Profile cover photo
Profile photo
Adam Bliss
975 followers -
The simplest person that cannot be described in ten words.
The simplest person that cannot be described in ten words.

975 followers
About
Adam's posts

Wait, how does this work?
http://my.router
I couldn't find ".router" in any list of IANA TLDs, and I couldn't find any RFCs about this shortcut. But it's a tough thing to google for. 

Post has attachment
RIP. But the link should have been: "What is the headline of this obituary?"

Post has attachment
I hope Mark O'Connell writes a novel like this article.

In older versions of android there was a file at /data/data/com.android.providers.settings/databases/settings.db that you could access directly with sqlite3 (if you had root). In Marshmallow this seems to be gone. Does anyone know where it went?

(I need to access it because the settings app finally removed the QuickLaunchSettings activity, so there doesn't seem to be any user-level way to edit bookmarks anymore.)

I think my 3yo kid just hacked Number Genie on the Google Home. He got it to respond:

"Well done! It is indeed %x."

Nooooo! http://m.xkcd.com seems to have added HSTS. That was the last site in my bookmarks to allow http connections, and therefore was what I always used to trigger redirection to evil wifi captive portals. And 99% of public wifi is evil nowadays.

What do you use? 

Post has attachment
I think that https://sandstorm.io is just the bee's knees, and probably the best possible solution to the (big, important) problem of centralization of webservers within the walled gardens of advertising megacorps, but...

But now I'm wondering, why do we need webservers at all?
I look at the webapps at https://apps.sandstorm.io/ ,
and think about the ones that are missing (google web search, facebook social network, twitter microblogging, etc.), and all i seem to see is:

65% static content
15% logins and ACLs
10% advertising
5% indexing and search
4% notification / pubsub
1% actual serverside computation that can't be moved client-side

Nowadays, I think that the vast majority of corpi that people actually want to index+search is small enough to actually slurp down and do the index+search client side. You still need google to run a server-side search because downloading the entire web to your client in order search it isn't feasible . But it seems like 99% of the "web of services" could actually be totally decentralized (with all content distributed and all execution local to your client) if we just put the right UI on it.

People are finally realizing that the webservices shouldn't be doing logons. And public-key crypto has a (slightly thorny, but perfectly functional) answer to ACLs.

So what am I missing?

Post has attachment

Post has attachment
Greenwald makes a convincing argument that "Russian cyberwarriors" is this decade's "Iraqi WMDs".

Proposed rule: Once a UI control has been displayed to a user on a touch screen, it must never be removed, or change its size, position, or function, except in response to a user action. (Changing appearance and enabled/disabled state is allowed but discouraged).

Seems simple enough to implement and would save a lot of user rage. What do you think?
Wait while more posts are being loaded