Hmm. A few of the videos were broken for me in Firefox, but worked in Chrome. In particular, the 3D model of the plane (caption "On a real plane these control surfaces are moved by large hydraulic pistons connected to the pilot's control stick."), and the PCB design software ("PCB design is quite fun. You start off by placing the ‘footprints’ of each of the chips, and then draw in all the connections.").
Ah, looking at it, it appears that these may not have webm variants. You might want to include a static fallback image or something of the sort for videos that you don't have webm variants for; in fact, static fallback images for all of the videos would be good for people who have video disabled, don't have video support, don't have appropriate codec support, or the like.
I believe (can't recall the source at the moment) that on Google computers, they actually watch all of your input for password input, and if you enter your password somewhere other than the official Google single-sign on interface, will make you rotate your password. They're pretty serious about not letting you type your password anywhere other than where you're supposed to.
That sounds like a pretty big security hole. Just "type" random letters until you get a warning saying not to enter your password outside of password fields.
There's no such warning displayed, because that would be a security hole.
This password security measure is a Chrome extension that's required by company policy to be installed on all corporate machines. It watches all input (to browser forms) and if it detects your password being typed anywhere other than an actual sign-on page, then the next time you sign on successfully you're required to change your password. I believe there's also an e-mail notification, but it's delayed.
This is actually a pretty good password security technique, specifically because people often inadvertently type their password into the wrong forms due to focus errors, lack of caffeine, etc.
Because I can't think of an efficient way to do that that doesn't involve having the extension have access to the password.
I mean, you could store the password hash + length, but then you're securely hashing every single overlapping substring of what you enter, which is not exactly fast. Especially as KDFs are designed to be slow.
And if you store the password hash then you enable an offline attack.
So then you're sending every keystroke people make to a central server?
Even assuming that the connection is secure (never a good assumption), that still means that there is a single point of failure. And one with drastic consequences.
True, but anyone that can gain privileged access to the computer is already king of the castle. Why attack it offline when you can just keylog it? I think it goes back to being one part of an overall security posture. Encrypt your workstations and people can't just pick them up and own them.
This seems really unlikely. Do they really calculate hashes for all words in a google document? What if your password is a sentence? Do they calculate hashes for all the fields in a spreadsheet too?
If you're being a responsible provider, you will be doing backups, in case your primary data dies. But backups are not all that useful if they can just be wiped out by an online process. There's a potential secret copy lying around.
Another possibility is that they run a caching system, that store cache on faster but still persistent storage (or even in memory on systems that don't reboot often). Does their caching setup ensure that it tells all involved caches to delete their copies of the data?
Many systems do some kind of transcoding, generating thumbnails or more highly compressed versions of uploaded video. Does requesting that a file be deleted immediately delete every file it was based on or derived from it?
Furthermore, when you delete data from most filesystems, it doesn't actually get overwritten, just the reference to it is removed. So even if you do delete all the copies that you know about, there's a good chance that the data is still on the disk, ready to be accessed by an off-the-shelf data recovery/forensics tool.
If you're a provider that runs a large amount of storage, you likely run through hard drives at a fairly regular rate, once they start throwing enough SMART errors. Do all providers properly securely erase failing hard drives before sending them off to recycling? Swapping out a failing drive and sending it off to recycling without securely erasing it first likely leads to other copies floating around in the wild that the provider no longer even has access to erase.
I would be more surprised to find providers that actually did securely overwrite every copy they had of a given piece of data when you asked for it to be deleted, than I would by a revelation that providers that "secretly kept a copy around".
It has nothing to do with the standards being strict; the standards can be as strict as they want. If the standards are strict and useless, no one will follow them, instead implementing something less strict and more useful.
For example, when downloading a file from a website, what default name should you use for it? There is a header to tell you, but not ever page supplies such a header; so the browser needs to do something. It chooses to pick the last component of the URL as that filename. However, URLs are somewhat more complex than you might expect, so this becomes more complicated and can lead to attacker controlled ways to manipulate this filename.
Now, you could make a more strict spec, for example by forbidding downloading files unless the filename is properly specified, or forbidding using any kind of default filename and making the user choose it themselves, or something of the sort. But if any browser vendor implemented this more strict spec, they would instantly annoy a lot of users who would find things breaking that used to work, and they would be likely to switch to another more permissive browser.
Security, compatibility, and robustness are hard factors to balance. Just blaming this on "slop in protocols" is a vast over simplification.
