And so I was only looking for specific users like github, gitlab and bitbucket, for example, um and so this is the screener but, like I said, I'm not sure I don't really feel confident in some of the respondents how they gave answers.

A

um So I was like I said I used uh see if you go back.

B

Like there's not a way to say, I guess you know, are you a gitlab user on, like the ultimate plan.

A

There's I mean there's no way to validate it right, so yeah.

B

Yeah sure you can just say yeah right.

A

Got it and that's the problem with it right, there's no way to validate any of this, so you just.

B

Take the assumption that.

A

They're, accurate yeah, and so you could try and validate it by saying.

A

Do you, which, which is a specific feature that you use then try and go through and label, specifically things and then very specifically only have the one you want to get feedback on.

B

Oh yeah yeah kind of like test them to see if they can.

A

They know the answer if they know what they're talking about and put on a bunch of other ones that are very similar, yeah yeah, good idea, uh let's see so this is a screen here, there's a pre! Oh it's the examples, ah that's not what I want. If you saved here it is so they have a bunch that github or excuse me, gitlab uh has saved, for we have saved for other people. So like the thing is like they have this one, which only has two answers, which doesn't make any sense.

A

um This one's a little more logical, has very specific stuff.

B

A

You can actually go through and see if any of these other ones match, but yeah. That was a deal for me. Like I said it's, um it's really. The screener is the hardest part of using user testing and making sure that users are valid. uh The testing part is easy. Yeah.

B

Got it that's good to know, keep in mind, I like your idea of the little test there kind of further validate whether they're uh real get lab users. What's that I like your idea of putting throwing in like a little bit of an extra test to that screener to validate whether they're real uh real good lab users.

A

Yeah, I think that's really the only way that you're gonna.

B

A

Keyboard yeah, one of my side windows that I had saved to watch tonight has I decided to sit on the keyboard and play that instead, oh she's gotta get out of here.

A

She's been very snuggly today, which means getting in my face on the keyboard.

A

uh So yeah, like I said that was the hardest part, was setting it up and then doing like a a custom screener test uh with, like a very specific question of like yeah. This is exactly what I'm looking for uh got it, but again you have no way to validate it right.

B

Yeah for sure, thanks for sharing that yeah that's helpful, should we jump into austin's.

A

B

Working out at nvc for protected branch configuration groups turn on whether I should see whether I should use the existing table, design for protected branch or for the form like when you are creating one good.

B

Yes, we can maybe do this. What do you think should we do this async or dive into it, though? Probably.

A

That one async at least while you're. Here we can talk about what you're working on.

B

Yeah yeah sounds good. Yes, I'm actually working on uh exploring and documenting workflows and trying to identify some, uh the most common kind of user stories for our slack integration and slack application right now, just a little bit of background, our slack application is more than like two years old and it doesn't see that much usage and uh we have a slack integration, but it's it kind of it works through the slack web hook.

B

uh So you're still using like the slack uh web hook app to kind of create uh these kind of interactions with slack. So it's not really that I wouldn't say it's like a fully uh a full integration per se with uh with slack you know, so you can't really. You still can't do a lot of things with it.

B

So the goal is to kind of come up with the set of most common user stories and then just refactor, the existing slack application and then uh kind of, and then I'm so I'm also working on kind of figuring out. You know: what's the balance, what should be? What should we keep kind of in the slack integration? What should we update in the slack? I'm sorry slack app versus slack integration kind of what are the differences there?

B

How are they being used differently, so I've been starting started to take a deep dive in that like a couple weeks ago, I just went through like a bunch of old issues and sorting through that I went through a chorus ai kind of looking through some customer calls and then now, I'm at the stage where I'm just putting together.

B

uh So I have a bunch of kind of user stories already listed out uh right now, I'm putting together this short survey that I linked here just in a google form survey to try to kind of solicit more feedback just internally, so I'm starting small, I just shared it so far with our two front, end engineers and marcel my uh my manager uh just to start collecting feedback.

B

So I tweaked the survey a little bit and then now, instead of sharing it with uh with this team or everyone basically on this call again to kind of get, take the survey and then get just feedback on this survey and then I'll what I'll do is uh hopefully, tomorrow I'll share it out in uh you know, ux weekly and then I was going to share it around uh throughout, like our slack through different channels and then maybe through some different pms individuals.

B

That, I think, would be that I'd want to get feedback from on this or just their actual. You know, input on that survey.

B

uh That makes sense or any questions on on the line of work or.

A

I'm looking at it right now and so I'm just going to go through the process. Yeah.

B

Yeah, let's do.

A

It imagine it's like a superpower, you could ask it anything about what is happening in get lab. What would you ask it.

A

I don't know, I think it matters.

B

A little worried that it might be that that question might be like a little too open-ended, so I don't know, but I did want to kind of general and kind of just what's your wish list, you know what you know, so it could be anything. It doesn't have to be real in some ways. I.

A

Would probably group or rephrase it by saying what are your top three like uses for slack and then what are your top three wants, or the top three things that slack does not do.

A

That might be better or more specific, because, like asking this first question, imagine slack has superpowers you can ask anything about what is happening in git lab.

A

So that's also kind of interesting because you're putting it on the context of gitlab operations day to day usage, features, support um who knows right and yeah you'd have to get like a lot of respondents to get like to see any sort of standard variation of it or to see any sort of numbers rise up.

A

So I would probably I don't know if you had a chance to talk with the ux research team, but I would definitely go through with them about the questions or the the specific phrasing of some of the questions that you have in terms of what you're looking to to find out.

A

um Let's see I'll just write something real quick see what the next one is.

