Browser testing sucks, it's time to start doing it right. | Hacker News Reader

Browser testing sucks, it's time to start doing it right.

saucelabs.com

19 pointsmikecotton16 comments

Internally here at Guidewire, we have a declarative widget-based web framework for our apps, and we test by automatically generating a Page Object model based on the declarative widget definitions (using metaprogramming techniques in Gosu). The rigidness of defining your application in terms of declarative widgets is not for most applications, but the resulting improvement in testability is probably the biggest benefit. Separating out a logical model of "widgets" on the page and writing against a Page Object model, whether or not your web framework has widgets, is clearly the right way forward to me. There's nothing worse than debugging hundreds of test failures because no DOM element with id "User_Search" was found on the page. Did the page not render? Did the test run too fast? Did someone change the ID? Did someone move the search button to a different page? Did they hide it behind some div I need to expand first? Did the page just break? There goes two days of development time . . .

Prior to that approach, we tried everything: we actually wrote our own test framework that used javascript to drive the browser from inside the browser, then we moved to using selenium, but they all suffered from that same problem of incredibly fragile tests. Over time, test maintenance is easily 10x the cost of initial test development, so it's a huge, huge problem.

So I see this as a promising direction, but building and maintaining that (essentially) Page Object model by hand is going to result in its own set of maintenance problems. How do you know when the model is up to date? If someone changes the model, how do you know which test that affects if you're in a dynamic language like Ruby? It solves one set of horrific problems, and replaces it with another set of problems that are hopefully, with enough developer discipline applied (and guidance from the framework), ultimately slightly less horrific. Being able to automatically generate the POM via metaprogramming by statically analyzing your application is awesome, but it's not a technique that really works with any standard web framework that I know of.

Web app testing really still has a long, long, long way to go.

There are some awesome strides about to made for generating/maintaining POM (or WOM, in our case) - sadly, marketing has shot down all of my suggested names :(

That said, it's going to be a lot of fun, and I'd love to have someone with your experience and insight join in the conversation. I'll have to think of a way to do it - a Test::Right convore, or mailing list, perhaps?

My personal e-mail is in my profile, so I'm happy to participate in whatever manner you deem is best. Just send me an e-mail and let me know (or if you're in the Bay Area, you can catch me in person at SuperHappyDevHouse on Saturday).

I think the goal of "impossible to write bad tests" is unreachable. It's like "idiot-proof": impossible because the idiots are too resourceful. I'll be really happy, though, if Test::Right accomplishes "easy to write good tests." Right now, nothing out there does.

I especially like the decoupling of intentions from mechanics. It lets us improve the mechanics without breaking or rewriting all the tests. Maybe the mechanics are implemented with JS DOM-traversal code, maybe it's CSS selectors, maybe it's SIKULI-like image recognition (http://sikuli.org/).

Looking forward to trying it.

> I think the goal of "impossible to write bad tests" is unreachable.

We all sat down a bit ago to reflect on 16 month of non-stop in-the-trenches browser testing for all kinds of people, to integrate it all. We talked about how we wanted to make things better for the people out there who know what they want to do, but are continually screwed over by the how.

The goal "impossible to write bad tests" floated out, and it was almost a joke. But we started to wonder it would take, and it grew more serious. It's a mantra now.

And we have so much cool stuff flowing out form that mantra in the next few weeks/months from that attitude, it's insane.

And excellent point about intentions/mechanics - Jason Huggins is all over Sikuli. We're watching very patiently :)

Well, I love big audacious goals. It's just that "impossible to write bad tests" could lead to a system where it's also impossible to write good tests. If you start shooting down good ideas because they could be abused to write bad tests, you'll end up with mediocrity. At the extreme, just prevent people from writing tests at all. Mission accomplished! Now, I know you guys are smarter than that, so I'll just leave it be.

Thanks for the work you've already done on this, and will no doubt do in the future.

I see a big problems in web testing caused trying to test drive from outside of the browser. I'm currently working on a solution that has tests written in javascript that run inside the browser. The test runner puts your web app into an iframe, so the test code can use jQuery, for example, to easily automate and assert over the UI.

JavaScript also makes creating a very clean testing-focused DSL easy.

I'm hoping to open source the project soon. If anyone would like to know more give me a shout (@andrewdavey).

This is the approach that Selenium 1 took, and there be many dragons down that path. It's workable (we've done a ton of cool stuff at Sauce Labs making it work), but it takes an usual mix of skills and insight to get it to really gel together. Selenium 2 was built with all that mind.

That said, it's not the important part. Selenium (and watir, windmill, etc.) is an execution tool, it doesn't help you avoid some very, very large pitfalls waiting for you down the road. Too often your tests are tightly coupled with the dom, data and selectors are strewn all about, you have manual pauses and timeouts set by someone who left years ago, and you have so many that the whole thing takes two weeks to run (yes, I've seen this a number of times).

Basically, I don't care about the tool to execute the tests anymore. They'll all become interchangeable with Selenium 2 anyway. The much more valuable piece here is in creating tests that are maintainable and a joy to work with.

Certainly best of luck, please be sure to share you experiences and insights! There's always more to be learned.

Reducing tight coupling is definitely a concern I'm looking into. I can't see there being any silver bullets, but a good framework should certainly encourage a sustainable way of working.

I've developed a rather nice asynchronous, sequential, execution model for test code. So they avoid the callback-hell you'd see with multi-step async tests. The spin assert idea mentioned in the article is very useful, and I'll be borrowing it for sure! :)

I'm wary of running tests in parallel, due to concurrency issues. However, I guess spinning up multiple web servers and in-memory databases would work fine.

I'm having a hard time pinning down the feeling, but there's a lot more declaration here rather than evidence.

In other words, shut up and show me. (This isn't an attack on saucelabs)

The github page is sparse on documentation, and the wiki page is basically a copy of the README.

So don't make vacuous claims. Write an article demonstrating how this system works, how it compares to X, and why it's better than X.

Code's there, example application with tests is there, go and try it.

We err'd on the side of getting this out (several developers were chomping at the bit even with just a bit of editing on the post) - rest assured that we'll have more coming. In particular, we'd like to have some more documentation, examples, and best would be quick screencasts talking about and demonstrating the techniques/benefits going on in here.

No worries about 'shut up and show me' - but the counter-challenge is 'dive in and try it' :)

tldr; "I hate browser testing. Did I fix the problems? No. I started writing opinionated software though, because my opinion should matter. I wrote it in Ruby because I know Ruby."

Hmm, that's not really a considered review.

Straight off, no on at Sauce Labs hates browser testing, that's just a stupid thing to posit.

Test::Right is a really cool mental model of testing, and it's going to make people's lives better. Robust tests, parallelism, immediate cross-browser compatibility (this is an issue people without a lot of experience don't really understand), a lot of cool shit.

I love the promise that if I write my tests in this pattern/framework, my tests will always be good. It's not there yet, but it's still damned cool.

We also have several volunteers who're excited to port this to other languages, but we started with ruby because it's a fertile playground, and the important piece here is the pattern, not the language/framework itself. It'll end up being the flagship browser testing framework de facto, but that's not the goal here.

I played a big part in putting some pieces of Test::Right together, full disclosure.

"ruby ... it's a fertile playground"

I like this a lot.

AFAICT, the whole point of Test::Right is precisely to fix the problems. As for opinion, well, if anybody is qualified to have an opinion about browser testing, it's the folks at Sauce.

Reply on news.ycombinator.com