Posts by tom dial 2187 publicly visible posts • joined 16 Jan 2011
Many years ago, around 1955, my parents' house was burgled while we were there. The police told us at the time that they considered burglaries of occupied houses especially serious because those burglars, as against those who took care to choose homes that appeared at least temporarily vacant, quite often were armed and presumed prepared to their guns if challenged.
This was in an inner ring suburb of a major city in which gun ownership was uncommon.
Actual evidence that violent video games lead to objectively antisocial behavior seems to be somewhere between rare and nonexistent. Statistical correlation, which may exist, would not establish a causal connection. It seems likely enough that those who become spree shooters exhibit much the same video game behavior as the far more numerous (by at least 4 orders of magnitude) gaming population who do not.
The New Yorker has an interesting piece from a couple of years ago that touches on this.
(https://www.newyorker.com/magazine/2015/10/19/thresholds-of-violence)
Roughly 99.9% (+~0.1%/-~.2%) of so called assault rifles in the US are used for target shooting or varmint/small game hunting. Contrary to an often expressed opinion, these are legitimate civilian uses. By current reports, the Santa Fe High School killings used a revolver and a shotgun (possibly shortened illegally), rather than an assault rifle, however defined.
Roughly speaking, the US media and those who refuse to accept the plain historical meaning of the second amendment (and the corresponding part of the English 1689 Bill of Rights) classify as "assault rifle" any self-reloading rifle that uses detachable magazines and has a pistol grip or any of several other mostly cosmetic characteristics that make it look like a military issue weapon. The most common variants, chambered for .223 or 5.6 mm ammunition, generally are unsuitable for hunting large game like deer due their low power, and are prohibited for that use in a number of states. At close range, like many other firearms, they make terrible wounds sometimes described in gory detail in articles, but they probably are not materially different from other weapons that deliver comparable kinetic energy. Limiting legal magazine capacity might have a minor effect, but it is worth keeping mindful that with practice it takes under two seconds to change a magazine. Banning these weapons is largely a waste of effort.
Meaningfully licensing gun ownership, with a similarly meaningful requirement for safety training, is unquestionably a good idea, but is unlikely to affect the occurrence rate of these awful rampage shootings other than, perhaps, making those who undertake them more skilled at their self-assigned task. The result of safety training failures is, if I remember correctly, in the order of 500 per year, some of which would occur even with a strict, frequent, and well-implemented training requirement. Classroom instruction, no matter how thorough, tends to evaporate over time.
Gun safes are widely used by responsible firearm owners, apparently including the parents of the of the Santa Fe High School shooter, and are, I think required in some jurisdictions. Biometric locks, suggested in one post, might be helpful, although there are numerous reports of methods by such things they can be and have been circumvented. It is not clear that more effort in this direction is likely to be very beneficial.
More extensive and thoroughly implemented checks preliminary to buying guns also is a worthy effort, but probably wouldn't improve things by much. Like credit checks, they will yield nothing for individuals with no history, and will not prevent those with no known history of bad behavior from acquiring weapons under whatever other control regime may exist. In the context of the US constitution, there also are due process issues related to abridgment of rights. A ban on firearms acquisition by "mental cases" or those on terrorist watch lists would be likely to be overturned by the courts unless backed by a requirement for specific evidence for each case. Psychiatrists also would object on the grounds that the great majority of their patients are not a threat to themselves or others. And while they generally are limited by professional confidentiality standards, they may and in some cases are required to report dangerous patients, but that is a judgment based on inexact knowledge and certainly will miss some cases in both directions.
To a first approximation, there are somewhat more than 30,000 firearm deaths annually in the US. About 2/3 of these are suicides, which some would argue exercise a fundamental and inalienable human right. Around 500 are accidental or by children playing with unattended or "found" weapons. Of the remaining 9,500 - 10,000, the great majority are related one way or another to what might be considered "ordinary" criminal activity of various kinds, including some family blowups that sometimes are classified as "mass shootings".
A small number, probably averaging around 100 - 200 per year, can by some standards be reasonably placed in the category of these school killings, the Las Vegas massacre last year, and others. It is not clear that these can be ended or significantly reduced by any actions short of repealing the second amendment and enacting seriously restrictive laws that result in confiscation (presumably with compensation) of nearly all privately owned firearms. In the US that is unlikely to happen any time soon, if ever.
