The obvious replacement of the OCR by humans seems likely to bring a positive error rate as well, quite possibly in the same range as the OCR system. It is not even unlikely that the human error rate would be larger as they get fatigued, unlike the OCR software.
1581 posts • joined 16 Jan 2011
Re: And when it seriously goes wrong?
Those who bothered to follow the link (and the further link to The Intercept) and actually read either article will have found that Talley's ex-wife, three acquaintances, and the teller he was accused of robbing in the second case identified him, presumably based on bank surveillance photographs. They also would have noticed that the FBI facial recognition specialist, based on personal examination of photographs found a likely match between the surveillance photos and others probably taken under better conditions. Those certainly would have been enough to justify an arrest, and quite possibly and indictment and trial. The fact that the only witness changed and augmented her statements during trial testimony does not alter that.
Talley may well have a good case against the arresting police for grossly excessive force, and possibly also for later procedural errors. He might also have cause for action against his public defender in the first case for slackness in checking with his employer as to his stated alibi, which brought him a two month jail stay.
What the articles do not do is implicate machine facial recognition, which was not used. Indeed, the Intercept article suggests, with reservations, that it is likely to improve the results that can be obtained by human analysts alone.
Re: Only half?
I wonder, if the US and UK are not "even remotely democratic" which nation states are. That is a fairly strong statement that runs contrary to conventional understanding and warrants a bit of substantiating evidence.
I stipulate that the US federal government is not a democracy, and was not intended to be; the authors of the Constitution took considerable care to prevent that, for reasons discussed at some length in the Federalist Papers. They did not, however, constrain state government in any way to prevent them from establishing democratic regimes of their own.
Re: Trump supporter
Corrupt != criminal:
having or showing a willingness to act dishonestly in return for money or personal gain.
"unscrupulous logging companies assisted by corrupt officials"
synonyms: dishonest, unscrupulous, dishonorable, unprincipled, unethical, amoral, untrustworthy, venal, underhanded, double-dealing, fraudulent, bribable, criminal, illegal, unlawful, nefarious; informalcrooked, shady, dirty, sleazy
"a corrupt official"
Whether Wikileaks, or Assange, committed a US crime would depend a lot on the exact text of any exchanges between them and Manning, as well as who initiated any exchange between them. So far, the only evidence for US extradition efforts seems to come from Julian Assange, who is unlikely to have reliable information. In any case, once the transfer was completed, it is unlikely that the publishing could be prosecuted successfully.
It may be worth mentioning that while whoever obtained the emails from the source probably* committed a crime under US Law, it is all but certain that publishing in the US is fully protected by the first amendment and did not violate any US laws**. In addition to the fact that cutting Assange's internet service will have no affect on Wikileaks' ability to continue stirring the pot, they have not violated the law unless they obtained the messages from the DNC and other servers themselves.
In the same way, they did not violate US law by publishing the material Bradley Manning gave them (despite the fact that Manning violated a number of laws), somewhat undermining Assange's claim of a secret plan to extradite him for US trial. That is possible, but it would be necessary to prove that he conspired in some way with Manning to break the laws, something I suspect he, and others associated with Wikileaks, would have been quite careful to avoid..
* Unless they were turned over by the author or one of the addressees.
** Except possibly those dealing with criminal libel or which on examination will be found inconsistent with the first amendment..
Re: Beware "Enable remote management" checkbox
Access to a network is not the same as control of any machine in it. Control of one or more of the networked systems, including the router, is the intruder's next logical step. Making it more difficult is a wise move.
From Agazzini's public announcement:
20/08/2016 - First communication sent to IBM PSIRT (psirt at us.ibm.com)
22/08/2016 - IBM Response, PSIRT Advisory 6345 assigned to the bug
05/10/2016 - Communication from IBM with fix information (PI62375)
07/10/2016 - Security Advisory released
Copyright (c) 2016 @ Mediaservice.net Srl. All rights reserved."
