Open "Developer Tools: Netmonitor", load any wesite supporting gzip compressed transfers (HTTP 1.1 or HTTP/2 doesn't matter), my example: mapyourtrack.com Expected: Transferred column should display actual size of information going through the wire (compressed) Got: Transferred column displays incorrect values (for gzipped transfers), roughly equivalent to Size column. Wrong column value confirmed by wireshark (capturing actual packets), also it is inconsistent with Content-Length response header.

Btw it was working fine a few months before (always on up-to-date Nightly), around May 2016. Windows 10 x64 1607 (14393.222) Name Firefox Version 52.0a1 Build ID 20161011030212 Update Channel nightly User Agent Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:52.0) Gecko/20100101 Firefox/52.0 OS Windows_NT 10.0

It appears only in non-e10s mode. Reg range: https://hg.mozilla.org/integration/fx-team/pushloghtml?fromchange=8173084bb8ebc4695a130e47575c8ff024d3f896&tochange=76a3bb10b02aa23b469ea1cc1c90d4b4b3619433 Tom, could you check this regression after your patches in bug 1244227, please.

Taking.

Track 51+ as regression in Developer Tools.

It turns out that this patch: https://bugzilla.mozilla.org/attachment.cgi?id=8783541&action=diff ... was only ok for e10s. We still needed this code for non-e10s. Also, nothing was testing the transferred size.

Comment on attachment 8801809 [details] Bug 1309523 - restore nsIInterfaceRequestor to network monitor; https://reviewboard.mozilla.org/r/86452/#review85176 Thanks, seems like a straightforward revert back to the previous path. ::: devtools/shared/webconsole/network-monitor.js:432 (Diff revision 1) > > this.request = request; > this._getSecurityInfo(); > this._findOpenResponse(); > + // We need to track the offset for the onDataAvailable calls where > + // we pass the data from our pipe to the coverter. Nit: convertor

Pushed by ttromey@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/2c81db21bb49 restore nsIInterfaceRequestor to network monitor; r=jryans

Tracking 52+ as this is a regression tracked for 51 and fixed in 52.

Hi, could you help verify if this issue is fixed as expected on the latest Nightly build? Thanks!

Hello. Yep, sorry, I checked it a week ago, it is working fine since the fix landed. Just forgot to mention it. Thanks for your work, guys! Regards.

Hi :tromey, Since this bug is a regression and also affects 51, do you consider to uplift this for 51 if this patch is not too risky?

I think it's reasonable to uplift. This patch is largely just restoring code that was there previously; and that was removed based on an incorrect belief.

Comment on attachment 8801809 [details] Bug 1309523 - restore nsIInterfaceRequestor to network monitor; Approval Request Comment [Feature/regressing bug #]: This fixes a regression introduced by bug 1244227. [User impact if declined]: The network monitor will show incorrect "transferred" sizes for compressed responses when not using e10s. [Describe test coverage new/current, TreeHerder]: During investigation we noticed that nothing was testing for correct values of the "transferred" data; this patch adds a regression test. [Risks and why]: This is relatively low risk. The patch simply reintroduces some code that was removed (based on an incorrect belief) in bug 1244227. [String/UUID change made/needed]: N/A

Comment on attachment 8801809 [details] Bug 1309523 - restore nsIInterfaceRequestor to network monitor; Fix a regression related to devtools. Take it in 51 aurora.