The "Vivaldi" in the User-Agent string is sometimes abbreviated to "Viv"
I'm using Vivaldi 1.10.867.42 on Debian 9 (Stretch).
The last few months I've had a strange issue with Vivaldi that I haven't had with any other browser. The API that I'm building at work wouldn't let me log in with Vivaldi, but worked with all other browsers.
After a few hours of debugging today I figured out the reason. I have made my API very strict when it comes to user sessions. If anything about the user's environment changes, the session token is immediately invalidated.
One of those environment variables is the user agent. This is the user agent string of my current Vivaldi install:
Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.104 Safari/537.36 Vivaldi/1.91.867.42
For a few requests, seemingly at random, it reports this user agent string instead:
Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.104 Safari/537.36 Viv/1.91.867.42
This causes my API to get suspicious and delete the user session token.
Is this a bug, or is there a legitimate reason why this happens?
Are the special URLs you fetch at this moment (when teh shorted UA appears) with Vivaldi or do you use a webpanel?
I will ask in developer chat.
sites matching this regex
Is VB-9275 and VB-21248
That is a fix because these Google's pages do not detect Vivaldi correctly.
@Gwen-Dragon You are definitely onto something. I took a look at the requests that use the alternative user agent, and they are all for the file "google-maps.js". However, this file is hosted on my web server, not at Google. The URL is:
This URL doesn't match the pattern you describe, yet the altered user agent is sent.
@Hubro Can confirm on my test domain, too. Vivaldi fails as its regex does not check only domains but filenames, too
Reported as bug
VB-30636 "Wrong useragent sent to non-Google domains"
The bug with wring UA spoofing is already fixed in todays internal tester version.
So next Snapshot and upcoming Stable will have the fix.
@Gwen-Dragon Awesome! Out of curiosity, do you know if it was fixed due to this bug report? Or was it already fixed by the time I posted it?
@Hubro It was fixed on 2017-08-02 around 16:00 CEST, some hours after i reported the bug.