dlitz

29 ms·

121.
by dlitz 13y ago
Most Bitcoin exchanges do this now. It's pretty standard anti-money-laundering KYC stuff.
122.
by dlitz 13y ago
Why should he have to avoid NJ, though?
123.
by dlitz 13y ago
TLS has demonstrated no such thing. Right now, you can still subvert most TLS deployments by subverting the DNS, because everyone trusts CAs who provide signatures that rely on unauthenticated MX record lookups followed by cleartext SMTP c
124.
by dlitz 13y ago
I guess this is yet another reason not to rely on firewalls and outdated notions of "internal" and "external" networks for any kind of real security. Increasingly, it seems that firewalls are doing less to improve actual
125.
by dlitz 13y ago
Debian unstable
126.
by dlitz 13y ago
You're not necessarily wrong, but uIP may not be the best example of a good minimal TCP implementation. I've seen it routinely put the wrong value into the TCP window-size field. This creates an illusion of packet loss, causing
127.
by dlitz 13y ago
I don't trust people who blindly follow orders.
128.
by dlitz 13y ago
Your argument seems to make sense, but why does everyone seem to agree that it would still be a felony, then?
129.
by dlitz 13y ago
There was a brief time when aptitude was smarter than apt-get (it was smart enough to remove auto-installed dependencies when you removed a package, for example). During that time, the advice was to use aptitude instead of apt-get. Those s
130.
by dlitz 13y ago
'except Felony'
131.
by dlitz 13y ago
The master/slave analogy isn't comparable. The terms "master" and "slave" refer to the relationships between technical components of the system, not to people. Every time inclusive language comes up, there are
132.
by dlitz 13y ago
Another problem with UEFI Secure Boot is that boot images can apparently only be signed by one key, which is why we see crazy things like Canonical going to Microsoft to get their Ubuntu images signed. (It's a nice, subtle, anti-compe
133.
by dlitz 13y ago
TPM chips are designed to enforce the rules of whoever controls the TPM chip over the will of whoever is sitting at the computer. This can be good or bad, depending on who those respective parties are.
134.
by dlitz 13y ago
... and then reverting the change when another committer landed it.
135.
by dlitz 13y ago
The problem is that people try to eliminate uncertainty instead of trying to manage it effectively. If you're trying to make estimates with more than order-of-magnitude precision, then you're probably doing something wrong. If we
136.
by dlitz 13y ago
It managed to insert invalid Unicode into a SQLite database, causing a subsequent SELECT to fail. That's at least a DoS attack.
137.
by dlitz 13y ago
Well then what's the alleged problem?
138.
by dlitz 13y ago
Filtering this stuff reliably isn't exactly free.
139.
by dlitz 13y ago
> Instead we're faced with the constant churn of semi-functional code, users working in silos and then showing up with patches well after it's too late to give them direction, animated GIFs instead of careful engineering discus
140.
by dlitz 13y ago
> Use encrypt-then-HMAC. Seriously. Yes really. No other choice is defensible in this age. If you think “oh but GCM” look more closely into its design. What's the problem with GCM?
141.
by dlitz 13y ago
That's odd. It should be fine, but here are some things to check: - Make sure that none of the filesystems on the USB stick are mounted before you begin. Otherwise, the filesystem drivers might corrupt something. - Use a larger block
142.
by dlitz 13y ago
Debian's CD images are actually hybrid CD/DVD/USB images: http://www.debian.org/releases/stable/i386/ch04s03.html.en It's been a feature of isolinux for a while now: http://www.
143.
by dlitz 13y ago
> or SHA-2 Nice try, General Alexander.
144.
by dlitz 13y ago
I don't think that would work. Bitcoin miners only need a small header (the root of a hash tree, essentially) in order to mine, but you need the full contents of the block in order for it to be broadcast into the honest network.
145.
by dlitz 13y ago
AV software is effectively a really complicated blacklist-based form of access control, which is a terrible way to do security in the first place. There's no reason for antivirus software of this kind of exist on desktop Linux, because
146.
by dlitz 13y ago
The problem is that if you actually look at how this AV software works (and at how many AV companies behave), there are clear reasons not to trust them. For example: https://lock.cmpxchg8b.com/sophailv2.pdf The problem of m
147.
by dlitz 13y ago
If you read about the technical implementation of the AV software that's been analyzed, you might not think it's that dangerous after all: http://www.cso.com.au/article/441070/google_security_researc...
148.
by dlitz 13y ago
If you're already familiar with those tools, the incremental effort to put them together is trivial. It's no different from saying that driving to the store is trivial: People who don't already know how to drive would disagre
149.
by dlitz 13y ago
That's what you get when you play numbers games poorly . A better way to do it is to maintain the same admission standards for everybody, but actively seek out more applications from women. It'll indirectly raise the number of
150.
by dlitz 13y ago
So does this: ChallengeResponseAuthentication no PasswordAuthentication no

More ›