Fixation on the "power" of the NRA is a distraction and an excuse. The NRA has, at bottom, the power of speech, combined with the funding necessary to support the advocacy staff, provided largely by voluntary membership dues. Their influence - the weight incumbents and those who would like to replace them assign to their advocacy - results from the perception, probably correct, that the NRA "speaks for" a large number of voters, combined with a similar perception of the logic of their arguments. Failure to recognize this, which also is true of most successful lobbying, is delusional, magical thinking.
For the record, I am not and never was an NRA member or supporter and do not and never have owned a gun.
In the eighteenth century, people in the US had the best firearms they could afford. As they did in the nineteenth century, and the twentieth, and as they do now, subject generally to the laws in effect at the time. The argument, so called, that the second amendment protects ownership of flintlock muskets and pistols probably is not intended seriously, but even if not is rubbish.
"The mechanism of voting is temporarily broken in America, as it is in much of the rest of the world." Is this a comment on the fact that voting outcomes are what (usually) a majority want and not what a minority want? (The electoral college is a special case that solved a political problem in 1789 and for which a rational case still may be made despite the fact that the intent behind it largely has been destroyed).
Again, is the NRA successful because they induce people to love and acquire firearms, or because a very large number of people already believe, correctly, that they have a right to do so and willingness to join - and fund - the NRA to advocate for them, in accordance with their first amendment right "peaceably to assemble, and to petition the Government for a redress of grievances?"
Is contemporary social media an extension of the self-appointed institutional "press" or is it a vast expansion of the old fashioned and effective rumor mill that was done largely outside of any formal control? (My answer: a bit of each.) And is its content really greatly different from attacks on Thomas Jefferson or Grover Cleveland (for example) in their presidential campaigns?
Yet enormous political pressure is applied to search engines, which (not without merit) nearly always are labeled "Google" to restrict search output by limiting return of "terrorist" or "old or irrelevant" (for example) results.
Next to no demands are made for Google and other web search portals to simply return a list of results ordered by some reasonable definition of applicability to the input query.
The underlying logic is quite straightforward and easily understandable.
If a patent application is denied, the applicant will revise and resubmit it. This cycle can continue for a long time, at least until the pruned down patent no longer has any meaningful claims; given the obfuscation built into the language used in patent applications, that can continue for quite a few years. For most of that time, the patent is part of the patent office's, and of some patent examiner's, backlog. Having a large backlog is not good for publicity, seeking of Congressional approval, or examiners' annual appraisals.
If a patent application is approved, however, it becomes a completed work unit for patent office publicity, appropriation request, and personnel evaluation purposes. All organizational and employee motivations are aligned, and the examiner's natural inclination to complete work and justify a high appraisal provides at least a gentle push toward approval and against denial.
In addition, of course, there also is the obvious fact that patent applicants often are going to have more and deeper technical knowledge in their particular field that the patent examiner. And the claims presented in general language in the application always will have been constructed carefully to be as broad and abstract as allowable (or more: after all, the patent, if denied, can be retooled and resubmitted), requiring the examiner to imagine use cases that were entirely clear to the inventor. (After all,
The system is broken, and was as soon as methods and software patents were let in.
The activity the California AG is going after certainly is reprehensible. They (and apparently Florida, where two of the individuals charged were arrested) have a law making it illegal to charge for removing the arrest reports from a web site, although from this article and several others it is not clear whether those laws are the basis for the charge. It also is not obvious that the activity is extortion or involves money laundering or identity theft.
The records at issue are public records, and it is not clear that the first amendment permits laws that prohibit their republication or require their removal, whether with a fee or without. Civil law may provide a better approach: Gabiola v. Mugshots, a class action lawsuit in the Eastern District of Illinois apparently is going forward and will, one may hope, lay on the owners of the web site a crushing financial burden that will put them out of business and deter others who operate a similar business. It is remotely possible that members of the class will receive some compensation.
