What is wrong in this test? :mad:

HTLM5 test, KM76proRCportable, out of the box: 419/555 points.
HTLM5 test, KM76proRCportable, w/ Firefox 55 UA: 419/555 points.
-
HTLM5 test, Chrome 60.0.3112.90, out of the box: 518/555 points.
HTLM5 test, Chrome 60.0.3112.90, w/ Firefox 55 UA: 517/555 points.

So, what is UA spoofing useful HTLM5 for? ¬_¬

https://html5test.com/

More interesting tests:
http://www.browserscope.org/alltests

Quote
J.G.
What is wrong in this test? :mad:

HTLM5 test, KM76proRCportable, out of the box: 419/555 points.
HTLM5 test, KM76proRCportable, w/ Firefox 55 UA: 419/555 points.
-
HTLM5 test, Chrome 60.0.3112.90, out of the box: 518/555 points.
HTLM5 test, Chrome 60.0.3112.90, w/ Firefox 55 UA: 517/555 points.

So, what is UA spoofing useful HTLM5 for? ¬_¬

https://html5test.com/

Interesting! I'm also spoofing Firefox 55, but in K-Meleon 75.0. That test comes out as 348 with HTML5 enabled. When I disable HTML5 and reload, I still get 348! :O

How accurate can that test be? By comparison, youtube.com/html5 shows the difference when I enable/disable HTML5.

Spoofing Chrome 60 in SRWare Iron 51 gives me a score of 486, but I can't disable HTML5 there to see what happens.

---
Gerry



Edited 1 time(s). Last edit at 08/14/2017 11:27PM by foliator.

@foliator, the HTLM5 test is very accurate, but I'm really very surprised about the results. :O

Also, I noticed a weird result with KM76RCpro /w Chrome 60 UA: 416/555 points only. Mostly important, I noticed that HTLM5 test detect the fake UA when Chrome is spoofing UA strings. When using Kmeleon, it always appears as "Kmeleon" (for example, while spoofing default Firefox 52 string UA by Rodocop, the test is unable to get a fake UA string as "unknown browser imitation"). While Chrome 60 is spoofing Firefox 55 it gives me this message: "You are using an unknown browser that imitates Firefox 55.0 on Windows 10" :mad:

Anyway, I am still asking myself about the advantages of using very recent UA strings, because the HTLM5 test results for KMeleon 76 are the same from FF 38 to FF 57 (Nightly Build). ¬_¬

Quote
J.G.
@foliator, the HTLM5 test is very accurate, but I'm really very surprised about the results. :O

Anyway, I am still asking myself about the advantages of using very recent UA strings, because the HTLM5 test results for KMeleon 76 are the same from FF 38 to FF 57 (Nightly Build). ¬_¬

The only reason I spoof recent UA strings is to prevent those irritating nags from various websites to update my browser. Those sites work, anyway. I haven't run across any sites that refuse to work when the original UA string is in use.

---
Gerry

@foliator, I wonder which is the minimum FF UA string version to avoid those warnings. However, I tested a very high FF version like 95 and it works perfectly in all the sites I visited, including a HTLM5 score of 517/555.

Mozilla/5.0 (Windows NT 10.0; rv:95.0) Gecko/20100101 Firefox/95.0

P.S.: The "Windows NT xx.x" partial string allows the Windows OS version.

* 419/555 score with FF 95, sorry for the typo.

I wonder what would happen if you gave it a Firefox version in the thousands, and perhaps an OS version of 13 or so.

I still can't understand why that html5 test shows exactly the same values with or without HTML5 enabled in KM 75. Unless, of course, it's looking beyond my current settings to determine what this browser is actually capable of.

---
Gerry



Edited 1 time(s). Last edit at 08/15/2017 03:12PM by foliator.

@foliator

Mozilla/5.0 (Windows NT 50.0; rv:1950.0) Gecko/20100101 Firefox/1950.0

419/555 HTLM5 test score, KM76RCpro w/ above UA.
517/555 HTLM5 test score, Chrome 60 w/ above UA.

All websites I have visited worked well for me, so I can't understand what's going on here. Probably the UA requirements are important only if they set lower versions, because machines are unable to understand when a "UA version level" is high or not, only low. :O

Quote
J.G.
@foliator

Mozilla/5.0 (Windows NT 50.0; rv:1950.0) Gecko/20100101 Firefox/1950.0

419/555 HTLM5 test score, KM76RCpro w/ above UA.
517/555 HTLM5 test score, Chrome 60 w/ above UA.

