Richard Pearse Reporter
	Description • 9 years ago

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:41.0) Gecko/20100101 Firefox/41.0
Build ID: 20150929144111

Steps to reproduce:

go to  eagle-i.net (https://www.eagle-i.net/)
Select "Explore All" in the upper right side under the search bar (https://search.eagle-i.net/central/#results?of=score)
Now select "All Core Facilities" in the upper menu (https://search.eagle-i.net/central/#results?t=http://vivoweb.org/ontology/core%23CoreLaboratory&of=score)



Actual results:

FF 41.0.1 returns an "invalid search" to this query while other browsers and earlier versions of FF did not.
Swapping the "%23" code in the above url for "#" resolves this issue so it seems that FF 41.0.1 no longer reads the "%23" code. This impacts every portion of our search that queries with a term containing this type of code (i.e. %27=', %20={space}, and so on)


Expected results:

Should have interpreted the %23 code as a #

	Alice0775 White
	Comment 1 • 9 years ago

Triggered by: Bug 1093611


It works if UA spoof as follows
user_pref("general.useragent.override", "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/43.0.2357.65 Safari/537.36");

So this should be the site problem

	Hallvord R. M. Steen [:hallvors]
	Comment 2 • 9 years ago

Hi Richard, can you try to contact this site and send them a link to https://www.fxsitecompat.com/en-US/docs/2015/urlutils-hash-no-longer-decodes-fragment/ which tries to explain the problem? Their code should stop special-casing Firefox. We're changing to be more compatible with other browsers and with code that doesn't make special adaptations for Firefox. This will make things easier for developers in the long run but right now when we're changing there's going to be a bit of trouble with old scripts.

	Richard Pearse Reporter
	Comment 3 • 9 years ago

(In reply to Hallvord R. M. Steen [:hallvors] from comment #2)
> Hi Richard, can you try to contact this site and send them a link to
> https://www.fxsitecompat.com/en-US/docs/2015/urlutils-hash-no-longer-decodes-
> fragment/ which tries to explain the problem? Their code should stop
> special-casing Firefox. We're changing to be more compatible with other
> browsers and with code that doesn't make special adaptations for Firefox.
> This will make things easier for developers in the long run but right now
> when we're changing there's going to be a bit of trouble with old scripts.

Thank you Hallvord. I'll have our developers look into this.

	Richard Pearse Reporter
	Comment 4 • 9 years ago

(In reply to Alice0775 White from comment #1)
> Triggered by: Bug 1093611
> 
> 
> It works if UA spoof as follows
> user_pref("general.useragent.override", "Mozilla/5.0 (Windows NT 6.1; Win64;
> x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/43.0.2357.65
> Safari/537.36");
> 
> So this should be the site problem

THank you Alice, This is helpful.

	Adam Stevenson [:adamopenweb]
	Comment 5 • 8 years ago

Looks like this has been fixed. Following the STR I don't get any errors in Firefox 49.
Closing.