GitLab Pair Programming - Frontend, 21 Jul 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Quick debugging session - What the diff!

Description

In this session we show that changes in an earlier MR fix local webpack issues in the capybara test environment. But! For some reason these changes were accidentally reverted by a completely unrelated MR :|

MR: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/93043

A

All righty, okay! So, let's see if the test works.

B

I expect it to fail now.

A

Do you have like the live debug there was it showing the failure to load assets on the terminal.

B

Yep it will, um we will see this. Do you see this in a second? It has. um Let me see.

B

The binding the price still in place it is in place um uh after the whole party ended kind of. So um it's not interfering with anything.

A

Sweet all right.

A

So it's it's running and loading right now.

B

Yeah we have to be a little careful, because the bummer is that this will open up a new chrome instance every time and I miss out on closing the old ones, but I'm expecting to I'm not closing one now, because I don't want to close the, um but we will. We will see one popping up and then triggering a leader reload in a sec. Here we are just here we go.

B

This is what is interesting. Data.

A

All right we're loading stuff.

A

Okay failed to resolve async component. Oh man, oh no! Look at all, but.

B

This is, of course, it is still funny enough um for what we're trying to test still kind of unrelated um like it is not great that all of those async inputs um are not working out, but the confirm model is for the feature that we're trying to test here. This is the crucial one.

A

Right but like even the top nav, we hit the menu uh yeah yeah.

B

Nothing nothing that comes.

A

These modules are getting loaded, okay, absolutely, and that is what we fixed a while ago. uh So can you open up that? Mr.

B

A

I think it's that one.

B

That's one correctly.

A

Correct, yes, so I think we can actually just copy. I think we can just apply this diff. If you copy the uh I bet we can just copy the url.

B

um Yeah we could, um I just have it in my in my stash. Instead, if that's okay for you, yeah.

A

However, you want to buy those changes.

B

So um these are the exact same changes that we've just been looking at.

A

Yes, yeah, I see you commented out that if the, uh if else block yep awesome did you did you apply it cool.

B

A

You just did get done.

B

A

B

Yeah absolutely so, let's run it again and let me quickly close all of this garbage um sweet so yeah this will. I expect this to work now because that's the that's the actual issue and while we added paul, I understood the that this was an issue before. But can you briefly explain what on earth is going on here, because I don't fully I fully.

A

B

A

What what we discovered, if you know your name's actually tied to this original lamar too, which is funny.

A

Okay, what we discovered was- um and I'm really curious to see what this network tab shows up, but for some reason we had code that, under a test environment, pointed webpack to a different host and it just didn't really make sense why we needed to and those async assets weren't living on that host. So we just undid that clause which didn't seem like it was needed, um and so now everything loads yeah. I see so that's it so yeah. Now now everything loads all right thanks.

A

What are we gonna say: yannick.

B

um I was just curious about some further information about this was reverted and we rewrote it. And since I got no idea, okay, yeah.

A

No idea why, when we looked at the commit history, it looks like it even might have been an accident, but I'm really concerned in why I want to see where that I got to follow that commit trail. I'm a little concerned. Why that diff wasn't showing what I thought so anyways.

B

Thank you thanks again, thanks.

A