It seems possible there would be some mileage in a claim of too drunk to understand what he was doing. It would, however take a bit of evidence to justify the claim. Otherwise, the standard Miranda Warning is short and explicit, and probably comprehensible on about a grade 6 or better level, and enough like the similar warning given in the UK for Mr. Hutchins to understand it unless he was really quite impaired.
“You have the right to remain silent. Anything you say can and will be used against you in a court of law. You have the right to an attorney. If you cannot afford an attorney, one will be provided for you. Do you understand the rights I have just read to you? With these rights in mind, do you wish to speak to me?”
http://www.mirandawarning.org/whatareyourmirandarights.html
I've been beating that drum for over 20 years, pointing out that unencrypted email is like post cards. In that time I found only one other person who had bothered to set PGP. I persuaded one more to use it; he wouldn't until he found an application that would let him do it within his web browser of choice and hehabitually uses HTML. It is this kind of thing that convinces me there is little mileage in the current moral panic about Facebook and election meddling, or Google and privacy.
As noted, those who genuinely need the security shouldn't be using HTML email or decrypting it on a machine that ever touched the Internet (or a USB key). For the rest of us, who are unlikely to be worth targeting, PGP probably offers enough privacy even before this gets fixed. And for the overwhelming majority, it is of no consequence at all because they do not care that their email is open to general inspection.
Years ago in an environment where we had to do some routine conversions between EBCDIC and ASCII I found that IBM's FTP implementation did a pretty acceptable job. My recollection is that it did so automatically based on the native character sets of the endpoint machines. It didn't work for records containing IBM packed decimal numbers however, and for one application we wound up using binary FTP, a Perl module from CPAN to unpack the numbers, and EBCDIC to ASCII transliteration using dd. (Some idiotic quasi-political reason having to do with production control prevented changing the database extract program to produce the numbers in display format).
The CTIA (1) shows a graph indicating 396 million connections, which seems in the context to be wireless telephones. A Pew study from 2010 (2) reports an estimate of about 5 calls per day for each cell phone. On the assumption that a number did not decrease significantly over the next 7 years it seems likely that the NSA collection reported amounts to well under a tenth of a percent of all US cell calls for the year, and a smaller fraction of all US telephone calls. That would have been a bit under 2 billion calls daily, and I have seen an estimate of 6 billion daily, although without references.
The numbers are large, but as a proportion of the total they certainly do not indicate serious mass surveillance of US persons. Although it is somewhat a matter of taste, it is not clear they even qualify as "quite extensive." Numbers of US targets are similarly unsupportive of mass surveillance, as is the information about probable cause orders, targets, and search terms.
The report is interesting, though, and good on The Register for linking it. Too few reports on such things do that. Indeed, a quick search suggests the Washington Post and New York Times considered the report too inconsequential to mention. In fact The Register seems to have the first non-government article (known to Google) on the 2017 report .
(1) At https://www.ctia.org/the-wireless-industry/infographics-library,
(2) At http://www.pewinternet.org/2010/09/02/cell-phones-and-american-adults/
I've received several class-action payouts over the years, none for as much as $10, and one, as I recall it, was for coupons for a discount on a product I no longer wanted to buy.
I'm still waiting in the PS3 OtherOS suit, for which I have submitted all the paperwork, including screen shots of the PS3 running Yellow Dog Linux. The final hearing is on May 29, so there may be hope; but payouts that large seem to be distant outliers in consumer class action suits.
This inevitably is a problem for escrow schemes, and was a criticism of Clipper and its data analogue Capstone. It is an inherent vulnerability of key escrow, just as is exposure of a private key in a public key system. In Ozzie's scheme, the repository in one possible implementation contains private keys matching public keys on the devices. The difference, as Green observes, is that the opportunity to get all the keys wholesale would make the repository a high value target for attacks.
The keys in the repository, if held by the manufacturer, would not have to be delivered to either domestic or foreign law enforcement or intelligence agencies. They would be used to decrypt the encrypted key those agencies obtained from the device and sent for decryption.
What is to keep committed privacy advocates, whether or not criminals, from implementing their own cryptography? A few, in order of increasing importance, are:
1. Phones and other devices on which the manufacturer has made non-"store" difficult to install "for the customers' safety" of course.
