Posts by John Sager 806 publicly visible posts • joined 28 Apr 2008
Having read the judgement, it looks very much like Judge Alsup has given Oracle very little wiggle room to argue on when/if they appeal this. They can only now appeal on the issue of judgement as to law versus a jury verdict on the facts, and Alsup has taken great pains to explain why it's all down to issues of fact that a jury has to decide rather than a plain direction by the law.
I'm still surprised that the appeals court threw out the verdict that the API was not copyrightable. We've all assumed that APIs were free to use and this was a bit of a shock. Perhaps now APIs should come with a 'free to use' licence and those that don't (unless in very specialised areas), fall by the wayside.
I'm in the 'sod them' camp. They've had plenty of time to register to vote, and there is a legally defined registration limit whatever the quality of the online registration process. I'm a bit more sympathetic to voting after the 10pm limit on polling day, if you have already joined the queue before 10pm.
Radio spectrum at the useful frequencies is a finite resource, even with some of the clever coding and MIMO stuff that's in development. Within a fibre the available spectrum is orders of magnitude greater, so in the longer term, as data rate requirements grow, fibre will be essential, not just in the backbone as it mostly is now.
It would be better if the comms infrastructure were classified more like roads & rail from a state aid perspective. I hesitate so suggest the network should be in public ownership as we would be likely back to the old days of "you'll get what you're given", but if the govt can spend all that taxpayers cash on HS2 without issues with Brussels, then why not also on a fibre broadband infrastructure at the network edges?
Incidentally, where I live in a country village, I really do have to have what I'm given - no-one other than BT would look at provision locally and it's looking increasingly unlikely that we'll get a fibre cab close by, even with BDUK cash. Gigaclear would probably not consider it either, having looked at their website & the conditions they require.
Cisco 1801s are fairly readily available on eBay. Alternatively, roll your own with an embedded Linux distro - e.g. LEAF/Bering - or go BSD with pfsense, on various hardware platforms. Of course all these have a learning curve - none of them are really plug 'n play. If you already know Cisco IOS then an 1801 is easy. Pfsense and some of the specific router/firewall Linux distros are fairly easy to configure too though you need to choose a hardware platform.
Well, he's a politician, and like the overwhelming majority of that breed, he knows eff all about nearly everything. You would think that ministers would do a bit of homework on their brief. I suspect, though, that none of the civil servants from whom he would deign to gather advice know anything about the subject either.
I would be seriously pissed off if anyone seriously offered me Internet over VSAT as a service. I might be glad of it in the middle of Mali or Botswana but not anywhere in the UK.
Well, the single track working problem was solved well over a century ago with the token system. Originally physical, with end-to-end interlocks between the token dispensers in the signal boxes, it's now often done electronically. No reason why they can't do that over satellites. That's what makes me so surprised about that head-on crash on a single track in Germany a few weeks ago.
Yup. I've got a 2010 model Panny TV, and it recently stopped working on ITV-HD via satellite when ITV changed a parameter of the satellite signal (still within the DVB spec). I've had e-mail discussions with P about a s/w update but the probability of that is infinitesimally greater than zero. P stopped updating the s/w about a year after I bought it, and virtually all the 'smart TV' services on there at the time are now gone. My next TV won't be Panasonic!
I've recently bought a Samsung Blu-Ray player and that phones home to Korea all the time to find out what to do. It works a treat now but I wonder how long for.
A letterbox sized screen is not ideal for reading books which, lets face it are more suited to a 4:3 screen ratio
Well, the first bit is true, but all the books I usually read have a portrait aspect ratio. The usual small size paperbacks in the UK are 198x128mm - an aspect ratio of ~3:2 in portrait. So you just turn the tablet on its side?
As for Nook, I bought one of these in preference to a Kindle as it could read a wider range of e-book file formats. I suppose I half expected some problems further down the line - the history of DRM content is littered with content lock-outs due to businesses going bust or getting out of the market. Perhaps B&N don't want to pay the publishers to licence content for the UK. I shall be royally pissed off though if any of the books I currently have get wiped by Sainsbury's because they haven't licensed it.
As always. Just to be sexist for a minute, is the lack of women in tech down to us cavemen putting them off, or are they just not interested & enthused by the subject?
I was down at the computing museum at BP last year, looking at the Harwell Dekatron computer. There was a school party there with quite a few girls, and the presenter, as a form of encouragement, was showing them pics of the early days with quite a few women in the team. In those days a lot of programming was writing code on sheets marked off into boxes with one character per box so that punch card operators could punch it up into a card deck. I did wonder whether the women were designing the algorithms or just writing the code on the sheets. In those days, 'writing code on sheets' was seen as a perfectly reasonable job for a woman, and they may well have seen it that way too, as just another sort of clerical function. These days that just doesn't wash, so girls generally don't want to do that stuff, and I wonder just how many of them really are enthused by the process of algorithm design as a prelude to the process of writing code.
The problem is that the UK government is also an unreformed wholly rotten organisation
At least it's our unreformed wholly rotten organisation, so we've got slightly more chance of reforming it than we have of reforming that thing in Brussels.
