I think it has to do with the FireFox “heartbeat” effort to get feedback from users, but even if that’s true I don’t appreciate that my website’s responses are being affected by code I didn’t enable.
This isn’t merely a problem with my browser; this is a problem with the results that webpagetest.org gets, and we can assume others using Firefox will get the same results: my site’s responses weighed down and slowed down by the self-repair code.
Can anyone tell me how to get rid of this stuff so my test results are clean again?
This will only affect your own browser, but you can disable it by
(1) In a new tab, type or paste about:config in the address bar and press Enter. Click the button promising to be careful.
(2) In the search box above the list, type or paste self and pause while the list is filtered
(3) Double-click the browser.selfsupport.url preference and delete the address (leave it blank) and click OK
If you decide you want to participate in Heartbeat again, right-click the browser.selfsupport.url preference and choose Reset to restore the URL.
The credit for this goes to angry.user on the Mozilla forums, I take no credit for it.
Thanks for responding, but I have done all you say I should do.
Here’s the problem: my browser has nothing to do with the code appearing in the result that webpagetest generates, evaluates and displays. In the first place the heartbeat system was already disabled from my browser. In the second and most important place: the testing visit is made by the webpagetest browser, not mine.
I have just performed a test using a computer that has never had the heartbeat code on board. I ran four tests using webpagetest.org against my website: richardgalli.com
When I ran tests from Dulles Virginia, the heartbeat code did not appear in my results when webpagetest used either a Chrome or a Firefox browser.
When I ran tests from Denver Colorado,
– the heartbeat code did not appear in my results when webpagetest used Chrome, but
– the heartbeat code DID appear in my results when webpagetest used FireFox.
In my mind, there is only one conclusion: the FireFox in Denver has the heartbeat code on board and that code is not on board in Chrome OR in the FireFox instance in Dulles.
Then I ran more tests from the webpagetest.org engine in Denver Colorado, and the Mozilla self-repair code showed up in the results for ALL of these sites:
My computer that has never had the heartbeat script on board is not the reason why Mozilla’s self-repair code is getting included in page-test results from my site, Google, Amazon and Apple.
I’d be happy to hear your thoughts; and I urge you to test yourself from the webpagetst site in Denver.
[quote=“gallilaw, post:3, topic:9612”]
I’d be happy to hear your thoughts; and I urge you to test yourself from the webpagetst site in Denver.[/quote]
It’s true that there are several issues with testing SSL sites with FireFox. I use Chrome tests only now for a while because of these issues. However, it’s not only the Denver instance that has these issues:
California - Firefox: incomplete waterfall, doesn’t show HTML and images on page, only clients.google.com and analytics: webtest.
Miami - Firefox: the self-repear issue: webtest.
Miami - Firefox: the incomplete waterfall issue: webtest (I did got a loadtime of 74 ms though )
Dulles - Firefox: the incomplete waterfall issue: webtest.
For reference, here’s the webpagetest from Dulles with Chrome that shows the same page as the above errorneus FireFox tests: webtest. As you can see in that test, there’s no issue with Chrome. Also, if I visit the tested webpage with FireFox installed on my pc, there’s also no issue. So I think there’s something off with the Firefox WPT instances.
So it’s certainly not just you Gallilaw that experiences these issues.
)