Maurizio Agazzini CISSP, CSSLP, OPST"
So IBM received notice on August 20, developed and tested a correction by October 5, and released the advisory on October 7, whereupon Agazzini immediately announced the details to the world.
Many shops have a regular patch cycle that varies in length, but would be unlikely to be less than a couple of weeks except for tiny organizations or very easily exploited patches with very high impact. Most have internal requirements for testing, even of security changes, and a patch cycle of a month probably is fairly common. Publicly releasing details of a major commercial product vulnerability on the same day that the fix is released falls well short of my idea of responsibility unless the vulnerability already is known and being exploited or a trivial mitigation can be applied until the full correction can be tested and installed.
There may be mitigating circumstances like vendor foot dragging, but this case does not show it. IBM moved from notification to correction release in seven weeks, which is not necessarily unreasonable.
Oracle's databases don't pollute
That depends on your definition of "pollute," which in an Oracle context is spelled "PL/SQL."
I made a pitch a few years back for PostGres as a partial substitute for Oracle. We had the rather bad habit of making a new database for nearly every application and PostGres would have done for 90 or 95% of the databases we were running or planning, but for our other nasty habit of using PL/SQL for a lot of the application code. The savings from replacing Oracle would have been large, probably enough to recover in a couple of years the cost of translating the PL/SQL to PL/pgSQL or redoing the necessary code in another language, even taking account of the need to reanalyze some and retest it all. But it was a government agency with no tolerance for risk, where C-level management had come through a couple of years earlier and in "all-hands" meetings had informed IT personnel that IT was not a part of their core (finance and accounting) business. Rejection was almost instantaneous despite the fact that the agency was under considerable pressure to reduce operating costs.
Re: Alternative solution
It still is Nexedi's choice. They can comply with the same rules as other developers and have their apps in the iStore, or not.
The lawsuit is rent seeking at its most obvious.
I hesitated between the downvote and the upvote I finally registered, mainly because the problem is a bit more complex and involves more slackness by the Senate and House of Representatives than the post suggests.
The President appoints judges and justices with the advice and consent of the Senate, which does not have to approve the nominee. The Senate and House together have passed laws that delivered a lot of power to the executive branch, irrespective of the President who heads it, and the presidents have welcomed it and run with it. The Congress does not have to do that (on paper) but over time have allowed the federal government to take on so many things that their failure to act creates deafening uproar and great indignation that uniformly has caused them to back down. Worse, they have passed laws that delegate to the executive branch legislative powers that, if they were doing their jobs, they would have guarded jealously. They even have allowed, and funded wars, for most of the last 70 years without the constitutionally required declaration of war. As a group they are feckless and spineless.
The expansive responsibility and power of the federal government does not go unchallenged, and the ensuing litigation dramatically increases the importance of judges at all levels, but most importantly the Supreme Court, and requires increasing politicization of judicial appointments to ensure "correct" decisions. It also explains much about why lefties are so fearful of a President Trump with the capability and presumable willingness to use that power "wrongly" and the Trump supporters are similarly fearful of a Clinton win. The truth is that if either one of them is elected, we are in deep trouble.
My neighbor has the short version on a yard sign:
Re: Coincidence or something serously wrong here...
Have an upvote for general correctness. However, I will make a few additional observations.
Federal government agencies like NSA are constrained by both staffing limits and general schedule compensation limits. The first may keep them from hiring enough people to accomplish their mission, and the second may reduce their ability to hire enough people in technical specialties that are especially well compensated in the private sector. It is very likely that both of these constraints bear on the NSA, as well as some other agencies; the National Geospatial-Intelligence Agency seems likely to be another example.
This situation is not the fault of the Obama administration alone. The Congress, in its sometimes misguided effort to (appear to) reduce the size of the federal government, is fully complicit. Furthermore, it certainly extends at least two administrations back, to the Clinton administration, and probably back further to that of Reagan. Democrats and Republicans are about equally culpable.