Hmm. A Windows-based router doesn't fill me with a lot of confidence but I see where you are coming from. However, although Microsoft's security processes are now quite good (though they misuse it regularly for other purposes), it took them a long time and a lot of mis-steps in the past to get there. It's also not cheap for them to manage, but only a small cost now compared with their revenue.
The same doesn't apply in the router market. One could argue that the fact that the big ISPs bundle a router with the product militates against good router security, as the ISPs demand a 'just good enough' product at a rock-bottom price. So the other manufacturers have to follow the race to the bottom to compete. Of the router mfrs, only the big iron guys like Cisco could support a MS-style security wrap and Cisco aren't really in the consumer market.
The later BT home hubs seem to have a good customer-based security wrap - a little slide-in card in the back with random Wifi and admin passwords. Let's hope the internal security config is as well thought out.
As the article states, the key to the data (128 bits, 256 bits?) is buried in the CPU, and the CPU will only use it itself to decrypt data on presentation of a valid passcode. So even though the flash memory could be cloned, that is useless without the key, which stays buried in the CPU at all times. So you need both the memory and that particular CPU running valid code to be able to get at the data.
The electromagnetic field modelling package NEC2 is a bit like this, though one part of the user manual does try to explain things. It uses magic numbers, particularly a high value (10,000) added to the segment number. Guess what happens when you have more than 9999 segments in a model...
Of course the code was written in the days when a 9999 segment model would a) take far more memory than the computers in those days could handle and b) take longer than the age of the universe to run.
Nowadays it's still used a lot by radio amateurs to model antennas so we run into these limits (amongst others in this package). I have hacked it to model up to 30k segments on Linux, but beyond that it runs into yet another, more fundamental limit.
You mean you actually accepted those meeting invites?
They weren't optional, and in those days it was all face-to-face. I was just a peon team leader at the time. Thankfully that project just kinda wasted away & I moved onto some more interesting stuff with a better management ethos (stayed out of my hair!).
The project manager on a big project I was working on used to do that. OK I suppose if you retire to the pub after, but that was not my scene. Friday evening in the pub is for a nice wind-down with some friends after tea at home with the family, not a work colleagues piss-up.
I doubt they would open the envelope in front of the world, and with the paper already printed in Nature, just to find "that was a test". I suspect the envelope-opening ceremony happened in private at a LIGO meeting some weeks ago, given the time that rumours have been circulating.
At the time he wrote that, the Internet was very much smaller & most people on it were geeks of one sort or another. So it was not exactly a dumb statement then. However AOL was connected about that time & the 'net started a long descent to what we see today, though there have been compensations along the way (Altavista & descendants, http, etc). PGP has always required some intelligent deployment. Enigmail might be just a plug-in but the real work is setting up & managing the public key infrastructure required to use it effectively as a day-to-day tool. Amongst a small circle of friends, acquaintances & colleagues that is manageable, but otherwise, forget it. And although there is now a halfway decent CA infrastructure for website certificates, that's still too hard to deploy universally for personal e-mail signing & encryption.
So non-TLS encryption is going to stick out like a sore thumb for a long long time, even TLS used in unusual contexts (not web, not IMAP etc).
Have Apple actually done wrong here? They were offered a deal by the Belgian govt, which the EU have now said is illegal state aid. So if anything it's the Belgian govt which is in the wrong. Apple can easily say they accepted the offer in good faith. No doubt the devil is in the details though.
http://www.aurora-service.eu/aurora-forecast/
According to that site it all happened last night (around midnight weds & the early hours thurs). Anyone see it? Lots of rain here & we're too far south for most of them. In the 30-odd years I've lived here (S Suffolk), I've only seen three.
My phone (Nexus 5, Snapdragon 800 SoC) will pick up both GPS and GLONASS sats, but I've only ever seen it use two GLONASS in the solution. I don't think the Snapdragon 800 will do Galileo, though I think the 810 will. Has anyone ever seen a Galileo in the sat list that some of the satnav phone apps have?
The sample of comments in the Crappygraph's article, plus another demolition job on it in Techdirt will get wider circulation than the original article, I expect. Not to mention all the blowback on Twitter. The riposte will be complete if she gets this:
CF: Hello, I'm Clare Foges
A.N.Other: Ah yes, the clueless idiot. Can I interest you in a copy of 'Cryptography for Dummies'?
Although our county (Suffolk) struck a deal with BT to supply 'Rural Broadband', colour me surprised when BT promptly upgraded the rest of the cabinets in the local towns & big villages that had previously been deemed to have 'no business case for upgrade'. Of course, it's not a picnic for them to add the extra infrastructure to make sure all lines in the sticks are <1km from a FTTC cabinet, but wasn't that what the taxpayers' shilling via BDUK was for?
You do what you need to do. Coober Pedy (AU) already has a lot of underground accommodation for that reason. I don't think the CAGWpocalypse is going to be anything like as bad as predicted, but even a degree or so rise in average temps is going to mean several degrees in peak temps in some places.
South facing windows? I don't think so in the Northern hemisphere!