2. Designing a secure encryption algorithm (especially one that is provably secure) is hard; many, including skilled cryptographers, have tried and fallen short.
3. Implementing a secure encryption scheme is hard and rarely done; many, including highly skilled and experienced programmers, have tried and fallen short.
4. Guaranteeing that a vulnerability free implementation of a provably secure encryption algorithm in an environment so that it is not vulnerable to operating system defects, both when installed and after all future OS modifications is hard, and since the nature of future changes is unpredictable, unlikely to be attainable.
5. Guaranteeing the entire system of (4) against all application software almost certainly will, over time, present opportunities to those intent on breaking in.
I doubt that managers at NSA or any other major SigInt organization worry much about this problem.
There might come a time, though, when it appears that online advertising, whether or not targeted, has less effect than its vendors would like to have their customers think. I sometimes am a bit mystified by rants about it until temporarily disabling my ad blocker; my wife has not received scareware (browser lockup with audio "Important message from Micro Soft") since I installed on on her laptop.
The Comey "October surprise" probably had little effect on actual voting. The claim that it does rests on the assumption that when he advised the Congress on October 28 of examination of Clinton email messages on Huma Abedin's personal laptop it caused an immediate and significant swing against her, but that the November 6 followup announcement failed to have the opposite effect. In fact, results from reelection polling completed before October 28 already were shifting modestly in favor of Trump, although weakly enough that several days later the Clinton campaign (publicly, at least) worried only about the presumptive victory not being a landslide.
The real cause of Clinton loss may be encapsulated in the New York Times article
https://www.nytimes.com/2016/11/07/us/politics/bill-clinton-campaign.html
which describes the campaign's failure to exploit Bill Clinton's demonstrated skills and knowledge, based on their analysis of the models they, the Obama campaigns, and the Democratic party had developed over the previous two election cycles.
Those eligible to vote are not stupid, but there is little reason to doubt that they are largely ignorant about public policy and generally deficient in the various related knowledge areas. Ilya Somin discusses this at length along with its basis and consequences in "Democracy and Political Ignorance."
We ought, however, to be quite skeptical about CA's expansive claims of influencing election outcomes by analyzing Facebook scrapings and delivering tailored political ads and using Facebook are things about which we should be quite skeptical. They are likely to be much more marketing hype than fact. Kogan stated as much, and I have seen elsewhere that a Republican National Committee spokesman said the voter information from the states was more useful and heavily used than CA's. For my money, CA's brilliant strategy was that it persuaded Robert Mercer, who is far from stupid, to pour money into their coffers.
Roughly 90% of Americans surveyed cannot name their US Representative and their two US Senators. A similar fraction of those who live in cities probably cannot name their city's mayor (or other chief executive) or their city council member, or their state representative or senator*, each of whom has more direct effect on them.
In that environment, in which a probably overwhelming majority have obtained their complete stock of information about this subject from entertainers like John Oliver and cartoon ads on TV, there is some reason to question their conclusion.
Elections are poor ways to decide major public policy**, and polls, no matter their design and conduct, are inferior to elections by orders of magnitude.
* Except Nebraska, where the only have Senators.
** Examples, in addition to network neutrality, include going to war, legalization of cannabis "for medical use" (a poor subject as well for legislatures), and both regulation of Facebook (and Twitter and other such) and "common sense gun control" in the US historical and constitutional context. All of these are too complicated and require too much general and domain specific knowledge to be decided directly by simple things like referenda or polls.
Internet users may need to ask, and answer rationally, what risks they face and having done so, they should respond accordingly. While there undoubtedly are some for whom the US or another government is the primary risk actor, it is quite certain that for the overwhelming majority the primary risk comes from criminals after their money.
Somewhere between, and toward the low end if the risk scale, is the rather ill-defined risk of private information gathering, of which the most likely effect is that it will be used to target them with advertising they do not wish to receive and the most damaging is likely to be its acquisition by criminals after their money.
Given the propensity of some billions of Facebook, Twitter, and other users to share their thoughts and personal information with something like the entire world, I doubt that Kaspersky or any other software can make much of a dent in anything beyond identifying and sometimes preventing unauthorized and nearly always illegal access to computer systems.
