---------------------------------------------------------------------- Message-ID: <20171008175020.19469.qmail@ary.lan> Date: 8 Oct 2017 17:50:20 -0000 From: "John Levine" <johnl@iecc.com> Subject: Re: Google sending balloons to help PR phone service In article <Pine.NEB.4.64.1710072140030.29901@panix5.panix.com> you write: >[Al Jazeera. Live with it] > >Google to use balloons for Puerto Rico phone service > >Alphabet Inc is sending high-altitude balloons to provide phone service to >island devastated by Hurricane Maria. It's also in the Wall Street Journal, so it must be true: https://www.wsj.com/articles/alphabet-gets-approval-for-giant-balloons-to-restore-puerto-ricos-wireless-service-1507409362 Some cynics have pointed out that all the whizbang balloons don't do much good while most of the residents have no power and no way to recharge their phones. R's, John ------------------------------ Message-ID: <20171008233909.GA17536@telecom.csail.mit.edu> Date: Sun, 8 Oct 2017 19:39:09 -0400 From: Bill Horne <bill@horneQRM.net> Subject: Verizon Wireless experiencing network issues Verizon Wireless experiencing network issues causing outages in N. Nevada & California by Kenzie Bales Verizon Wireless is currently experiencing network issues that are causing widespread outages throughout the Northern Nevada and California areas, according to a spokeswoman for the company. More than 100 reports have been filed in the Reno area since 8 a.m. on Saturday, Oct. 7, according to the company's outage site. http://mynews4.com/news/local/verizon-wireless-customers-experiencing-widespread-outage-in-reno-area -- Bill Horne (Remove QRM from my email address to write to me directly) ------------------------------ Message-ID: <ANCdndPjvr1gO0TEnZ2dnUU7-RHNnZ2d@posted.internetamerica> Date: Sat, 07 Oct 2017 23:18:05 -0500 From: gordonb.fox5s@burditt.org (Gordon Burditt) Subject: Re: White House wants to end Social Security numbers as a national ID > White House wants to end Social Security numbers as a national ID > > US government is examining the use of a "modern cryptographic > identifier." > > Rob Joyce, the White House cybersecurity czar, said on Tuesday that > the government should end using the Social Security number as a > national identification method. I'd like to suggest a few objectives for a replacement for a Social Security Number. The number should be long enough and confusing enough that dictating a NewSSN over the phone without error should take more time, on average, than the average lifetime of a person who holds one. (This will hopefully stop scammers from asking for it over the phone, or banks from trying to use it as a default password.) This might mean, for example, a 100,000-character NewSSN consisting of the following base-23 alphabet: The digit 1 Capital I Lower-case i Lower-case l Vertical bar Left bracket Right bracket Capital I with acute accent Capital I with grave accent Lower-case i with acute accent Lower-case i with grave accent Lower-case l with acute accent Lower-case l with grave accent The digit 0 Capital O Capital O with acute accent Capital O with grave accent Capital Q Capital Q with acute accent Capital Q with grave accent Lower-case o Lower-case o with acute accent Lower-case o with grave accent (Someone once wrote a program that generated random Microsoft Product Keys with a similar alphabet, but limited to ASCII, as a joke and complaint about how it was difficult to accurately type them. To Microsoft's credit, they avoided characters that looked alike, and they only required 25 characters, not counting the -'s which you didn't have to type, the form would do that for you.) Note: as far as I know, no existing Unicode character is a capital Q with any kind of accent. Or, you could just dispense with a human-readable representataion of it at all, so asking someone for their NewSSN will get a blank stare after they get the card out and look at it and find no number or bunch of characters. There should be *NO* personal information encoded within the SSN itself, unlike the current SSN which seems to have state of registration (which often implies state of birth) and year of birth within a few years for a fairly good percentage of the numbers. The Social Security numbers of families registering for numbers at the same time should be unrelated (e.g. *NOT* consecutive). Now, this probably applies to immigrants and multiple births only, but back in the 1950's or so when kids started needing one because of laws going into effect, it was not uncommon for all the kids in the family to get SSNs at once, and possibly end up with consecutive SSNs. Also, there should be *NO* changeable personal information encoded, (marital status, weight, current