Forgot your password?
typodupeerror

Comment: Nuclear is dangerous - Fukushima overlooked (Score 1) 380

by HongPong (#44913937) Attached to: Its Nuclear Plant Closed, Maine Town Is Full of Regret

It seems most commenters on this thread overlook the severity of Fukushima. It's an uncomfortable reality for techies that tech like nuclear is not really run in a responsible - more expensive - way. In Minnesota the Monticello nuclear plant is the same GE Mk II as Fukushima with the terrible spent fuel chamber design. Fortunately MN has low disaster risks compared to say the uber-dangerous Cal Edison San Onofre plant - but look how even one nuclear accident proves impossible to contain or bring to a close?!
See http://en.wikipedia.org/wiki/San_Onofre_Nuclear_Generating_Station - imagine if a tsunami had hit this rickety place - it took tons of local pressure to shut it down finally. Nuclear plants sometimes degrade unpredictably like San Onofre did. Meanwhile in Japan, a country world renowned for its robot expertise, has been utterly unable to mobilize a Soviet style Chernobyl like response. Chernobyl popped once and then had a moderate fire, but Fukushima remains in slow meltdown, and if another typhoon or tsunami happens to hit the area just right, the remaining fuel rods could finally go off and metropolitan Tokyo could have to be evacuated. In this country the EPA has radically raised the 'acceptable' levels for radiation, and who knows how many nuclear incidents in the US go unreported??

I agree coal and oil based plants trigger major environmental consequences and natural gas plants drawn from fracking now cause geological & chemical damage. We need to focus on driving down aggregate demand for electricity and patching together intermittent sources (a new water-based heat cylinder idea for example could help w storage and peak, vertical windmills are safer for birds, clever plastic lenses cheapen solar etc), while phasing out catastrophe-prone technologies. In the new SimCity nuclear plants are safe if the workers are educated, if only real life were so easy :P

Comment: 700MHz Radio Spectrum battle continues (Score 1) 115

by HongPong (#38757110) Attached to: Post-9/11 DOJ Tech Project Dying After 10 Years?

I heard last year that first responders are trying to hang onto a chunk of radio spectrum that the telecoms want. I don't think it was really about encryption so much as making sure that it could do trunking correctly - units could bring in radios across the country and have working interoperability. Encryption is its own ball of crazy. I for one would rather have the fire fighters have better radios, the fuzz can generally get good radios if they want them.

This is apparently the "D Block" which is next to existing 700MHz public safety frequencies.
http://gcn.com/articles/2011/03/31/first-responders-public-safety-d-block-spectrum.aspx
later: http://www.nydailynews.com/blogs/dc/2011/06/911-first-responder-radio-bill-clears-committee

Comment: This code should be open-sourced? (Score 0) 199

by HongPong (#38757050) Attached to: Man Charged With Stealing Code From Federal Reserve Bank

There should be a git repository for all the code used for such core functions as the US Treasury ledger. Of course that would cause reporting to improve -- imagine if each budget operation got spit out in tweets or API-compatible calls. That would really mess up the routine at the Federal Reserve for laundering drug money & creating credit lines for foreign criminal banker arch weasels, so it's going to be closed source as far as they can take it.

Comment: SAIC builds out tracking systems roads-panopticon (Score 1) 310

I recently obtained info about SAIC participating in building a new tracking pilot system called IntelliDrive. Basically they are there to profit (cost plus) from approving the system. It's a huge industry to install military industrial tracking systems at every level of society. Story here:
http://tc.indymedia.org/2011/may/tcimc-exclusive-contracts-intellidrive-mndot-military-industrialu-m-plan-gps-track-all-cars

Comment: Consult zerohedge for cyber/market spam nexus (Score 4, Interesting) 51

by HongPong (#35181314) Attached to: Subtle Cyber Attacks Could Tilt Global Economies

The majority of stock activity is exactly this electronic noise, it's the rule, not the exception. The whole equities market (and other ones) is a turbo-hyperactive instant messaging system of bid/ask channels and they are all basically crapflooded now at about 60-70% of daily trading volume. The "Carbon Market" is another huge scam handy for passthru moneylaundering & fraud operations. Unknown binaries have been lodged in key NASDAQ systems. The messaging "order flow" is arbitrarily frontrun (i.e. man in the middle message intercept).

