I have coded several projects with Selenium, which worked perfectly for all my needs, then I needed a headless browser and coded just a small test using PhantomJS and it worked just fine. However, PhantomJS uses way too much memory for me to launch hundreds of instances, which is what I would need in my next project.
After reading and re-reading this question and many others: Headless Browser and scraping - solutions
and doing tons of online searches, I am still unsure which headless browser might suit my needs.
What I need:
- JavaScrtipt, Ajax, HTML 5 support
- proxy support
- low enough on memory and CPU usage so that I can have at least 100 instances running simultaneously
- running on Windows machines
What I hope for:
- C# .Net wrapper (not a deal breaker)
- no installation
- good documentation
- Webkit based
For now, ZombieJs and HTMLUnit are the two solutions that I want to test, but before I invest considerable amount of time in either, I hope someone has already done a similar project and can point me in the right direction.