>> For example, when downloading a file from a website, what default name should you use for it? There is a header to tell you, but not ever page supplies such a header; so the browser needs to do something. It chooses to pick the last component of the URL as that filename.
Yeah, and that's slop in the protocol. If the header was required everything would still work, web sites would just have to fill in the header. What's easier to do, comply with a protocol where your site brakes if you don't, or to have swiss cheese and then make site developers learn a bunch of security best practices and hope they get it right?
Also in there is the good old "this site wants to blah blah" and ask the user to decide. If you have to ask, the answer is "No! fix your site so it's not on the user to decide". Broken certificates? Not my problem, browser should just say "sorry site security is busted" and leave it at that. It's an old debate, but AFIAC there is no debate, only lazyness.
Yes, the author mentions that in his paper. He contacted both Google and Microsoft; sounds like Google rolled out fixes before publication, while Microsoft is still working on them:
On March 2014, I reported a security feature bypass to
Microsoft which enables batch files (“bat” and “cmd”
extensions) to execute immediately without warning the
user about the publisher or origin of the file. Hence,
RFD malware that uses the bypass will execute
immediately once clicked.
...
Microsoft is working on a Defense-in-Depth fix to solve
this issue.
And:
This is the exact problem that multiple Google APIs
suffered from until I reported it to the Google security
team, leading to a massive fix in core Google components.
Only if you have some existing interest in the hardware in question. This is a hobby retro-computing project, for a chip that was popular 30 years ago.
If you want to get into a low-level, open source project that is relevant today, I would try RISC-V (http://riscv.org/). It's an ISA and family of processor cores designed to be competitive with the niche that ARM usually fills these days, but fully open source, with a freely implemtable ISA. Now, it's pretty new and you can't buy RISC-V chips yet, but it's done by a team led by David Patterson, who is one of the fathers of the RISC architecture, and it looks pretty promising as a new open ISA and family of processor cores.
Or if you want to work on hardware that is actually available today, ARM would probably be your best bet. Maybe try porting Linux, or an RTOS like NuttX (http://www.nuttx.org/), to an ARM SOC that it doesn't yet run on.
I haven't used rr, but I have used GDB's built in record and replay ability.
GDB's was quite useful, but painfully slow; I had to do a good amount of bisecting down to a small enough input to reproduce the issue without having to wait ages to execute before it was usable. Once I did that, though, it let me find the issue a lot sooner than I would have been able to otherwise.
rr promises to be substantially faster the GDB's record and replay; if so, I imagine it will be quite useful, though it is only once or twice a year that I actually have to debug a problem that these kinds of tools are relevant for.
undodb gdb (a commercial product) works a lot faster than gdb (which is so slow it's unusable). We used it with success to debug programs as big as pypy
Yes, I meant faster than gdb's native record and replay feature.
The nice thing about rr is that you can still use it with gdb, so you have the full power of gdb along with close to native speed record and reply. Once the x86-64 support is ready, it'll probably be a common tool for me.
In seti@home, they are all being controlled from a single location to work on a single task, while the entire internet consists of independent machines operated independently. That's what allows you to talk about seti@home as a single supercomputer, even though it's somewhat tenuous as the problem is so embarrassingly parallel, it would never compete on any actual supercomputer benchmarks which require low-latency communication between nodes.
The first cg company to use 10GbE? I find that hard to believe. I develop storage boxes for video production, and 10GbE is pretty much standard these days, you only go with gigabit if you're being really cheap. We are now starting to sell 40GbE systems on the high end, as well as bonded 10GbE for those that need a bit better than 10 but don't want to shell out for all of the really expensive 40G gear.
It is the first CG company I have heard of or seen using it. You are talking about a different industry. Not everything in visual effects studios is higher tech than anywhere else. The bottleneck is often expensive network storage and not the network itself.
I would imagine that right now studios are looking at 10gb ethernet, especially since SSDs should be making their way into the enterprise level gradually. Many times reading a scene goes at 50MBs.
It helps to realize that these studios are not flush with cash. They also might not always spend money in the best places (although usually they know what they are doing or they go out of business). Articles paint them to be super high tech, but really everything is built out of commodity hardware except for the backbone disks and routers.
Also upgrading to 10Gb means upgrading thousands of boxes, not just a couple. It could also mean running new cable but you would probably know more about that.
Ah, looking at it, it appears that these may not have webm variants. You might want to include a static fallback image or something of the sort for videos that you don't have webm variants for; in fact, static fallback images for all of the videos would be good for people who have video disabled, don't have video support, don't have appropriate codec support, or the like.