The real question is what miserable slice of the activity actually represents rationally allocated capital, rather than this message crapflooding. My fave site for this topic http://zerohedge.com/ consistently nails so many anomalies, flash crashes, order stacking rule loopholes (which bids are bumped off to 'dark pools' under less than optimal logic) etc.

It would be better if there was a 5ms 'quanta' for market prices, at least that would set a minimum time-to-live for price quotes. It becomes less 'rational' as the time horizon asymptotically approaches zero.

Comment: FBI has shutoff all non-terror resources basically (Score 1) 57

by HongPong (#34220708) Attached to: The Great Cyberheist

The thing is that the FBI has basically diverted all their white collar crime resources, and probably whatever might be used to track hacking / financial crime stuff, into stupid counter-terror campaigns. This whole mess is really a permutation of white-collar crime.

They haven't sent a single greater-than-pawn level obvious fraudulent white collar criminal to prison in like a decade. They catch a couple hackers running large creditcard schemes but they haven't done jack about the industrial espionage, which as you note is going 'all the while.'

I am mainly just sad that all this context is lost, the one primary thing feds are good at is 'making an example' and making sure that it appears to be a broad enough example that they are getting to the core of the matter.

Comment: Please check out the new Army domestic ops manual! (Score 1) 406

by HongPong (#32941496) Attached to: US Deploys 'Heat-Ray' In Afghanistan

Bad news brewing in here
http://cryptome.org/dodi/fm-3-28.zip
New Army Field Manual draft -- all this stuff is coming home as NORTHCOM-commanded Full Spectrum Dominance type doctrine. Please read this new revised Army field manual to have a better idea.

These domestic military operations are rapidly expanding - in recent weeks, mass scanning/stops in NY state and now in CA border areas. You *need* to study the details before something like the G20 descends on your city -- I have seen these domestic military crackdown ops up close and personal and it's really, really bad.

+ - MS Windows 7 Law Enforcement guide on Cryptome.org-> 1

Submitted by
HongPong
HongPong writes "In a continuation of the excitement around Microsoft's confidential Law Enforcement guide hitting Cryptome.org, now several more Law Enforcement Sensitive PDFs about Windows 7 have been posted, including a lot of detailed information about examining BitLocker drive encryption and potentially cracking it: "We can also see the Recovery Key ID number" and a series of hex addresses, it says (win7-bit-spy.pdf p 67). With all the guides Cryptome has posted for PayPal, MySpace, AOL, SKype, Yahoo! & others, one can certainly get a clearer picture of implementations of government demands, but also these training manuals created by the companies clearly illuminate their own intent. Also, who else has had this information? Isn't it deceptive marketing to peddle products with such backdoors or intended weaknesses?"
Link to Original Source

Comment: Inline doc help via wiki? Usability & design u (Score 1) 769

by HongPong (#30322302) Attached to: Is Linux Documentation Lacking?

K I skimmed this whole thread, the core problem & solution are elusive. Part of it is a decline in the 'harmony' of Linux app/desktop design integration, part is the information 'rot' of obsolete threads found on Google. The Gentoo wikis are pretty much the only bright spot here, no one can even cite a good GUI linux app documentation.

I'm not a Linux expert but I spend a lot of time dealing with Drupal which is also GPLed and regarded as a tough learning curve. They have dedicated a ton of effort into not just the documentation and forums but also U of M usability research. I met Dries at the U of M before they went in and looked at how peoples eyeballs scattered in panic because a RED ALERT BOX was worried their user creation password was not secure enough. They got a draft usability plan out of the research:
http://groups.drupal.org/node/9252 - and even video of eyes mapped around the screen.
In this case the information, inline documentation really, came in perceived as too hot by being RED so they changed it in Drupal 7 to light orange bkgnd. You structure the information to direct attention appropriately and then deliver snippets when the environment changes.

Think about it: we have totally divorced 'documentation' from even considering how important little snippets of text are, delivered correctly *with the correct level of detail* AND *the ability to seek up down and laterally in the conceptual environment*, instead thinking of man vs info vs annoying old threads. Probably the most important documentation, definitely for non-GUI Linux, are the small, less-than-ten-line, instructions and advisories that come before prompts. And usually these have HTTP links included for big deals. If everyone tripled their effort here it would work a lot better than just cleaning up the disastrously wrong (or certainly obsolete) design of man and info pages. Could familiar man pages spit out more examples and not exhaustive list of flags? (well it has to if you believe man must only be one page of stuff with all the programmer hooks, signals &etc. Where's the non-programmer material to be found then?)