The solution is, and always has been, to increase appropriations without increasing staffing limits, ignoring the fact that, as another poster noted, that it increases the cost of federal government operations quite substantially. As this post notes, it also allows agencies to hire people to critical positions at rates above what the general schedule will allow. According to his claims, Edward Snowden was paid far more than the GS-12 or 13 rate that would be the range for his job, based on his known CV. More often, however, contractor employees receive less, and sometimes much less, than the GS rate for their positions. The contractor firm bills substantially more than they pay the employee (they are in business to make money, and actually incur expenses for management, payroll, and sometimes fringe benefits). It is not uncommon for them to bill more than the fully burdened cost of a civil service employee in a position.
I've worked on both sides. As a manager, I found the ease of filling vacancies under an existing contract extremely helpful, but the low rate paid in some cases was quite distasteful, and I encouraged contractor employees to apply for civil service vacancies that came up. The contractor employees we got, though, were as good on average as the civil servants. Later, as a contractor employee (under the same contract) I was paid on a par with my civil service counterparts. My employer was a subcontractor to the primary, whose contract limited them to a rate that may well not have allowed both them and my employer a profit. That probably did not bother either of them, as they supplied quite a few others at lower rates and could make their contracts profitable as a whole.
Re: does size matter?
This comes close to saying that a company that becomes successful to the degree that its product exceeds 50% (or some other, possibly larger market penetration) is, for that reason, to be subjected to a different set of rules, better described as constraints, to inhibit their growth. That to prevent them from abusing their dominance.
Without further qualification, that condenses down to "punish market success." I suspect EU competition law is intended to be a bit more nuanced, perhaps to requiring that the market dominance actually be exploited in a way that constitutes abuse. Has Google actually done that, say by first giving away the Android OS and then, once it has gained dominance, changing the contracts to require the App store and other pieces at issue? Or did they, beginning when they first entered the smart phone OS market, offer a bundle of the OS, App store, etc., preconfigured and immutable by contract, to phone manufacturers and carriers? And on the basis of that did they then become dominant without changing the bundle other than upgrading components? Or was it, as seems likely, somewhere between, with new components added, arguably to foreclose competition.
In the first two cases, Google probably have done nothing either wrong or illegal; they simply offered a product that helped the manufacturers and carriers succeed in collectively dominating the market for smart phones. They may have been somewhat high-handed, but they did not, after all have to offer the product at all. The third case is more interesting, arguably wrong, quite possibly illegal, and highly beneficial to the attorney class.
Re: You can not be more wrong
Nothing about "open source" (even the GPL) requires that those who use it distribute what they use it for or any code that they modify, or that they develop based on the open source. The fact that some of them may not share their modifications or extensions to open source code does not refute the claim that they have built infrastructure or applications on it.
I was trying to maintain civility in hope (dashed) that they could provide a decent workaround for accessing the new browser version of their scheduling and TV management application which, as of Tuesday, September 27, depends on installing the current version of Flash. The prior version did not require flash at all. I am more than a little unhappy with them and would drop their service in a heartbeat if there were a decent alternative.
This estimate clearly must be in error. I know this because a Comcast CSR told me just yesterday with a perfectly straight face that the STB the rent me for $20 a month costs them $1600. The notion that they could find a lower cost supplier of a box with the capacity of a couple of Raspberry Pi s, a terabyte disk and an OTS tuner, together with a hand held controller with the capacity of another Raspberry Pi, more or less, and a handful of control buttons, apparently escaped them completly.
$100 probably isn't quite enough, but $200 - 250 should be possible for a large run and allow for the power brick and decent profit. Two sanity checks: first, their assertion that they will provide a piece of computer equipment at a rent so small as to take nearly 7 years to recover the cost, and second, that the power brick for the STB is rated at 36 Watts.
Re: I can understand a little bit of bias
There should be no discrimination as such between "natural born" US citizens and naturalized citizens however recent. A citizen is a citizen. If that is a requirement for hire, as often is the case where defense or national security is involved it should be enough at the first pass. If the applicant is seeking a position that requres a security clearance, any relevant questions about when an applicant became a citizen, or how, can be answered connection with necessary background investigation.