All websites I have visited worked well for me, so I can't understand what's going on here. Probably the UA requirements are important only if they set lower versions, because machines are unable to understand when a "UA version level" is high or not, only low. :O

Ha, ha! I shudder to think what Windows 50 would be like, and I doubt if the human species will be around long enough for the development of Firefox 1950.

---
Gerry

Quote
foliator
Ha, ha! I shudder to think what Windows 50 would be like, and I doubt if the human species will be around long enough for the development of Firefox 1950.

UAs are just a text string that the browser sends as his "name tag", but has no function in itself. It's up to website authors if they just use this (fake?) string to decide which elements and scripts to send across, or whether they actually do a check which functions exist on the client and which not.

And of course there are lots more page elements which may be influenced with it, not just HTML5 ;-) (for example those "obsolete brower" warnings)

PS: guess the current (fake) strings shows up in Help>About KM

PS2: if I recall correctly (80%?) and rodocop didn't change it lately, especially the UA macro script in KMPro is very stubborn: it stores an own UA for every site where it was toggled! That's embedded somewhere deeply (special omni.ja) or such, so changing it requires editing that file. And last time I looked, it's "default" UA for the other sites was also hardcoded in that special script.



Edited 3 time(s). Last edit at 08/16/2017 12:39AM by siria.

KM 1.6: 127/555 points (couldn't resist, LOL!) :cool:
"you are using an unknown browser that imitates"... my fake.

Meanwhile remembered: if your KM is recognized despite UA, it could also be those "vendor" prefs.

Quote
siria
KM 1.6: 127/555 points (couldn't resist, LOL!) :cool:
"you are using an unknown browser that imitates"... my fake.

Try it with the UA of a recent Chrome version and see what happens. Maybe htmltest.com will give you 500 points.

That site tells me "You are using Firefox 55.0 on Windows 7." That's what I have it set to in K-Meleon. I think Google doesn't notice, either, but if I use SRWare Iron 51, which is based on Chrome 51, and spoof it to look like Chrome 60, Google complains about an outdated browser, although version 60 is very recent, actually the latest version, I believe.

Also, they downgraded me by three points since this morning; now they gave me 345 instead of 348. LOL

---
Gerry



Edited 2 time(s). Last edit at 08/16/2017 01:50AM by foliator.

Before this thread becomes the longest and also most senseless one on these forums

As Siria already mentioned, the UA is only a text string sent to the server you are connecting with.
The UA has absolutely no impact on such tests! A text string can't influence the way a browser engine is working.
The points you'll get are depending on how the browser is dealing with the code it gets.

Yes, it's logical after all, being called TEST sites :cool:
That means they must not blindly believe the name tag (as most other web authors, also most video sites), but must actively test which features are really working or not.

Just in general and for future readers too, want to point out again a second possible mis-assumption:
Such tests do depend on other PREFS/SETTINGS!
For example if single features like javascript or media or mp4 etc. are toggled OFF, or some hidden prefs are OFF, which most users are unaware of, then the result will claim the tested browsers were "not capable" of the blocked stuff. Logical too, since blocked features shall not work obviously.

Prob is, in the Mozilla engines of FF+KM etc. there are are TONS of different prefs influencing video stuff. Mostly hidden ones, accessible only via "about:config", and a few not even showing up there! And some even influence each other. Several dozens, it's a real JUNGLE, Mozilla seems to want to make it as hard as any possible for users to choose stuff as simple as "I want html5 only" or "I want Flash only"
That does not mean I'd be an expert, far from ist! Only more aware of that jungle, and very cautious when any users blindly believe a "non-capable" feature in a test would automatically mean "not existant and not possible at all in my browser".

And a 3rd important reason to consider:
the user's system and hardware can be responsible too, not just the browser itself! For example those "webm" prefs have something to do with Intel hardware acceleration, and those windows-media-foundation prefs depend on the Windows OS version.

Quote
Yogi
Before this thread becomes the longest and also most senseless one on these forums As Siria already mentioned, the UA is only a text string sent to the server you are connecting with. The UA has absolutely no impact on such tests! A text string can't influence the way a browser engine is working. The points you'll get are depending on how the browser is dealing with the code it gets.

