Profile cover photo
Profile photo
Michael Ploujnikov
110 followers
110 followers
About
Michael's posts

Post has attachment
The Lenovo X1 Carbon comes in some really stylish packaging!

#lenovo #x1carbon
Photo

Post has attachment
Kinda did this on a whim and it turned out better than expected: http://youtu.be/RvJv4gElxnw

Post has attachment
I've been declining updates to the GamePad app ever since it started to require "network access". I'm sure the GamePad developers have a legitimate reason for it, but I'm paranoid. In general, I don't think an input app should have access to the network to avoid intentional or unintentional "keylogger" behaviour.

Today I'm saddened to see that Google, in an attempt to simplify and hopefully improve security (https://support.google.com/googleplay/?p=app_permissions), is making it harder for me to secure my phone. At least it still asks if I want to allow the update rather than automatically installing the update along with other apps that don't need new permissions.

I am left with an uneasy feeling.

#android #security #keyloggerapps  
Photo

Post has attachment
Frosted glass
Photo

Post has attachment
This was pretty useful to me while recovering from the latest eCryptfs fiasco: https://github.com/clark800/dropbox-restore

Background: I had my home directory encrypted with eCryptfs since it was so easy to enable in Ubuntu. Unfortunately, it didn't handle out of disk space situations very nicely. When I ran out of space, eCryptfs pretty much erased important documents from the disk. This would have been easy to fix because at the time I still had copies of those documents in my Emacs buffer. However, I didn't realize that they were lost until I rebooted my machine. I was syncing a portion of my ~/.Private/ to dropbox, but again, by the time I realized the problem, even the dropbox copy was corrupted. That's where Dropbox' file versioning and dropbox-restore became really handy!

#dropbox #eCryptfs #Ubuntu

Post has attachment
And it still vibrates every time I receive a message. How do I stop it?

#BBMAndroid #BBM
Photo

Post has shared content

Post has attachment
My work laptop's USB port proved to be too tight for this Patriot USB stick. It lost its protective cover as I tried to unplug it.
Photo

Post has shared content
Icculus is at it again!
So Sam Lantinga and I are writing documentation for SDL2. We feel like the code is pretty good for a 2.0.0 release, but then we looked at the wiki and wanted to make sure people were more comfortable with using this code, so we got down to work.

We've been banging the hell out of this documentation for two days now, and that has proven to be a learning experience.

- We're discovering things we wrote ourselves do not work the way we thought. I don't even mean how they are implemented, but how they were intended to be used when we first wrote them. I totally had to relearn the audio conversion API today, like I had never seen it before.

- I found myself writing detailed documentation about a function that wasn't actually implemented, as if, in a fever-dream, I had actually written that code.

     http://hg.libsdl.org/SDL/rev/c6b4fab28e00

- When verifying that something works they way we planned to explain it, we have uncovered several bugs, like how some things only initialize if you call SDL_Init() vs SDL_InitSubSystem():

     http://hg.libsdl.org/SDL/rev/5706d7fc10d0

- Also when verifying, we're checking some basic assumptions. Case in point: dlopen() can take a RTLD_LOCAL or RTLD_GLOBAL flag which makes it behave in drastically different ways. There's a default behavior if you specify neither flag. Do you know what it is? If you looked at the Linux manpage then you know something different than if you looked at the Mac OS X manpage. Good thing Sam and I were using different OSes when we double-checked this:

     http://hg.libsdl.org/SDL/rev/4ea0b0177a57

- We found out things we thought were true weren't anymore. I had no idea SDL 2.0 removed the "parachute":

     http://hg.libsdl.org/SDL/rev/e9a7333d86f0

- We found out that the world marches on even if your code doesn't:

     http://hg.libsdl.org/SDL/rev/063a932a264f

- We found out you have to be able to say what will happen in corner cases, because corner cases happen:

     http://hg.libsdl.org/SDL/rev/b7219f6cf67a

- And we found out that a lot of things that seemed like good ideas really suck. Documenting something with design flaws makes you realize the problems, because you feel like an asshole having to explain those flaws. We jot down a note to fix this or that on the 2.1 TODO list, explain the best we can, and move on to the next wiki page.

Post has attachment
Wait while more posts are being loaded