GPS coordinates, firearm license, awake/asleep status, citizenship, etc.) unlike current Medicare claim numbers which consist of the SSN followed by a single letter. T indicates you have Medicare but you are not receiving Social Security (yet). Since people usually enter Medicare at age 65 and the standard retirement age (for getting Social Security) is 66 for people going to retire around 2017, a lot of people will have T for a year and then change to something else a year later when they will start getting Social Security also. The NewSSN card needs to be *READ ONLY* and machine-readable (and preferably NOT human-readable) but it may *NOT* be readable from a distance of more than 0.5 mm (no RFID) from the card. NewSSNs must not be re-used until all previous holders of that number have been dead for at least 100 million years. The chance of guessing a NewSSN (issued in the past, active now, or issuable in the future) by generating random characters in the appropriate alphabet must be less than one in the number of particles in the universe (estimated as 1.e+78 to 1.e+82). If you're using digits as an alphabet, that means at least 82 check digits. The design should avoid dividing the NewSSN into "check digits" and "the real number", where the check digits can be calculated from "the real number". There probably should be several levels of check digits - some public, some classified. The ultimate check is against the database which will indicate whether the number has been issued. NewSSNs should be treated as "private medical information" under HIPAA laws. The minimum damages for a data breach is $100,000 payable by the holder of the data to each owner of the NewSSNs involved, or double actual damages, whichever is higher, plus 1 year of jail time per NewSSN. This amount doubles every 30 days after the first breach until it is paid. So, if you don't admit to the breach for 6 months, that raises the penalty to $6,400,000.00 per number. NewSSNs must not be revealed to Equifax, current Equifax employees, or former Equifax employees who worked for Equifax after Jan 1, 2016. This means that Equifax and its employees or former employees must not have access to THEIR OWN NewSSNs (or NewTINs). Including a NewSSN in a credit report when that credit report was requested using search criteria that didn't include the entire NewSSN is a data breach, even if the recipient of the report is the subject of the NewSSN. Including two NewSSNs in a credit report on a couple when that credit report was requested using search criteria that didn't include both NewSSNs is a data breach by the credit reporting agency, and it may be a data breach by one of the couple against the other if the one whose NewSSN wasn't included in the search criteria wants to press the issue. Being the parent, guardian, or spouse of someone is *NOT* a defense against giving out their NewSSN without their permission. ------------------------------ Message-ID: <barmar-D4462F.14110208102017@reader.eternal-september.org> Date: Sun, 08 Oct 2017 14:11:32 -0400 From: Barry Margolin <barmar@alum.mit.edu> Subject: Re: Verizon to discontinue legacy services across seven- state area In article <513952e1-8ec0-43f3-ba45-068b5a0d2936@googlegroups.com>, HAncock4 <withheld@invalid.telecom-digest.org> wrote: > On Thursday, October 5, 2017 at 8:53:58 PM UTC-4, Bill Horne wrote: > > U.S. telecom behemoth, Verizon Communications Inc VZ is reportedly > > seeking permission from the U.S. telecom regulator Federal > > Communications Commission (FCC) to discontinue four legacy interstate > > DS0 services across parts of seven states. These legacy voice and > > low-speed data services are Voice Grade Service, WATS Access Line > > Service, Digital Data Service and DIGIPATH Digital Service II. > > > > The affected states include Delaware, Maryland, New England, New > > Jersey, New York, Pennsylvania and Virginia. Verizon has about 10 > > wholesale customers and approximately 67 retail customers for these > > services in the affected areas. > > > > http://www.nasdaq.com/article/verizon-vz-to-shut-down-legacy-voice-services-in-7-states-cm855014 > > > > What will happen to people who don't fibre service to their home? > Lots of places do not have it. Will they be forced to go over > to Comcast? In my area at least, Comcast is notoriously unreliable. This isn't talking about regular consumer phone services. Did you miss where it says there are only 67 retail customers for these services? -- Barry Margolin, barmar@alum.mit.edu Arlington, MA *** PLEASE post questions in newsgroups, not directly to me *** ------------------------------ ********************************************* End of telecom Digest Mon, 09 Oct 2017