Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror

Slashdot videos: Now with more Slashdot!

  • View

  • Discuss

  • Share

We've improved Slashdot's video section; now you can view our video interviews, product close-ups and site visits with all the usual Slashdot options to comment, share, etc. No more walled garden! It's a work in progress -- we hope you'll check it out (Learn more about the recent updates).

×

Comment: Re:I'd suggest to recommend uninstalling windows t (Score 3, Insightful) 134

by blueg3 (#49101665) Attached to: Homeland Security Urges Lenovo Customers To Remove Superfish

That may be true.

It's not applicable in this case, because this is OEM-installed adware. Everything it does can be implemented just fine on a Linux system. The solution is really the same for this sort of thing regardless of whether you're talking Windows or Linux -- don't use the OEM-provided pile of crapware that comes with the machine; install a brand-new copy of just the OS.

Comment: Re:The headlne and the text say different things (Score 1) 114

The headline that they are "tied to NSA"... but TFA says that "researchers stopped short of saying Equation Group was the handiwork of the NSA."

That's a clever turn of phrase. Kaspersky pointedly calls them out as NSA, but doesn't explicitly say "this is a group at NSA".

Comment: Re:Dunning Kreuger effect (Score 1) 809

by blueg3 (#49060327) Attached to: Ask Slashdot: What Portion of Developers Are Bad At What They Do?

They certainly are not. Some database implementations require a filesystem, but plenty do not -- they work with raw block devices.

Further, you can build a filesystem using a different filesystem. Take, for example, glusterfs, unionfs, EncFS, or Samba's "NTFS features on top of a non-NTFS filesystem" implementation.

Comment: Re:Dunning Kreuger effect (Score 1) 809

by blueg3 (#49060301) Attached to: Ask Slashdot: What Portion of Developers Are Bad At What They Do?

I suggest taking to heart the words of Fred Brooks -- or of numerous other computer scientists who have said similarly:
"Show me your flowcharts and conceal your tables, and I shall continue to be mystified. Show me your tables, and I won’t usually need your flowcharts; they’ll be obvious."

Good data structures, particularly for a data-structure-oriented system (like a filesystem), imply the algorithms to be used with no further comment.

I can, for example, tell you the structure of the Volume Header and of the Catalog B-Tree file in HFS+ and you could use only that information to implement reading data from HFS+ (for files with fewer than 8 fragments). Little more is required for writing (and little more is required for fragmented files). Data structures is all Tech Note 1150 gives you, really, and it's enough to implement HFS+.

Comment: Re:Dunning Kreuger effect (Score 1) 809

by blueg3 (#49055947) Attached to: Ask Slashdot: What Portion of Developers Are Bad At What They Do?

a database has a way to address individual "items" that exist on a far lower level than a "file" with the ability to read and update or delete them

So the items in a database are in theory smaller and there are more of them. That's a practical and minor difference, not a fundamental difference. After all, plenty of filesystems have far more items than most databases. Lots of files are much smaller than many database elements.

Note that you can implement a filesystem using a database and vice versa.

I appreciate your distinction between "how it works" and "what components it consists of", but I think that unless you're being excessively pedantic, there is not a significant difference when it comes to software and especially when it comes to things like filesystems that are collections of data structures. Organized data structures generally imply exactly how they're used with little additional explanation.

Comment: Re:Dunning Kreuger effect (Score 1) 809

by blueg3 (#49051993) Attached to: Ask Slashdot: What Portion of Developers Are Bad At What They Do?

You don't know how file systems work,

Says you, sans evidence.

ignores both the OS-specific driver stuff and all the messy details. That's the thing about storage - the job is 95% about the messy details.

You must've missed "high level summary". They explicitly ignore the details in order to discuss the overall architecture. Which is important. You can't reasonably start with the details and expect to understand it -- though lots of people do, and end up understanding the details without understanding the overall structure (resulting in saying ridiculous things like "filesystems are not databases").

Filesystems are not OS-specific and they don't need drivers. They're bits on disk (or any other storage mechanism). Or blocks on disk, if you prefer.

I find too few candidates can even describe to me why doing file access in 2 threads helps.

Maybe so, but that doesn't have a damn thing to do with how a filesystem works.

Comment: Re:Dunning Kreuger effect (Score 1) 809

by blueg3 (#49051979) Attached to: Ask Slashdot: What Portion of Developers Are Bad At What They Do?

I'll bite -- what feature or lack thereof makes a filesystem not a database?

Bear in mind that "database" is a quite general term and that I didn't say it was any particular type of database (e.g., transactional, relational, key-value, etc.).