There is a lot more to "mainframe" than executing the same instruction set and running the same operating system.
About 1993, a subsidiary of Gould challenged my (DoD) agency to match their cluster, based on Motorola four-core processor, bundles of memory, and a 70 megabit/second (!) backplane running Unix against our (at the time) IBM ES9000 running MVS/ESA. After some months and a good deal of preparation that included running without the "overhead" of ACF-2 because the challenger had nothing comparable, the contest was engaged.
I do not remember the entire result, but one chart from the report stood out. It showed response time vs. transaction rate for a simulated interactive application. On the challenger's system, the response was well under a second up to a point that was fairly high, after which it increased at an increasing rate until at the end of the test it was four or five seconds and rapidly increasing. The same test run against the ES9000 showed a response time of about 1.1 seconds. At all offered transaction rates. We kept the ES9000, which also offered reliability, availability, and service that the challenging system did not.
The statement that big iron eventually will be eclipsed, presumably in I/O capacity, is rather badly in need of justification. What are the reasons to think that mainframe designers and vendors like IBM or Fujitsu will be unable to match the assemblers of giant stacks of Intel or ARM commodity processors?
Anyone who truly cares about secure communication should not trust Microsoft or any government to protect it. That necessarily entails the parties to the communication taking care of encryption and key security for themselves. Every method or procedure that relies on other people or organizations reduces security.
Oracle has been a major beneficiary of DoD purchasing and probably still is. When I left government employ at the end of 2011, Oracle had a fairly secure lock on DoD large and medium database business and we had a lot of little ones on it as well. There were a few DB2 large mainframe databases and a sprinkling of DB2 for Linux/Unix/Windows and SQL Server on Windows. I had tried without success to stir up a little interest in Postgresql, for which commercial support then was about $700 per year as against the extortionate Oracle rates.
The transmitter in a US located base station presumably would require an FCC issued license, and that seems to be essentially the FCC's formal position. Things being as they are, the Canadian Space Agency probably would work out the other details with NASA and others, as in section 5(3)(d) of the Canadian Space Agency Act.
So what? I read the text (at ThinkProgress); there is nothing in it that would have raised an eyebrow if it had appeared in the New York Times which, as a matter of fact, employs a full-time staff member (Margaret Sullivan) whose duties include receiving comments about incorrect, incomplete, or biased reporting.
The "middle-class as we knew it from the 1950s through the crash of 2008" originated from the huge scientific and industrial investment occasioned by WW II together with release of large amounts of semi-compulsory personal savings resulting from the concurrent war financing and rationing. It was given an extra kick by the GI Bill and the consequent step increase in the number of people with significant post-secondary education.
The post-war growth spurt was plainly on its way out well before 2008, as evidenced by the fact that from the 1970s or so maintaining the life style came to require two employed adults per household, where through most of the 1960s one often was sufficient. The crash of 2008 (and the somewhat similar savings and loan debacle of the late 1980s and early 1990s) resulted to a considerable degree from attempts, encouraged by the federal government, to kick the can down the road for a few more years, augmented by financial ignorance in the population, dishonesty of real estate lenders and buyers both, and probably a bit of bad luck.
It is not at all clear why one should attach significance to Sinclair's potential capability to reach 72% of the US public. No similar concern is expressed over the fact that ABC, CBS, Fox, NBC, and PBS each now reach 95% or more of the US public. Sinclair stations' local news reports in one broadcast area are unlikely to be passed to other broadcast areas, and certainly nowhere near 72% except as they report items of regional or national interest ilikely to be covered by the big 4 commercial networks and PBS.
Forty or fifty years ago, the type of thing the Russian government stands accused of was called propaganda, broadcast in various languages in the medium and high frequency radio bands, and sometimes fought with jammers. I never had trouble receiving the English language services of any of the major players, but there were enough active jammers to make me suspect that East German, Russian, and other Eastern bloc countries were trying to blot out Western broadcasts on a regular basis.
I do not remember anyone getting into a panic about this propaganda, then, that seems to be the standard response in the last couple of years.
