4 ms·
I've thought of that too. The trouble is that Browsh's text renderer is more of a science than an art. It's quite hard to actually get a guarantee of the same t
by tombh 8y ago
I've thought of that too. The trouble is that Browsh's text renderer is more of a science than an art. It's quite hard to actually get a guarantee of the same text twice, which is a pretty important condition of testing.
But yes, I'd love for Browsh to be used in such a way!
- brlewis 8y agoGuaranteeing the same text twice might not be so important. There are pixel comparison testing tools for popular browsers. Your advantage is probably speed. Would cypress end-to-end tests run way faster in Browsh than in a graphical browser? If so, you've got something of value to big companies. You might want to partner with or contract for Cypress. EDIT: Since Browsh depends on Firefox I'm curious whether it's actually faster than a graphical browser.
- annabellish 8y agoAs an anecdote, somebody at my workplace has just spent a day tearing their hair out trying to figure out how to get our selenium tests running automatically on a headless machine, because the resolution flat refused to go above 1024x768. Even then, they're slow and heavyweight. A very fast browser not tied to a graphical display with the bonus of being able to put error screenshots into all of our existing tooling could actually be quite beneficial here, whether the text is character-identical every time or not.
- michaelmior 8y agoSince browsh requires Firefox so I'm not sure you'd really get much of a speed benefit over Selenium
- JadeNB 8y ago> more of a science than an art. Given the rest of your sentence, I think you mean the opposite.
- tombh 8y agooops, yeah