Re: An interesting legal question...
It is legal in the US for the US government to conduct such activities in other countries with which the US government does not have treaties that govern them; otherwise not.
It is legal in the US for a foreign government to conduct such activities in the US if a treaty approved by the US Senate authorizes them; otherwise not.
I am not aware of any treaties that allow such activities in the US by any other government (or, for that matter, any laws that would allow it by either the government or private sector actors. The hack was illegal whether done by a foreign government, foreigners, or US residents. Blaming it on a "state actor" is misdirection that one supposes is intended to increase the scariness and reduce Yahoo!'s perceived culpability in the matter.
Re: Dar Yahoo Customer:
And roughly $320 Million (at 30% of the total settlement) to be shared out among the plaintiff class attorneys.
Re: It is all in a name...
An autopilot would not have to be all that good to be better (and safer) than a great many human drivers I have observed over 50 years or so. Come to think of it, I can recall some occasions when I have to admit that a relatively simple machine might have driven more safely than I.
I think we can assume that Skorobogatov was not represented in the market as seen by US federal agencies, and that those who were charged a good deal. $1MM still seems high, but four months of part time work clearly understated the overall cost to him, as it skips by the fact that as a senior research associate he undoubtedly had considerable relevant background knowledge before starting. And in any case, four months of part time work for hire by anyone certainly would be much costlier than the $100 stated in the article as the cost of the hardware Skorobogatov used.
The drift of the article seems to be that the cost of developing the attack, which evidently took Skorobogatov quite a few man hours of what seems to be highly skilled analysis and electronic technician work should be ignored because the result can be replicated for a small amount going forward. That is somewhat like saying the design, development, engineering, and testing investment in a SOC should be ignored when setting a sale price for the end product, even if the projected demand is for only a few thousand units.
Re: Just a "clerical function"
Ignoring the clerical class can be and often is a great mistake. Josef Djugashvili, for example, later and better know as Stalin, arranged his rise to the top of the USSR hierarchy from his position as General Secretary of the Central Committee of Communist Party of the USSR.
Re: ...expose customer data
Like many issues, this one actually is not plain and simple, and in many locations not a monopoly either. Where I live, everyone has the option of either Comcast or CenturyLink, and a large and growing number not far away have the additional option of Google Fiber for cable TV. Nearly everyone also has good visibility of a number of OTA transmitting towers and could access broadcast material for the one time $20-30 cost of an antenna. Finally, anyone with broadband internet service from one of the above cable providers can access Netflix, Amazon, Hulu, or Acorn at fairly reasonable rates.
Have an upvote, and thanks for pointing me to the standard that justifies my habit for quite a few years.
Maybe because some of the national security collection requests (roughly 1/3 of the total) arise out of US referrals?
I am a bit skeptical about this report. It is my impression that auxiliary power ports in automotive vehicles provided power only with the ignition on or in the auxiliary position, and I confirmed that toe be true of my two - a Honda and a Toyota. The Jeep Grand Cherokee might be different, of course, but if not it appears we are asked to believe the owner left the key in the ignition lock and the car running or on auxiliary. That seems a bit unusual, and very likely to risk theft of the vehicle.
Re: standard operating procedures
Judging by the easily available pictures at
It seem plausible that the tesla's crash speed was considerably greater than the maximum speed attainable by a Prius or similar car.
Re: If this wasn't a Tesla, it wouldn't be a story
Bursting into flames is fairly common among automobiles that crash into solid objects at speeds that, in this case, appear likely to be over 100 mi/hr - irrespective of their energy source. This is news notable only because it is a Tesla and the attendant possibility the driver was going at such a speed "no hands."
I've been out of touch with federal data security standards for a few years, but as of 2011, BleachBit would not have met the standard for handling disks containing Sensitive but Uncalssified data such as Social Security Number, let alone disks that ever had held any data with a secret or higher classification. The agency for which I worked handled no data classified higher than SBU, but the disposal requirement was degaussing followed by physical destruction (the agency had a shredder for the purpose. This despite the hefty additional charge for failure to return dead disks to the vendor, which we mitigated by purchasing new disks at retail to hand the CE instead of the failed disk. I an fairly sure the standard has not been relaxed.
