Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?

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:cost/gflop (Score 1) 68

by Carnildo (#49279237) Attached to: GCHQ Builds a Raspberry Pi Super Computer Cluster

RPi: $35 per node * 64 nodes = $2240

i7-4790: $500 per node * 64 nodes = $32000

If you're testing cluster-computing algorithms, it doesn't really matter how fast your nodes are, but there are situations you'll encounter with a real networked cluster that you can't simulate on a single compute node, no matter how fast it is. It's much like how you can use multitasking on a single-threaded processor to simulate a multicore processor, but there are entire families of contention issues you'll never encounter.

Comment: Re:A sense of scale (Score 1) 24

by Carnildo (#48937135) Attached to: Spire Plans To Use Tiny Satellites For More Accurate Weather Forecasts

Somebody is certainly missing a sense of scale.

Traditional Earth observation is done using a small number of satellites at a large distance, traditionally in geostationary orbit (35,786 km away). Using a large number of satellites in low orbit (300 km away), you can use low-power transmitters and commodity cameras. Sure, without cooling, you lose the thermal IR range, but in return you gain a great deal of resolution in the other bands.

Comment: Re: Relays, not exit nodes (Score 1) 80

by Carnildo (#48937027) Attached to: Mozilla Dusts Off Old Servers, Lights Up Tor Relays

Let me ask you this, since you obviously didn't think about it...if Tor is so good at protecting privacy and traffic, how does the DoJ know what percentage of ANYTHING is going through it?

That's easy: you set up an exit node and watch the traffic going by.

Tor only promises to protect the data as it travels between your computer and the exit node. If you want protection after that, you'd better use SSL.

Comment: Re:I actually warned the FBI... (Score 3, Informative) 110

by Carnildo (#48835755) Attached to: To Avoid Detection, Terrorists Made Messages Seem Like Spam

You alerted them to actual spam.

The purpose of the suffix was to evade simple subject-line spam filters, while the "word salad" was an effort to evade word-classifier spam filters by drowning out the "spam-like" words with "non-spam" words, or to poison the classifiers and render them useless by loading up the "spam" wordlists with words that usually appear in non-spam messages.

Comment: Re:I've got an idea !! (Score 2) 248

by Carnildo (#48835663) Attached to: SpaceX Landing Attempt Video Released

Parachutes don't have the accuracy needed to land on a barge, and splashing down in the ocean means complete disassembly to get the residual salt off all the parts.

The Shuttle SRBs could do parachute recovery with ocean splashdown because they consisted of a small number of very large parts, and needed pressure-washing to get the fuel residue off anyway. Taking a liquid-fuel rocket apart is a much harder task.

Comment: Re:I'm amazed (Score 1) 169

by Carnildo (#48763033) Attached to: How Long Will It Take Streaming To Dominate the Music Business?

from your figures it appears your script is including the theoretical purchase price of music that Pandora chooses to play at you rather than just the musicyou actively selected. I mean it could play something you don't even like or would ever buy but your script would still include the cost.

I've got statistics on that, too. Pandora is 99.65% accurate at picking music I like (by play count), or 98.03% accurate (by track count). Doesn't change the cost by much.

What about the extra hidden cost of your internet connections themselves and the necessary extra bandwidth usage?

The amortized cost of my Internet connection probably doubles the effective cost of Pandora, but even if the entire cost were added, it would still be many times cheaper than the iTunes cost.

Comment: Re:I'm amazed (Score 1) 169

by Carnildo (#48751779) Attached to: How Long Will It Take Streaming To Dominate the Music Business?

1) You can't listen to your music when you dont have an active internet connection.
2) You're basically paying regularly/multiple times to hear the same music you could just pay for/download once.

I've been running a script to track my Pandora activity for almost eight years. According to it, my "collection" of music would cost me somewhere between $22,000 (iTunes) and $150,000 (CDs) if purchased, versus $300 or so for a Pandora subscription.

Yes, purchasing the music would let me play what I want when I want, even in the rare instances that my nearly-always-on Internet connection is down, but it's not worth a 75-fold increase in price.

(21,934 distinct tracks from 11,050 albums by approximately 6,596 artists, for a total of 190,330 tracks played.)

Comment: Re:"suspected pedophile" (Score 1) 150

by Carnildo (#48519343) Attached to: UK Authorities Launching Massive Child Abuse Database

When they claim that there are "tens of millions" of images in this database, I wonder how many are of victims and how many are cartoons found on 4chan or scans of children's clothes catalogues and that sort of thing.

I want to know how many are of teenagers. Reportedly the single largest source of child pornography these days is teenagers with cell-phone cameras taking steamy self-portraits.

Comment: Re:The only solution I can think of (Score 3, Informative) 136

by Carnildo (#48389391) Attached to: 81% of Tor Users Can Be De-anonymized By Analysing Router Information

Not really. Random jitter can be dealt with statistically: collect more data, compute the mean, and use the mean where you would have used the exact timing.

In order to defeat timing analysis through noise injection, you need to introduce a large amount of variation compared to the number of packets being sent; for any realistically-sized data transfer, this requires jitter on the order of minutes to hours.

Comment: Re:Put the SMART stats to the test (Score 1) 142

by Carnildo (#48375315) Attached to: Data Center Study Reveals Top 5 SMART Stats That Correlate To Drive Failures

Google did this about seven years ago. Of the stats, a drive with a non-zero scan error count has a 70% chance of surviving eight months, one with a non-zero reallocated sector count has a 85% chance of survival, and one with a non-zero pending sector count has a 75% chance of survival. For comparison, a drive with no error indications has a better than 99% chance of surviving eight months.

Overall, 44% of failures can be predicted with a low false-positive rate, while 64% can be predicted with an unacceptably high false-positive rate. 36% of drive failures occur with no SMART failure indications at all.

Comment: Re:Top #1 Indicator That Correlates To Drive Failu (Score 1) 142

by Carnildo (#48375247) Attached to: Data Center Study Reveals Top 5 SMART Stats That Correlate To Drive Failures

If you go by Google's definition of failing (the raw value of any of Reallocated_Sector_Ct, Current_Pending_Sector, or Offline_Uncorrectable goes non-zero) rather than the SMART definition of failing (any scaled value goes below the "failure threshold" value defined in the drive's firmware), about 40% of drive failures can be predicted with an acceptably low false-positive rate. You're correct, though, that the "SMART health assessment" is useless as a predictor of failure.

They did a study on this a few years back. It comes to about the same conclusions that Backblaze's study does, but with more numbers (and a larger data set).

Comment: Re:OpenPGP (Score 2) 63

by Carnildo (#48323009) Attached to: The Fight Over the EFF's Secure Messaging Scoreboard

The scorecard gives negative marks for both PGP for Mac and PGP for Windows, for both "Are past comms secure if your keys are stolen?" and "Has the code been audited?" Both negative marks are quite wrong!!

I don't know about the auditing, but the negative mark for "Are past comms secure if your keys are stolen?" is quite right. They're talking about forward secrecy, and PGP doesn't implement it. The basic idea of forward secrecy is that even if all the long-term secrets (passwords, keys, etc.) involved in a conversation are stolen, the person who stole them cannot go back and decrypt the encrypted messages.

Nothing in progression can rest on its original plan. We may as well think of rocking a grown man in the cradle of an infant. -- Edmund Burke