A CME is a large (for very large values of large) flux of charged particles, mainly protons, moving at high speed. This causes radiation damage to electronics outside the atmosphere, hence potentially killing satellites. When it hits the atmosphere it causes all sorts of mayhem in the ionosphere, hence auroras, and stuffs up its reflectivity for HF radio. The large currents created in the ionosphere induce similar large currents in long-distance cable systems. Those (DC) currents in electrical transmission systems can saturate the magnetic cores in the transformers, reducing their inductance so that the combined overload can kll them if the circuit breakers don't work fast enough. In any case the electrical network shuts down (cf Quebec March 89).
I have no use for skype so why should my email be slowed down because some one is using skype?
But do you not use Netflix, or iPlayer? The same argument applies. Some types of traffic need timely delivery (on the scale of milliseconds or even microseconds) whereas others could be delayed by seconds or even longer. Some kind of QoS-based delivery goals would be good. But how should that be policed so that customers & networks don't cheat?
As an example, I have a femtocell gateway to provide mobile phone coverage in the house (we live in a hole). It sends the mobile data over the Internet to our mobile provider in an IPSec tunnel. I did think about using my firewall to mark outgoing packets of that stream with an appropriate QoS category, but talking to my previous ISP they said they took no notice of such markings. I ran some tests to other endpoints on other ISPs and the markings often got set back to 'best effort' anyway.
As others have said, the system should be allowed to classify traffic with different flow characteristics and treat them appropriately but not to differentially favour traffic in the same flow classes for competitive advantage.
Set the rules properly and the engineers can come up with effective solutions.
The recent issue with Diffie-Hellman is that the standards, and a lot of implementations, use one specific 1024-bit prime known as 'Oakley Group 2'. The conjectured hack is to calculate a lot of specific data from this prime which can then be used to rapidly break any shared keys generated by D-H using this particular prime if the D-H message exchange is observed. The counter is not to use that particular prime. More modern implementations tend to use Elliptic Curve Cryptography (ECC) which, as far as is known publicly, is secure with large enough fields over which the calculation is done. For example, OpenSSH has for a while used in preference an elliptic curve algorithm called Curvep25519, which is supposed to be as hard to break as 128-bit AES, i.e. impractical currently.
Note: Although thethere has been a recent debacle over the NIST Dual_EC_DRBG random number generator which uses elliptic curves in a specific, and conjecture to be hacked, way, this has no bearing on the general security of ECC.
I found the same issue recently with a phone app from a financial organisation who shall remain nameless. It was a server-side issue. Credit to them, they fixed it pretty quickly in comparison to the usual big org timescales after I alerted them. Since the app has been around for a while I was surprised no-one else had found it before I decided to use it.
Now I'm old, lazy & cantankerous, I only shave every couple of days or so, so I've mostly got the 'hobo look'. My wife isn't impressed but no-one else seems to care. When my daughters were little I grew a beard a couple of times as both my brothers-in-law were bearded & the girls didn't react to that too well. It was OK in the winter but it had to come off in the spring - too itchy by half!
That was most interesting. I had read about JTIDS years ago but I didn't realise Welchman was instrumental in its development. They interviewed John Scarlett and as you might expect he was still pushing the GCHQ line. That seems to be a major problem with history of this stuff. Scarlett's view was "we know whether this stuff is still sensitive or not, so we should make the decision to declassify". He has a point, but it does seem hard for them to rationally appraise secret stuff to decide the balance of risk/benefit of declassification. You can see that with the whole story of BP - the history of WWII looks a lot different now in the light of those activities.
As for Welchman's book, 'The Hut Six Story' was withdawn by its publishers and copies are now like gold dust - look at the prices on Amazon!
And what good will that do, since those hosts don't speak v6 and so won't be able to communicate end-to-end with v6 hosts on the network?
The point is that the v4->v6 in the home router works in tandem with the proxy at the ISP to go v6->v4 again. The ISP could, of course, hand the v6->v4 function off to a third pary if they are themselves a v6-only ISP (more and more likely in the future).
At some point ISPs will have no more v4 addresses to hand out, either fixed or dynamic. At that point new customers will *have* to have a v6 prefix. The way I see that working to support legacy v4, both in the home and in the Internet is two bits of kit.
1) In the home, the router supports an internal v4 rfc1918 network that NATs to a specific v6 external addresses in the prefix range, so v4-only hosts can connect out. radvd or dhcp6 will identify v6-capable hosts internally that can just pass through the router/firewall. Although the v6 hosts would also get an internal v4 address from the router, DNS64 would make all their external traffic go via v6, and they would only use v4 to connect to internal v4-only hosts.
2) At the ISP, run proxies with a mix of DNS64, NAT64 & 464XLAT to manage the connection from v6 hosts in the home to v4 hosts on the Internet via temporary v6 addresses allocated at the proxy.
There are probably edge cases that don't fit this model but that happened with v4 NAT and handlers got built into the NAT gateway code. Similarly this will get solved here.
Unfortunately I have a suspicion that some ISPs will instead go the v4 carrier-NAT route (mobile operators have already done this, at least in the UK), which at this point is rather more mature:(
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
