Posts by James 100
392 posts • joined Friday 26th June 2009 20:45 GMT
Posted in Google adds validation to DNSSEC
Do as they say, not as they do?
Google.com? No DNSSEC there. Google.co.uk? Same. Likewise OpenDNS.com.
It's depressing: when even the DNSSEC *advocates* aren't actually enabling it themselves, who will? (FWIW, my personal domain has DNSCurve and DNSSEC, as well as IPv6 - it's truly disappointing that Google don't!)
Anonymous nuisance calls
One thing I'd like to see changed is a bar on non-personal use of 'number withheld'. EU rules give individuals the right to make anonymous calls free of charge - but does and should this apply to businesses too? I don't think so. That, and make anonymous call rejection a no-cost option to be offered to all customers at installation time, along with being ex-directory etc.
The £90k fine is a welcome first step, but nothing like enough: disconnection from the phone network - no more phonespam - would be better. The US approach of charging hundreds/thousands of dollars per illegal call documented would be good too.
Maybe grab their outgoing call records (from the telco; should go back at least a few months for billing purposes anyway), check against TPS; for every hit, require them to provide proof of the 'pre-existing business relationship' or specific consent required for that call to be legal or pay £1000 for the illegal call.
Anything less, they'll just shrug it off as a cost of doing business - hell, they probably pay more than £90k for their electricity each year!
Fair use?
The idea of needing to pay for "title and excerpt" of a web page sounds alarming to me - all too close to the shake-down news companies have been trying in Germany with Google, demanding that Google pay for the privilege of including their pages in Google's search results.
Maybe Meltwater's "excerpts" were too big, but it still worries me in that context. If they were copying whole articles, fine, that's a clear-cut violation, but ...
Big mistake?
"whoops, we lost £21m down the sofa" ... I can imagine getting 10k, maybe even 100k adrift on a big company (depreciation calculations, things like that) - but even on an NHS/government scale, £21m is a big accounting hole to explain.
I got quite worked up enough when the last quarter was about £300 out (couple of misplaced expenses forms)!
Double-fault?
Nice to be quoted in a Reg article there*! From their later updates, it seems the original problem was excess route entries overflowing the TCAM (fast lookup memory in big routers), which makes them fall back to a much much slower routing approach - which can't keep up with the level of traffic you see on an ISP backbone.
That was problem #1: the lab stress-test leaked out and flooded the live network, bringing it to its knees. (IPv6 itself was unaffected - it's routed separately anyway - but since the line is authenticated over IPv4, as soon as you disconnect, your authentication packets get lost, stopping the PPP link coming back up.) Like their autopsy says, filtering out the routes then re-booting the routers cleared the problem - then they discovered a Cisco line card was still misbehaving.
The wonky line card seems to have been problem #2: it seems to be working again now, but they're keeping most traffic away from that card until Cisco can figure out why it misbehaved in the first place.
* Apparently the router problem took out their VoIP lines and access to their own status system - downside to "eat your own dogfood" as an ISP, when your own services are down, you can't use them to communicate about the problem - normally, I've found Entanet to be very helpful and responsive. In this case, the live traffic graphs on noc.enta.net were enough to tell me it wasn't just my line that was dead - how many other ISPs give you that kind of detail?
Re: Not flashy but owns a market...
Apple were in on ARM from the very first steps out of Acorn, using ARM's first (post-Acorn) CPU in the Newton. The *financial* resources, back in the very early days before ARM had a revenue stream to support them, probably made a lot more difference than Apple going on to join Nokia and co in using ARM cores in mobile phones.
Strategically, it looks to me as if ARM's in a pretty good place right now: pretty much owning the smartphone/embedded sector, with solid backing behind a push into the low-end server market (for colocation in particular, all those LAMP systems could move to ARM very easily; I wouldn't be surprised to see Google, Facebook etc making that jump in the next few years). Intel's great strength is "we can run all your existing x86 code unmodified": great when you want to run Windows with all those ten year old legacy applications, but a waste of silicon for running MySQL, PHP, Node, Java...
Posted in Paying a TV tax makes you happy - BBC
Missed opportunity
I wish they hadn't leaped at the chance to remove encryption support from Freeview STBs when they took control; if the BBC encrypted the TV output and issued decryption cards/keys/whatever on payment of their subscription, they could abolish the whole "enforcement" setup: no payment, no BBC services, so you're not getting anything without paying for it.
For the tiny amount of output from the BBC I watch, they represent atrocious value and I'd cancel in a heartbeat (I don't care that the content is ad-free, since I have no interest in it anyway!); £1 an episode for Dr Who would be much better for me. I'm happy to pay for a Sky subscription, because I actually get value from that: I watch those channels. Lose those, I'd get rid of the TV too, or just keep it for DVDs and the odd console game - for some odd reason, my house doesn't even have a TV aerial. (It was pre-wired for cable, in the early days of the cable build-out; my inner cynic suspects the cable company persuaded the builders to skip fitting an aerial...)
Storage problems
Like others have mentioned, studios were quite happy using massive reels of expensive single-use film stock to store their footage; compared to that, even the top-end 3-6 Tb/hr is trivial: the equivalent of 1-2 big SATA HDDs per hour. Throughput's much too slow on a single drive, of course, but make, say, a 12 drive RAID 6 and you're fine, with better fault tolerance than film stock had even before you make a backup.