You are not right.

Chrome 60 with its default UA: 518/555 points.
Chrome 60 with "Chrome 60" UA (just these two words): 503/555 points.
Chrome 60 with "Chrome/60.0.3112.90" UA (just this word): 518/555 points.
Chrome 60 with "Firefox 50" UA (just these two words): 517/555 points.
Chrome 60 with "Mozilla/5.0 (Windows NT 10.0; rv:55.0) Gecko/20100101 Firefox/55.0" UA: 517/555 points.
Chrome 60 with "Firefox/50" UA (just this word): 517/555 points.
Chrome 60 with "Mozilla/5.0 (Windows NT 10.0; Trident/7.0; rv:11.0) like Gecko" UA: 518/555 points.
Chrome 60 with "Trident/7.0" (just this word only): 506/555 points.

Kmeleon 76rcpro with its default UA: 419/555 points.
Kmeleon 76rcpro with "Chrome 60" UA (just two words): 419/555 points.
Kmeleon 76rcpro with "Firefox 50" UA (just these two words): 419/555 points.
Kmeleon 76rcpro with "Siria" UA: 419/555 points.

So @Siria is completely right. :cool: :cool: :cool:

Quote
siria

Prob is, in the Mozilla engines of FF+KM etc. there are are TONS of different prefs influencing video stuff. Mostly hidden ones, accessible only via "about:config", and a few not even showing up there! And some even influence each other. Several dozens, it's a real JUNGLE, Mozilla seems to want to make it as hard as any possible for users to choose stuff as simple as "I want html5 only" or "I want Flash only"
That does not mean I'd be an expert, far from ist! Only more aware of that jungle, and very cautious when any users blindly believe a "non-capable" feature in a test would automatically mean "not existant and not possible at all in my browser".

Neither am I an expert myself nor am I keen to defend Mozilla.
However, some things are worth to be mentioned nevertheless.
- "about:config" is not meant to be used by the average user - at least not as long as he knows exactly what he is doing.
- the default settings in "about:config" are those which should work best for all or at least most users.
- the more settings "about:config" offers, the better the chances for the advanced user to tweak the browser to her/his like.
- why some default settings in "about:config" are changing over the time:
New implemented features are usually experimental and disabled in "about:config". Those testing the browser can enable them and report if they are causing problems. As soon as those new features have been tested enough, they will be enabled by default in "about:config".
Even so, thanks to Google's infernal release cycles, it's almost impossible to test the browser reliable. Not even Google does/can. As a result they fix dozens of (security) bugs during their updates. The average user will say: "wow, so many bugs fixed, Chrome must be the most secure browser that ever existed." Nobody will ask how and why those bugs have made their way into the browser in the first place and for how long they have been there before getting fixed. Mozilla is damned to keep up with those infernal release cycles, otherwise people will complain that the browser doesn't work on Google's sites and services like Chrome does. I'm not even speaking about the financial burden imposed by such infernal release cycles...



Edited 1 time(s). Last edit at 08/16/2017 02:04PM by Yogi.

Quote
Yogi
Even so, thanks to Google's infernal release cycles, it's almost impossible to test the browser reliable. Not even Google does/can.

Google has plenty developers and 60% market quote. Firefox doesn't.

However, we can't blame the others for their release cycles. They can release crap every single day, or just release good things each month. Also we can't blame the others because if we blame the others, the others would blame us for the same reason. I think we all would be so happy if Kmeleon 77 with Mozilla 52.3 ESR inside was released some day. Probably our release cycle would seem surprisingly slow for any Chrome or Firefox user.

And I don't think that this topic is "nonsense" and "longest". I'm a newbye to Kmeleon and we all deserve the opportunity to show our opinions and ways to contribute to make a better useful browser. I have demonstrated that Kmeleon scores are the same in the HTLM5 test whatever the UA spoofed, a proof of its robustness. Also it seems that the UA string in Chrome is very important (just notice the HTLM5 test while Chrome is spoofing Internet Explorer 11: it gets 518 points, just like the default Chrome UA string, 1 point more than FF 55 UA). :O

I think we should enjoy every single line wrote in this forum.

P.S.:
Ubuntu: new complete free ISO builds every six months.
Microsoft Windows 10: new complete free ISO builds every six months.