"Describe a filesystem" is different from "how does a filesystem work".

Everything after the first sentence (which is an introduction) is, in fact, how it works.

Comment: Re:Dunning Kreuger effect (Score 1) 809

by blueg3 (#49050331) Attached to: Ask Slashdot: What Portion of Developers Are Bad At What They Do?

No, you can explain how it works in five minutes, given some background in data structures. (You cannot cover the details of how a particular filesystem, particularly a fancy one, works.)

It's a database that manages the allocation of fixed-size blocks on disk to files and stores metadata about those files. It generally has a header at a fixed position on disk that identifies the filesystem, stores filesystem-wide metadata, and contains a pointer (rather, offset and length) to the index of files. The index of files is a data structure (varies per filesystem; example: B-Tree) that stores a record per file on the filesystem. The record contains metadata for the file. Metadata varies per file system, but the key metadata stored is the collection of blocks on disk (and their order) owned by (allocated to) that file. Generally, every file gets a unique identifier and directories are implemented as lists of the unique IDs of files contained in the directory (plus, potentially, other metadata), though some filesystems implement directories differently.

Knowing too much about filesystem should not prevent you from being able to describe how they work at a high level. If it does, the problem is not knowing too much, but focusing too often on details in a context where the details are not warranted.

Comment: Re:Dunning Kreuger effect (Score 1) 809

by blueg3 (#49049827) Attached to: Ask Slashdot: What Portion of Developers Are Bad At What They Do?

'How file systems work' would span one book, minimum.

How file systems work at a high level takes about five minutes and a small whiteboard.

At a slightly more detailed level, a chapter out of any standard undergrad-level operating systems textbook.

The details of how one particular filesystem works, at a level such that you could reimplement it, takes about one book.

Comment: Re:More than a little retarded (Score 4, Insightful) 129

by blueg3 (#49022175) Attached to: The Technologies That Betrayed Silk Road's Anonymity

This is true.

I mean, the "cybercrime" investigators that work for the FBI are not stupid and they're not incompetent. If you're running a large, well-known drug-selling site, they probably will put resources into finding you. On top of that, the deck is really stacked against you -- as a criminal, you need to avoid making any mistakes, while the investigator only need to wait for you to make a mistake. They're patient. (And "investigator" is not just people working for the police -- it's also anyone who might both have reason to dislike you and also motivation to reveal your identity to the police.) So, it may well be possible to hide indefinitely from prosecution, but it's not easy.

Comment: Re:thank god for the poor states (Score 1) 297

by blueg3 (#48995657) Attached to: Mississippi - the Nation's Leader In Vaccination Rates

You dont have to search for a clinic that provides the free vaccines, it doesn't change if I move states, your records move with you easily.

Here, you don't have to search for a clinic that provides free vaccines, either. It's a local government department.

All local services change if you move states. You certainly don't see the same GP after you move, no?

Comment: Re:thank god for the poor states (Score 4, Informative) 297

by blueg3 (#48988377) Attached to: Mississippi - the Nation's Leader In Vaccination Rates

Everywhere I've lived in the US, vaccinations are provided gratis by the local health department.

People with insurance usually go to a doctor and get their vaccinations through them, but the health department will also do it for free. (That's the same health department that will run free STD tests.) Often, the real battle is communicating to people that these resources are available, fighting the stigma associated with getting free services from the government, and the practical issues of getting a working person over to a busy government office.

As many childhood vaccinations are practically mandatory in the US, as they're required for attending elementary school (which is also mandatory), it makes sense that they're freely available.

As a result, I think, of Obamacare, all childhood vaccines and most adult vaccines (including flu) are free to anyone with insurance.

Comment: Re:It's not about the presenter. (Score 1) 227

by blueg3 (#48806833) Attached to: Lawrence Krauss On Scientists As Celebrities: Good For Science?

It depends. You can get anyone to read a script on a recorded television show, but that's usually not the point. You can't really do an interesting interview with someone who is a good communicator but isn't familiar with the subject matter.

De Grasse Tyson is often talking about tacheons, wormholes and white holes and other claptrap that's horribly speculative, wildly unusupported, and very probably untrue.

It depends. All of those are supported by theory. They're only "speculative" in that they are permitted by theory but have not been observed and, as a result, it's an interesting question as to whether or not they exist. (If they don't or can't exist, can that information be used to improve our theory?) So, depending on the context and what you say about them, they are potentially interesting lines of discussion, because they offer insight into what (some) physicists are still looking in to.

When the weight of the paperwork equals the weight of the plane, the plane will fly. -- Donald Douglas

Working...