Also heavily implicated are both Security and IT staff at State. The FBI report makes it quite clear that they knew pretty well what was going on and did nothing to stop it and forcefully discouraged those who questioned it (described more fully in the State IG report a few months back). It also makes clear the general sloppiness at State, and to a somewhat smaller degree at some other government agencies, in handling classified material. The DoD component that employed me for a number of years handled only sensitive-but-unclassified data (Personally Identifiable Information) yet was far better by about 2004 than State five full years later, both in technical protection and employee behavior. They clearly could use a good purge, although if, as many suppose, Clinton is elected President we might expect to see distribution of performance awards instead.
Re: FBI records
My cursory and as yet incomplete read of the FBI report has it that (a) she (through her minion) applied for the clintonemail.com domain about the time the Senate was interviewing her for the position, and (b) she did, in fact, change her email address to use the new domain very shortly after confirmation to the position. Couple that with the fact that there aren't a lot of email addresses that would be cooler than, e. g., firstname.lastname@example.org, "cool address," "used for a while and don't want to give it up,'" or "didn't want to change email addresses" simply won't do.
Add the fact that the servers were seriously non-compliant with longstanding federal law, and with FIPS and State Department standards; apparently were quite insecure in their configuration; and were known within the first two years to hackers and probably foreign intelligence services, and you get quite a mess.
i put it down to a sense of personal entitlement, combined with a disturbing casualness about following established laws and rules, that we should be very leery of in choosing a President, even if, should she be elected, she chooses "email@example.com" for her email address.
It ain't what you don't know that gets you into trouble, but what you know that ain't so. (Attributed, probably wrongly, to Samuel Clemens, AKA Mark Twain).
For various reasons, US internet service may be poor in places, but arguably does not rate that badly. US average connect speed (12.6 Mbit/sec) ranks about 14, roughly midway between Germany and Denmark; peak speed (,57.3 Mbit/sec) ranks about 16; and for connections better than 15 Mbits/second (24%) the rank is about 15 (full confession: based on Akamai reports for the third quarter of 2015 as reported by Wikipedia). That puts the US generally in the same group as much of Europe + Canada, and well ahead (for example) of Australia, New Zealand, and France. It could be better, and in many places is, but certainly does not qualify as "dismal" (or even abysmal) as it is.
The state of internet service in the US also is not a result of market failure. In many areas, possibly most, it results from lack of or severe restriction of the potential market, as Google's entry into several areas has shown by inducing existing providers like Comcast and at&t to improve service and reduce prices in some combination.
My own experience, first with Cox and lately with Comcast, has been of increasing speed over time (measured courtesy of ookla) at nearly constant price over a period of about ten years. The outlier was at&t, which failed to match Cox (as promised) albeit at a lower price. This, of course, is to be expected as a result of normal equipment replacement where the new equipment is inherently better than what is replaced due to general technology improvement.
Re: Michael Brown
Michael Brown did not commit an armed robbery.
I do not know what police patrol practices are in "every other country." In the US, use of two officer patrol cars fell into disuse beginning around 50 years ago because an increased number of cars with one officer each provides much better coverage at constant cost or similar coverage at reduced cost. The rules of engagement do not prohibit an officer from making a stop, but officers normally will report they are doing so (as I believe officer Wilson did in the Michael Brown incident) and nearby officers will converge on the scene.
Before routine use of patrol cars, foot patrol often was done by solitary officers, but those were different times.
Re: Michael Brown
Michael Brown was not shot "while" or "for" running away after committing a crime. He was shot after assaulting a police officer who had stopped him while he was running away. When shot, he was advancing toward the officer.
The officer involved probably made a tactical procedural error during the stop, in putting himself in a position of some disadvantage in opposition to a larger and considerably heavier opponent, but Michael Brown made larger errors in assaulting and later advancing on an armed police officer.
