stephenbrooks.orgForumOtherBrowser WarsStephen still here?
Username: Password:
Search site:
Subscribe to thread via RSS
waldi
2019-02-04 13:15:33
Hi Stephen, are you still around? 
It's really cool that this site is still up, but it looks as if it's breaking more and more on the edges (no pun intended).  I've no idea if more people would visit and play if the bugs were ironed out (ouch).  Of course half the current browsers are detected as mozilla and the other half as chrome, so it's not too attractive for "browser warriors" any more.  Otter than that (ok that's it now...) I hope you're fine and thanks for a looong time with this great game. 
Zerberus
2019-02-06 04:05:54
It's too quiet.  Most links still work, but Stephen seems to be gone.

The place itself has lost a bit of its chromatic shine, so I visit it once in a pale moon.

Maybe he'll fix it some day.  The Firefox icon is broken since ages.
Stephen Brooks
2019-04-26 01:15:07
Well, I'm still here.  To be honest, I'd assumed that once I shut down the Muon1 project indefinitely, nobody would hang around my site any more.  It's kind of a pleasant surprise to see some people still here!

Maybe I'll have some time to update the BrowserWars icons this weekend.  A few months back they also migrated the whole site to a newer version of PHP, which broke certain things, but they are usually easily sorted out.  I got rid of the most obvious errors but I still see some on this forum for example.

BTW, I post a lot on Twitter. I retweet pictures of ducks mainly.
Stephen Brooks
2019-04-26 01:15:52
Also if you're on Windows you might want to look at dMaze. It's just a test right now but it's possible I'll extend it in the future.
Zerberus
2019-04-28 04:01:08
I still run Muon1 occasionally, for entertainment.  Though, my version 4.47dev is probably way outdated now.  Pity we didn't get to version 5, with the really cool stuff.

Another thing, the site has IPv6 now.  This will probably fudge with the IP detection stuff.  Makes no sense to detect the whole IPv6 address, the first half (64bit prefix) is enough.  ISPs assign whole IPv6 prefixes of 64bit or lower, so the user can choose at least the other half of the address freely.
Stephen Brooks
2019-04-29 06:33:14
Muon1 also has a Twitter account
Zerberus
2019-04-30 02:45:51
Yes, that's basically the really cool stuff I talked about.  Seems it evolved nicely.  That would have been fun simulating.  :\

As you are in the progress of changing the site over to TLS, please don't forget the feed.  It's still HTTP.
Stephen Brooks
2019-04-30 03:02:42
It's a compromise at the moment: the forum is HTTPS.  Muon1 has to stay HTTP because that's all the client can communicate with.

The RSS feed?  It seems fine via HTTPS, although I might not have given it the right "content type" in the server config.  What problem are you having with it?
Zerberus
2019-04-30 04:21:21
When I subscribe to the feed (forum, all topics), the entries still lead to the HTTP URLs.  It tries to redirect, but my browser is set so I have to confirm every redirection.

Well, yes, muon1 (our old clients, at least) does not support TLS.  Not over HTTP, and the Windows ftp.exe does, neither (it never evolved in any way).  That also prevents using stuff like HSTS on the site.

Do people still send in results?  Does it still make sense?  From what I've seen on muon1 Twitter, we are way behind, now.
[Edited by Zerberus at 2019-04-30 04:22:06]
Stephen Brooks
2019-04-30 04:57:29
I fixed the URLs in the RSS being HTTP just now, when I also found some non-PHP7-compliant stuff in there, so it should work better once your cache is clear.
Zerberus
2019-04-30 07:12:18
Indeed.
Stephen Brooks
2019-05-04 00:02:33
--[Another thing, the site has IPv6 now.  This will probably fudge with the IP detection stuff.  Makes no sense to detect the whole IPv6 address, the first half (64bit prefix) is enough.  ISPs assign whole IPv6 prefixes of 64bit or lower, so the user can choose at least the other half of the address freely.]--

I noticed Board 7 had Chrome lines of 4 but didn't register a win.  It appeared to log a lot of Chrome v6 IPs with the first 64 bits the same and the lower 64 bits different.  I don't remember how I did IP detection and whether this causes a problem, though.  I thought a win for Chrome should be a win even if "different" people did it...
Stephen Brooks
2019-05-04 02:01:41
Wonder if it could be something stupid like "a Chrome has won, but not your Chrome", so the board doesn't reset... Does IPv6 generate random strings for the lower 64 bits?
Zerberus
2019-05-04 23:08:08
What I meant by "IP detection stuff" is the anti-cheat where one user should not be able to play as multiple identities even if switching user agents.  If a Chrome wins, it should not matter whose Chrome it was.  It's "Browser wars", not "User wars".

IPv6 is a bit more complicated.  A device supporting IPv6 can get its IPv6 addresses (yes, having multiple is common, now) in several ways - e. g. by DHCPv6 or router assignment.  In any case, the ISP usually assigns at least a whole 64bit prefix.

Or, more specifically, an IPv6 address has 128bits (8 blocks of 16bits each).  The first 64bits (or first 4 blocks) are the interface part, assigned by the ISP or router.  ISPs can decide to assign less static bits to the users, meaning bigger regions of address space.  Example:

Getting a /64 prefix assigned means you can run only one network segment, like 2001:bf8:0:1::/64.
I can get up to a /56 prefix from my ISP, allowing me to run 256 subnet segments freely (2001:bf8:0:0::/64 - 2001:bf8:0:ff::/64).


The other 64bits are essentially free to choose.  The device can even assign it by itself (stateless autoconfiguration).  Originally, that part of the IPv6 was derived from the MAC address of the device, however, nobody likes exposing the hwaddress of the devices to the net.  Currently, IPv6 with privacy extensions uses random identifiers and temporary addresses to mitigate the issue.

What you, as website owner, will see from us IPv6 users is a temporarily generated IPv6 address, with the ISP-assigned prefix as the first half and a randomly selected second half (which usually changes at shutdown/reboot and every 24 hours, by default).

That's why I wrote to ignore the second half and just flag by the first, ISP-assigned one.  That one does not change too often.

tl;dr: Yes, the lower 64bits are randomly generated if "privacy extensions" are enabled.  And they almost always are.
Stephen Brooks
2019-05-14 03:39:04
If *I* go to a board (with Edge in this case) that already has rows of 4 and 5 of Chrome and Mozilla on it, I win as soon as I put down a single square!
Zerberus
2019-05-16 19:31:06
That's probably another issue and has to do with the system accepting negative board numbers.

As proof, take any board that hasn't completed, forge the URL so the board number is the same, but negative, like changing 8 for -8.

Surprise, there is an underground game scene where everything is upside-down!  Make a move there and you will likely f*ck up the normal board, too.
[Edited by Zerberus at 2019-05-16 19:38:02]
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel charactermstdn.io RSS feed

Site has had 26768236 accesses.