Raborazzi shipped iOs snapshot test. I asked AI to...
# testballoon
e
Raborazzi shipped iOs snapshot test. I asked AI to run spike about it. I’m digging to details. Looks like there is main thread lock with Test Balloon. Not human confirmed yet.
👀 1
o
Anything new (and real, human-confirmed) about that one?
e
Sorry, not yet
I parked iOS snapshots but forgot to dig in a bit
o
No problem. Can you refer me to the Roborazzi announcement about iOS support?
e
It was time ago. Here is already release with improvements github.com/takahirom/roborazzi/…/1.66.0
o
Thanks! I'll be looking into it.
Oh, and BTW, I did a Roborazzi comparison between TestBalloon and plain JUnit 4 based on your article. I've tried to add helper code for both variants in order to understand the minimum amount of boilerplate needed in each case: • ScreenshotTestsViaTestBalloon.kt • ScreenshotTestsViaPlainJUnit4.kt With larger test volumes, I also found out that TestBalloon can run Roborazzi tests up to three times as fast as plain JUnit 4. Larger test volumes also require a memory optimization due to ship with TestBalloon 1.2.0.
e
Nice!
I have only one project with Raborazzi where I started from start with Test Balloon
So I choose the best strategy already 🙂
o
Yep. The only thing where you could have relaxed things a little bit is passing parameters directly instead of using string lookup.
e
Yep. The only thing where you could have relaxed things a little bit is passing parameters directly instead of using string lookup.
Thanks! Let me change it. Still I have matter of seconds for the whole suit.
o
Found it
Thanks! So that was quite a while ago. I should probably have scrolled down the README, where it's mentioned.
Still I have matter of seconds for the whole suit.
Yeah, it's not because of the performance, just better readability if you can use the parameter classes directly.
👍🏼 1