Re: Who in their right mind ...
Nearly everyone in the southeastern suburbs of Salt Lake City, for one example. My present alternatives include Comcast (high price, pretty good speed) and CenturyLink (lower price, max available speed "up to" 20 Mbits/sec). Goole have started turning on fiber in SLC proper, but never got around to stating plans for the suburbs.
As an aside, there was a multicity consortium, UTOPIA, which pretty much stalled 5 or more years ago before reaching many of its potential customers. So much for government doing the job.
Re: Why bother?
You cannot (realistically) ban mathematics. If you are an effective national government, however, you probably have the power, and possibly the authority, to regulate the legal use of cryptographic systems, and to fine, jail, or otherwise punish those under your jurisdiction who decline to follow the laws. That goes quite a bit beyond "make difficult."
1. Real evidence of actual abuse by governments is pretty thin in most countries with regimes that are generally regarded as liberal democratic (small l and small d) unless having the capability to abuse is taken to be equivalent to abuse itself. Indeed, that probably is true even under most regimes commonly thought of as oppressive, although the range of behaviors such governments ignore may be quite limited.
2. Security of such data clearly is a risk, but one that admits mitigation. Various key escrow arrangements that have been suggested included provisions intended to reduce the risk and increase the difficulty and cost of escrow database compromise. Risk never is zero, and all one reasonably can require is that it be quantifiable and small enough.
3, 4. There is no real basis to argue that key escrow would make encryption more difficult or less convenient, as collection, indexing, and storage necessarily would be automated. It could present additional points of failure (or not) depending on whether failure to escrow would cause failure of the basic communication. Communication for commercial transactions may not be an important issue, as for legal trade it often will be possible to obtain the details from at least one of the participants by a suitable court order.
5. The obvious answer would be to provide the escrowed key, as national laws may require, to the government of the originator and recipients. In many or most cases, that would be at most one, since most communications do not cross national boundaries. That obviously would present issues, but for most people and organizations they would not necessarily be overly serious. Those who wish to shield activities from any of the governments demanding escrowed keys would have the most reason for concern, followed by those with reason for concern about security of one of the repositories against criminals or competitors. Increasing the number of repositories clearly would increase risk of control loss, however.
6. The customary government approach to refusal to participate would be criminalization, with a combination of detection procedures and penalties sufficient to discourage it.
The point of the original post was not to argue for key escrow, which has very little to recommend it, but to note that it would not be less private than plain text communication and might not add a great deal of risk, for most people, most of the time, compared to encrypted communication without escrow. Other approaches to law enforcement access include enforcing backdoored encryption systems, probably a much worse choice, and judicial warrants demanding delivery of the decrypted message by the originator or a recipient, depending on details of jurisdiction and treaty arrangements, with punishment for noncompliance.
Re: Fight them at every turn!
In the US, the USPS has collected metadata for all first class mail for over a decade, following the anthrax letters that killed a few. That metadata, to be sure, is not nearly as reliable as communication data, since only the destination address information is functional.
I call BS on the often stated "It's maths, stupid" and "It's magical thinking" themes.
Encryption, in today's customary usage, certainly is based on mathematics, but that is largely beside the point. A completely trivial key escrow system in which a communication metadata and session and key are deposited with a government custodian is more secure than communication in the clear despite being subject to the same kinds of vulnerabilities, and clearly would meet the stated need of law enforcement authorities. Nothing about this represents magical thinking, and it does not depend on a weak encryption system. Volume is a potential problem, but there is good reason to think that national signals intelligence agencies have developed effective ways to deal with it.
The fundamental problem is one of lack of trust combined with arguably excessive government authority, or at least power. Many people believe that law enforcement officers and agencies spy on nearly everyone without any particular reason, and do not trust them. And in most countries there is evidence of some government misbehavior. However, such misbehavior is not new and almost certainly would not be made simpler or easier by even a trivial or badly designed key escrow system. In most countries, too, those who are law enforcement targets are likely to be surveilled, and if important enough, prosecuted, sometimes irrespective of guilt. The number of laws on the books offers plenty of options for prosecutors. Use of encryption that the authorities cannot break, if legal, might delay the outcome but would do little to prevent it; and if illegal it could be a useful substitute charge leading to an easy conviction.
