Hello,
First thanks for this great comparison of Parsers! I don't read russian so I read it through translation tools (I hope I didn't misunderstand things)
As a JMeter commiter I wanted to clarify some notes from the release notes of 3.0.
What has been improved in 3.0 is:
the connection simulation
the throughput of resources downloads
The parsing of CSS resources which didn't exist before
We never pretended that we downloaded what a browser does. We always write "JMeter is not a browser".
We don't download any JS loaded resources and will never do because we're not a browser.
Besides, from a Load Testing perspective, all resources that hit 3rd party servers (yandex, google analytics, ....) are not useful, we only download resources that match a regular expression that you enter.
Now your results are very interesting and I highly encourage you to report:
To Jodd (http://jodd.org/) a bug on the difference between downloaded resources compared to JSOUP
To report to JMeter the recursivity issue that you faced, with an example
Of course any patch improving JMeter is very welcome.
Hello,
First thanks for this great comparison of Parsers! I don't read russian so I read it through translation tools (I hope I didn't misunderstand things)
As a JMeter commiter I wanted to clarify some notes from the release notes of 3.0.
What has been improved in 3.0 is:
We never pretended that we downloaded what a browser does.
We always write "JMeter is not a browser".
We don't download any JS loaded resources and will never do because we're not a browser.
Besides, from a Load Testing perspective, all resources that hit 3rd party servers (yandex, google analytics, ....) are not useful, we only download resources that match a regular expression that you enter.
Now your results are very interesting and I highly encourage you to report:
Of course any patch improving JMeter is very welcome.
Regards
Philippe M.
philmdot