If you assume that so-called "friends" would be able to spread your Facebook messages, what earthly reason justifies an assumption that third party apps those friends use would not have access to them as well?
I always assumed that anything I put on Facebook was beyond my control and immediately and widely available. Accordingly, I put very little there.
I think if I were a manufacturer, I would warrant the product only against malfunction caused by defects in assembly or of original parts or repairs by my authorized agents using replacement parts I approved. I see no reason to compel a manufacturer to extend a warranty to cover a product repaired by agents not under his control or supervision, or using parts, even if they exceed original specification, that were not explicitly approved for the purpose when installed.
The Teslas described above might better be handled by requiring Takata to provide the rebuilt salvage Tesla owner a replacement air bag and the same compensation it would provide Tesla if they were doing the installation. On the other hand, the law governing recalls may require otherwise, as may the laws of various states that cover rebuilt salvage cars, which normally seem to require a specific title as a result of being declared a total loss and fairly extensive detailed inspection before approval before retitling for road use. Those laws might dictate a result.
"Set up a commission of independent 'respected academic experts'" - chosen by Facebook? "who will develop a research agenda to assess how social media impacts democracy, and then review and sign-off on research proposals from other academics."
"Facebook need to ensure that they give researchers access to their system to get informed consent." That's to say, a probably biased data set for which the bias is essentially unknown as to both magnitude and content - an unpromising starting point indeed. It is equally likely that the respected academic experts, and the other academics who will conduct the research, will have their biases. The main difference is that the direciont of academician biases is somewhat more predictable.
Moreover, access to the FaceBook data alone is quite insufficient to answer the key question of whether Cambridge Analytica actually succeeded in swaying votes to a different degree than other methods based, for example, on state provided registration lists and voter surveys such as have been done regularly for at least the last 75 years, or even the experience-informed judgments of candidates and their campaign staffs. At the least, it would be necessary to identify and survey a reasonable sample of voters (i. e., people who actually voted) and nonvoters targeted by CA and not to try to estimate the significance of CA's efforts as to whether they were induced to vote or not and whether Cambridge Analytica's carefully designed and targeted material worked better than other methods. The ongoing moral panic over this and the US election in general probably has poisoned the respondent pool enough to make accurate, reliable, and statistically significant results hard to obtain; as a guess, I suspect that a random sample with a 3% sampling error would break for Clinton by at least 50:44 despite the actual break of about 48:46.
In short, they will arrange to have "social science" research done that will expend a good deal of other peoples' money but is quite unlikely to show anything very significant or useful.
The category "licensed media" is seriously at odds with the first amendment apart from things like FCC licensing of broadcast radio and television as necessary to allocate a somewhat scarce bandwidth resource and impose technical restrictions to prevent interference.
It would be nice if someone would explain how requiring those who have firearms to store them safely and unloaded, and to store the ammunition separately, is going to prevent a spree killing. What prevents an unhinged individual from unstoring the gun and its ammunition, bringing them together, and doing a large scale shooting?
It appears there may be something else going on here for which these sensible rules do not account.
I am a consumer. I do not write assembler code any more. I think it is extremely likely that only OS implementers do that, and then only for certain critical pieces of code where detailed profiling has shown it will significantly benefit performance. Please tell me why I should care whether my C or other compiler level language is executed by an ARM, Intel, Motorola, MIPS, IBM, or other processor.
If the facts of this are as the article states, Intel should replace its HR director with one who will do a thorough reorganization of the department to ensure that in the future it enforces compliance with federal and state employment and civil rights laws. That one of its few legitimate purposes, most of the others having to do with accurate record maintenance and payroll management, which are largely clerical functions that do not justify the high salaries HR staff sometimes receive.
You are not alone in this sentiment. I have wished for that for quite a while, and suspect there are quite a few others. The most recent US example seems to be Ronald Reagan, a third of a century back, although Clinton #1 showed some inclination. For roughly the last 50 years, though, the major parties and their candidates seem to have drunk the Kool-Aid of the pollsters and advertisers to the point where they believe they cannot win election except by pandering to the least informed, selfish, and prejudiced of the electorate.