Re: You think _that's_ bad?
The maximum - and minimum - stored password size should be the fixed length output size of whatever cryptographic hash function is used on the salted concatenation of user name and password.
Re: Microsoft or Americans?
Many of those in New York City and adjacent parts of Southern New England tend to forget there is a world west of the Adirondacks, or regard such parts as may possibly exist as probably uncivilized.
Re: English is wonderful
There are rather a lot of people on the US West coast of Japanese or Chinese ancestry or origin, and not a few of them likely enough are employed by Microsoft.
Somehow, it does not seem this should have happened.
Re: At this point, not a 'snooping' flaw
It might also be worth observing that the flaw described in CVE-2016-5696 was introduced to correct or mitigate a previously existing, and perhaps much more serious, vulnerability.
Also worth noting is that the probability that any large piece of software contains no errors is operationally equal to zero. This vulnerability, like large numbers of earlier ones, will be mitigated or eliminated, and others will be found, and some of them will have been introduced in the correction process.
Re: Passwords need to be rethought
The requirement to change passwords periodically (every 60 days when I left government service) has less to do with crackability and much to do with limiting exposure time if either user passwords or the hashed password file is compromised.
Microsoft has open-sourced PowerShell for Linux, Macs. Repeat, Microsoft has open-sourced PowerShell
Re: Why is ssh built in?
As an additional note on the clunky A:, B: and the like, I seem to recall that an intermediate version of MS-DOS (3.2, I think) had a built-in command, join, that enabled the user to do the rough equivalent of the Unix mount command, with the same beneficial result of being able to treat all the disk resources as a single directory tree. MS lost me when they removed it, and a number of other useful items, in the standard part of the next version and made it a $60 or so additional utilities package
I switched right then to Xenix, which I picked up used at an amateur radio swap meet, and never after paid more than the minimum necessary attention to MS operating systems. I did note that Windows hid the clunkiness rather effectively, and disks grew quite rapidly, so that for many it made little difference.
at&t probably is in a formally competitive market almost everywhere it does business. It is either the phone company (e. g., in most of Ohio) or a late comer trying with less than perfect success to compete with the local cable franchise for both data and telephone service. Cox, where I used to live in Ohio, had an exclusive local franchise (which at&t tried to break up) that provided faster data service at every speed they both offered (at&t was unable to compete at the highest available rates).
Comcast, for the present, is the incumbent in my area southeast of Salt Lake City, but has a competitor (CenturyLink), which definitely is less expensive, at an advertised price of $20 a month for up to 12 Mb/s; Comcast's rate starts at $30 for up to 10 megabits, and may be losing a few customers over this. But Comcast also offers other capacities up to 250 megabits at $70 a month, which I typically measure. My "up to 150 Mb/s" service often yields better than that and typically measures around 130 Mb/s at the inside of the router attached to the modem. For all the whining about Comcast's poor service, here, I can say only that the service has not been quite as reliable as Cox's was in Ohio, but the unscheduled down time over three years certainly has been under 0.05%.
So Comcast, here at least, is not quite a monopoly, but we certainly are looking forward to the benefits of competition from Google, which has started its build out not too far away.
Re: Google are switching to OpenJDK...
Would the OpenJDK API not be essentially identical to that of Oracle's JDK? I don't do Java programming and don't know the answer, but it seems possibly relevant. If the API is the same (or near enough) could Oracle reasonably claim infringement by Google without also claiming infringement by everyone else using OpenJDK? Or would the fact that they have not (and maybe because of licensing cannot) claim infringement by OpenJDK developers and maintainers invalidate any claims they might be making?
Those who don't know the email metadata are in the clear are a danger to themselves and others, and should be kept under tight supervision in any security context.