Good wiki pages for software documentation usually break text into similar less-than-ten line sections, and do so in an up-down-lateral hierarchy of headers. Fortunately this can get exported and stashed into the app. If you had wiki paragraphs XML tagged to land in certain dialog boxes and other points, you could pipe wikified micro-documentation into the apps, even desktop apps. Hell if you just put a "WIKI??" button right there in the modal dialog box or prompt at least half the users would get it immediately. If a clever crew was handling this 'help string wiki' it would work out fine probably.But you'd have to control yr wiki or yr enemies would put in bad Windows YES/NO dialog boxes' help ("Do you want to print or save? YES/NO") just to mess w you.

Anything you present an end user should be structured in the newspaper article style - a pyramid structure leading with 'what does this do? how can i do the 3-5 basic things? why does this matter? what is this related to?' Beyond that you should be able to reach an overview of that part of the system architecture. Like apache2ctl would get you to rc and rc.conf and note what the runlevel is and why / which daemons are at runlevels.

There should be a clear ontology between nodes and levels of information. There is usually no explicit way to back out from a command to where the command fits in the system, or something you can run to lookup what a weird file does. maybe also the Apple 'receipt' type file that is a breadcrumb for packages could be used as a way to pull out documentation from different versions, another big gripe/snag here. There is not a lot of unity between Linux packaging systems and documentation and window managers. Obviously packaging info is already quite helpful but once things are installed it doesn't 'appear' anywhere useful, to other apps (imagine special warnings for diff versions INSIDE of a gui of something else) or the user. Package hierarchies are also useful and should appear inline somehow - help clarify the awesome world of Linux DVD writing ;-).

Another side of it is the kitchen sink app design, GIMP the leading example. Since a hierarchical executive isn't dictating design, usually the complex apps lack an overarching 'vision' or metaphor. (You don't need executives but they end up making the usability decisions traditionally) No one it seems has read 'The Design of Everyday Things' which describes the importance of letting users build a mental model of the object's function, which in turn provides them the cues for interacting with the object. The user doesn't need to know how it's wired but they need to perceive that lowering microwave power lowers its heat.

The problem is that Linux of course crams in as many functions as possible to each command, so the average user can never 'perceive' tar command at all. And GOD FORBID (because UNIX tradition tells us so) that the tar man page start off by telling you how to make a .tar.gz file of a directory... because it would be noncanonical to use tar the way 90% of the time you use it AND it would be non canonical to start with a working EXAMPLE of the most common 3-5 use cases. And the man pages are apparently purposed as tjwhaynes depressingly put it "not help pages". Why?! It's a manual to pipes, flags and hooks that can lead you to no other pages in an easy way.

Who decided this and how can it get changed? The fact that no 'front desk' of the OSS community could ever actually get the entire format of man pages straightened out. This is a structural problem of the open source world.

So I decree from the mighty /. thread: Man pages should include the leading 3-5 use case examples. The programmers and doc people have to actually determine these use cases because usability is central. The examples should be at the beginning. And this should somehow magically become 'canonical' in the OSS world. :)

Obviously it would be best over the long run for all packages to expose blocks of help for searching so that you can have an OSX or Google-like exposed help search. It could include all config file names, all flags and commands. The hierarchy to organize this would be similar to the form that the Gentoo wikis take, installation, networking, signals, hardware interfaces and cards, particular apps, etc.

You need to think about points of access to the help system. Where would you come into it? You should be able to drop error numbers and other obvious phrases into a searchytastic box.

I haven't been that deep with Linux lately, let alone touching desktop Linux. But this kind of usability problem plagues everyone - focus on inline documentation and usability, provide examples and let people move around with up-to-date and package/version tailored info. Every OSS project should have 'use cases' and 'skill level' practically as separate extra fields, next to 'license' and 'language'. And you could make a wiki-style participatory user interface dev app (lets wiki up these checkboxes). which would also be a good way to get people to develop add-ons/plugins.

Wherever you go...There you are. - Buckaroo Banzai

Working...