B

Yeah I mean for that question, I'm not necessarily looking for to start generating a pattern on that more just like it's more of a just degenerative kind of uh you know, exploration of you know, what's what what does everyone? What does everyone think or come up with you know? Is there anything unique out there or any kind of edge cases to consider.

A

Yeah I mean first like the first thing I think about. Like would be our open slack channels. um I would assume asking for support or asking you know. Why is this thing not this way or something just on my first because, like that's mostly what I use it for is trying to fix my gdk. You know how do I, how do I get this thing fixed because I always break it somehow.

A

uh So, if I look at the next questions, there's a series of questions, a new issue has been created and you are assigned to it. What would you do next?

A

The pipeline has failed for a job in an mri you were assigned you next. You were selected on your role and time zone to potentially review an mr. How would you act on this information.

A

Someone responded in a comment and you can see their comment in slack. That might would kind of be interesting. Number four, an issue you created 30 days is still open and not scheduled, so I wouldn't know how I'd be able to find that information or how that information would come to me right.

A

So I yeah I mean like answering these questions on the top of my head. Some of them, I'm not sure it's because, like um number two apply fine, failed and for a job in an mri. You are a sign um I don't ever get assigned. Mrs for pipeline review. Pipeline failures. Just doesn't look good, so I probably wouldn't know that or in a educated manner.

A

Yeah, um but otherwise, like these seem fine.

A

Also, I don't think you have a the questions is set to mandatory, so they can just skip next next next, without answering anything yeah.

B

Yeah everywhere, that's fine! Okay! I didn't want to like force someone down like to answer, for example, that question: maybe if they don't get pipelines right or they don't they don't deal with. Mrs really that much.

A

You could also put like um I'm not familiar with this as, like.

B

uh Yeah, that might be better yeah because I didn't want to like spend too much time on putting in logic like whether they chose like their role. You know, depending on their role, maybe have specific questions for yeah well, but yeah that'd be just kind of too much. I like that suggestion, actually just saying like I'm not familiar with this, you know check next.

A

Kind of thing yeah and just move on that'll, be my only thing um I'll just fill out some something real, quick. um Well, my question is so imagine you get a notification, so that was actually hard. I didn't notice. Maybe I did the the intro question. Imagine that you get a notification from slack telling you oh yeah 135.

A

You might want to break that up or make it more prominent, because I missed that on the first.

B

Okay got it that's good to know and maybe break that up in an individual, maybe pages.

A

uh I think I mentioned that before I think who was it. Amanda had a question like this. She was trying to do something with um using google form feedback to see, and that was the same thing. So people had missed that top intro question got it.

B

Got it yeah? I should just repeat that just make five separate pages, I think for that it's kind of easier, probably.

A

Well, the other thing too, is that what are you? Are you trying to do a specific thing with slack integration? Are you looking for all of these features to be as a part of the stock integration.

B

No, no, so these are just kind of the the five that I've I've kind of identified that you know seemed like the most common, so I just wanted to get feedback start getting feedback on some of these common user stories here that are kind of so I wanted to kind of validate them in some way but yeah, but I'm also still it's still at the phase, where I'm very much of the learning phase, not that much validating.

B

So I think after this survey I'll have enough information to kind of uh really uh prioritize these a little bit better and then uh and then validate those again and that's what I was hoping to do a little bit more with like either either user testing session or some kind of uh more externally. I guess with other lab users all right.

A

Was there uh another page after this, the the fifth page or is it just a thank you because I was filling up these questions.

B

There was just one I think, the bonus one I don't know, did you see that little bonus.

A

Section um I'll get to it, one second uh sure.

B

That was the last page. There's just like a little bonus asking uh if you want to, is it? Is there anything else you want to tell me and then, if you want to remain like uh kind of in the loop or stay uh stay in the loop, you can add your email address there to participate like for future.

A

Studies, uh the last one number five, an issue you created 30 days ago- is open still and is not scheduled.

A

uh I want to know if the issue who was assigned to because it was not assigned to me, do I care um that's something I want to know.

A

So I think context around how you what do you use issue scheduling for.

B

A

I've just been taking notes here on paper yeah.

B

A

I see on the other.

A

A

Oh anything else yeah, and would you like to be a participant yeah? Those are my only feedbacks, um otherwise it seems fine.

B

Yeah great, that's yeah, that's great feedback, thanks yeah, I'm gonna. Also it was a good call and asking uh reaching out to uh someone a researcher just to get their sense out, like the way that I phrase. Those questions, I think, is a good idea.

A

You could also, um I don't know how.

A

Like I swear I've seen like twitter surveys or something like if you wanted to try and reach out on twitter and do like a impromptu survey that way, um because then you get direct feedback from people in the industry more so I think than user testing. They could tag you on a hashtag, gitlab or, like you know, atkit lab or use it as that way. I think it's something that I personally don't think I've seen many people doing like utilizing.

A

The gitlab twitter account to.

B

A

Feedback on these things, right, which I think would definitely get to you- the target audience of people who know what they're talking about, rather than using the screener surveys right on user testing, we're having to do like set up on a quite a complex form on google uh surveys right.

A

um But that would be all the other proposal right.

B

No, I like that idea too. I was thinking about that at some point, so this is kind of more internal. I was thinking about reaching out to like twitter or reddit to maybe someone that has like more yeah, like they're, more industry, specific kind of participants out there. So I like that idea.

A

Yeah, that was just it for me um just going through the survey and looking at the questions, I couldn't really think of anything else.

B

Oh, this is a great great feedback. Thanks for that dude any problem uh yeah, that's all I got. I guess we're seven minutes over two. It's all good! Okay! Sorry about that. No.

A

Worries I enjoy helping cool.

B

Yeah we'll enjoy the rest of your night over there and uh yeah. We'll talk to you then take it. Easy have a good.

A

One cheers.