Candidates of the better class can be found among the Libertarians, Greens, Socialist Workers, Natural Law and others, minor parties all, with no chance of winning a national election, although one or another occasionally wins a state or local office.
The point, which seems quite valid, seems to be that intervention by any foreigners is potentially problematic. It may be illegal (under US law) as well. It is not clear whether work hired out to foreigners is in the same category. I conjecture that the Democrats pursuing "collusion" so enthusiastically do not think hiring Christopher Steele's firm was problematic.
Robert Mercer is a US citizen, with the same constitutional rights as any of the rest of us, despite the fact that he has a lot more money than nearly all of us.
My one vote is indistinguishable from any other, so my psychological profile probably is not worth more than any other. Indeed, because I have moderately well thought out and consistent opinions about public policy and the proper functions and operation of government that probably are quite hard to change, my profile could well be worth less than many others. And as stated that could be as true of the mean Register commentard or Mensa member.
The overwhelming majority, at least in the US, are largely ignorant of nearly everything to do with the Constitution or laws; the authority and operation of federal, state, and local government; the relations among these different layers of government; the formation and execution of public policy; even the names of their elected officials at all levels. Accordingly, their "public opinion" is shaped largely by stereotypes and prejudices and easily molded by the wording of questions into the answer the poll sponsor wants and the polling organization, for business reasons, wants to produce.
Cambridge Analytica, and their critics as well, claim they can use unsolicited information scraped from the likes of FaceBook or Google, together with other publicly available information, to arrive at a better understanding of voter behavior than other available methods. They claim as well that they can use the results of such analysis to design precisely tailored advertising that will be more effective than ads produced using other available methods; and further, that they and similar businesses can use FaceBook and maybe Google and other facilities to deliver the ads precisely to those who will be most responsive and thereby change voting behavior as their clients are paying them to do.
Cambridge Analytica's apparent touting to a potential client of centuries old practices like negative rumor spreading suggests they do not really believe the claims, and neither should the rest of us without far more evidence than the hand waving that so far is offered in its place.
Whether the US government has honored the natural, constitutional, or legal rights in all cases has nothing to do with the case in question or any of the hundreds of other similar cases where the government has lawful possession of iPhones and warrants authorizing them to be searched for specified information. In these cases, a law enforcement officer submitted an affidavit to a court describing the reason the owner was a suspect and the information sought on the target iPhone, and a judge of the court had approved the warrant as the US Constitution requires.
Doing it once certainly would have been the first of many. The FBI and various state prosecutors had several hundred such requests already in the queue. It is not obvious why that is a problem, since they had fourth amendment warrants for them all, including the Farouk iPhone, for which they also had the owner's permission. In fact, Apple had previously been guided toward its action in the Farouk case by a judge in the Eastern district of New York, I think in connection, with a drug case. With a proper warrant in hand, the government is authorized to seize and search for what the warrant specifies, and the subject US person has lost the presumptive right of privacy to that extent.
Reference to RICO and civil forfeiture expansion, which certainly are abused in major ways, are nonetheless not comparable to either the individual warrant approach that law enforcement agencies actually have taken or a hypothetical demand for a general purpose iPhone cracker. There is no legal basis for the FBI to demand such a thing, and Apple could simply refuse to do so if they didn't want to; the courts would not enforce such a demand. Apple also are at liberty under current law make the suggested product changes which, if I understand correctly, would make them unable to comply with such warrants, a plainly valid justification for refusal; the law cannot require the impossible. But when the Farouk and other similar warrants were issued they had not done that, and certainly had the capability to assist the FBI with a single iPhone, repeated as necessary for warrants on other iPhones, all within the limits of the Constitutional restrictions.
"the court order was based on a request from the FIB that they knew full well could only be done in such a way as to be a universal lockpick."
If Apple's code signing process actually is effective, that seems to be a bit of a misdirection. The court order did not require code turnover to the government, or even that the unlocking be done on government premises. It seems likely enough that any of a few thousand people in the world would be able to write and compile the code necessary to do what the subpoena required, and it certainly could be done in a way that would be generic, at least to the combination of iOS version and hardware type. However, under the order, Apple would have had full control of the code, and the presumption at the time was that Apple had to sign the executable code in order to install and run it/
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
