Change google.com to wikipedia.org

[aleksandrs-ledovskis]

While Google's front-page is undoubtedly the single most known site on the Internet, authoritative details of underlying back-end operations that result in HTML page brought to end-user's device are either unknown, scarce, outdated or non-verifiable. For the sake of "No[t] skipping out on anything", it is a must that we describe processes taking place both on client and server side as detailed as possible.

One can't argue that Wikipedia is also a very widely recognized website, however best thing about it in comparison with Google is that it is a non-profit organization and runs its operations on a known OSS stack. I would also expect that inquiring Wikimedia sysops/admins about some additional specifics would be more successful than trying a same feat with Google.

@alex What's your take on this?

[kojiromike]

:+1: I wasn't far from making the same suggestion myself. Should it be https://wikipedia.org/ though? Also, what happens to details like SPDY and HSTS? It'd be a shame to lose them altogether…

[aleksandrs-ledovskis]

@kojiromike Judging by this ticket SPDY HTTP/2.0 support is in development pipeline for Wikipedia. On HSTS, well, last time I checked google.com also doesn't set it. So, it makes no difference - either way we would desribe it from what if point of view.

[gitcnd]

Yes - we absolutely need to avoid using "google.com" - because (eg: see my DHCP bug) - this happens to be the default landing domain for search queries as well, which makes the distinction between "direct" and "search" really confusing to explain. (For example - if you use DHCP, and wake up your laptop and really quickly put "wikipedia.org" [enter] in there, you'll end up at: https://www.google.com/search?q=wikipedia.org )

I also like the ",org" ending - much better than .com for explaining stuff.

Wikipedia is not a good substitute though: it's not secure, and HSTS is really important.

[kojiromike]

What about example.org? The more I think about it, the more I think this is exactly what example.com and example.org are for. The only possibly compelling reason to use anything else would be to cover nonstandard protocol or configurations, as secondary examples.

[aleksandrs-ledovskis]

Well, the IANA test domains are pretty good candidates, and I at first considered them to replace google.com but their ops side is not much known [the main reason to change from Google, remember]. Running simple curl -v example.org also brings some questions:

• What is the webserver software? Server: ECS (ewr/1584 [or] phl/9D54) is pretty much unheard of
• How is caching designed? X-Cache: HIT/Etag: "359670651"
• Custom headers? x-ec-custom-error: 1

Also, there are no Cookies/HTTPSHSTS/Javascript/geo-DNS and list goes on...

[unwiredbrain]

How about:

• the Electronic Frontier Foundation (EFF)
• the OpenSSL project
• the Internet Engineering Task Force (IETF)
• the Linux Kernel Archives

[aleksandrs-ledovskis]

@unwiredbrain Any site would do (even google.com) if there's publicly disclosed information available regarding ops.

[ghost]

We could include both under two separate files.

[uschwar]

Absolutely! And once it's done, it should also be published on wikipedia.org. Actually, why not put it there straight away?

[StevenBlack]

@uschwar the Wikipedia deletionists won't allow this. Lack or references is one reason, and others like no original research.

What I'm saying is, when it comes to posting to Wikipedia, nothing is simple.

[kojiromike]

Nothing is simple, Except the above statement. Well, it seemed simple.

[ghost]

@StevenBlack I am very involved in Wikipedia and we could easily fix those issues but the largest issue would be content that is not wikified or is Unencyclopedic.

[PiRSquared17]

The current format is unencyclopedic. You could try to get it on Wikiversity though, if it qualifies as a "learning resource" (I think it does).

[TheInitializer]

Lol. When I type "g" into my URL bar github.com pops up...

I kinda agree with this.