@Duncan Macdonald: from your own link, "4K" is 4096x2160 for cinema purposes, 3840x2160 otherwise; your 66 Mbyte/frame is for 8K not 4K. Divide by ~3 and you have 200-odd Tb to archive, not 700. Probably LTO not disk for archival and backup - but bear in mind another relevant figure above: Superman II, weighing in at six tonnes of film. Does it really cost much less to store that weight of delicate material in secure, climate-controlled space than to store a hundred LTO-6 tapes? (They hold 2.5Tb without compression, weigh 200g and have a rated life of 30 years, 1,000,000 passes, meaning you replace six tonnes of film with 20kg of digital tape. Store with something like RAID6, re-scan and checksum all the blocks annually, regenerating any damaged blocks as needed: nothing like £40k/yr.)
Multicast?
If only multicast were available with wireless IP... (For that matter, it won't work as well as it could with ADSL or VDSL, because of multiple ISPs: if I and my neighbour are both watching a multicast video stream at the same time, we'd still have to get separate feeds from separate routers, down the same BT fibre as far as the street cabinet, unless we both happened to be using the same ISP.)
In terms of fairness, though, if mobile phone companies have to pay Ofcom "rent" for their spectrum, why should TV companies get essentially the same resource handed to them for free? (Personally, I'd think prefer if spectrum *were* allocated free, or for a cost-based handling fee to cover running a small office somewhere to track the allocations, but charging one group an eye-watering price then giving the other a freebie is clearly unfair.)
Location, location, location...
I remember Microsoft refusing to disclose the location of other bits of their cloud in the past, though the Office 365 service they provide to the university I work for is definitely located in Amsterdam, i.e. within the EU, which should avoid most protectionism about "data protection".
Pulling the plug on China would seem fair play from one angle - they get awkward, just go home - but unfortunate for businesses caught in the middle of such a dispute. Of course, they should be used to that by now...
People move on...
The 15 m won't all become unemployed overnight, it'll happen over the course of years. The 50 year old drivers will just work on and retire, the 20 year old would-be drivers won't get hired as drivers in the first place so they'll train for and find other jobs.
It won't mean 15m unemployed, any more than cars meant millions of unemployed horse riders/breeders/farriers etc - just a drop in hiring with the number doing that job gradually declining.
Of course, if the delivery robots are ParcelForce-based, they'll just fire the parcel at your house as they drive past at 20. CityLink will do less damage to your house, since they'll just fire a piece of card saying "come and collect the parcel from a sink estate three cities away, because we're the only outfit without a depot in your city".
I wonder if unmanned vehicles could usefully use the rail network for faster inter-city travel? Easy enough to form up batches of 20-100 small cars/vans into a "train", no issues with other traffic getting in the way since the rail network's already fully monitored/controlled...
There was a message "looking for the free comics? Click here" after I finally got logged in ... but it gave me a PHP error message. From Daniel B's comment above about the promotion being suspended, presumably that's supposed to be a message "sorry, our servers were flattened in the rush, back once we have decent hosting" - but they don't even have the hosting horsepower to deliver that message, let alone the promised comics.
It finally loaded - very slowly - just now: apparently "a number of technical issues have arisen", further updates to come. I hope they figure out how to get it working in the end, rather than give up on the idea entirely.
Posted in BT engineers - missed appointments
Third time lucky
I got FTTC last October. At the first attempt, two nice guys from Openretch (apparently a roving team of FTTC installers) showed up - with the wrong paperwork, and despite getting confirmation it was a BT error, they were told from higher up not to complete the installation, since a box not being ticked on the order form is something they can't fix on a same-day basis. They rebooked for later that week ... no-show: apparently, BT's ordering desk had booked the visit with too little notice for the paperwork to come through to the engineers, so nobody was sent. The following week, a local guy showed up right on time (calling first!), swapped the sockets round, hooked up and tested the new kit and went away.
BT wrote to me today to announce that FTTC's available here now. Having been using it for five months, I would certainly hope it's available...
Hardly "6 billion" making use of ham radio, TV's higher up the dial anyway ... are there really more radio hams than people wanting home networks? I doubt it. Just have an equipment buyback, trade all the ham gear in for a VoIP handset, problem solved ;-)
Personally, I'm interested in the G.hn coax option, running a gigabit or so up the disused drop of coax Vermin Media left between floors a while ago. (Yes, having cat 6 would be nice, but I draw the line at drilling through walls/floors myself. I'm guessing once I have a few hundred quid lying around I could get a local electrician to poke a dozen drops down from the attic and stick a GbE switch up there, but until then I'll stick to the Solwise powerline stuff.)
Apparently, thanks to the Internet use of contraception is no longer enough to stop me having to raise kids. Everything from busy roads, to newsagent top shelves, to the bleach aisle in the supermarket is dangerous to leave kids unsupervised - so parents are expected to do their job of stopping kids having unsupervised access. The minute they get online, though, suddenly it's everybody else's problem: we all have to put our own resources into keeping feral brats away from anything not child-safe.
If I left a 5 year old kid unattended in the park to chat to every passing stranger, it wouldn't be the strangers or the park getting investigated - and rightly so. Why is it suddenly different with a virtual park and strangers from all over the planet not just the local area?!
I quit Vermin back in September of last year. I was promised they'd send me a jiffy bag to return the STB and cable modem, but it never showed up. Not really a big deal, so I didn't bother chasing.
A while later, I pointed this out to their Twitter rep, who said to phone about it. This was in late December. Drone said "OK, I'll send an engineer round to pick it up ... how is next Tuesday afternoon for you?"
"You mean ... Tuesday, the 25th of December?"
"Yes, would that do?"
I was tempted to go for it to see what happened, but moved it back a few days. It's now February, and the kit's still sitting in a corner of the spare bedroom awaiting collection.
When I had it, the performance was fine - as far as the speedtest.net nodes. Complete junk for accessing real-life sites, of course: video streaming, software downloads, SCPing data to/from my colocated servers for work (probably tripping some anti-P2P measure?). As soon as I switched to a 60 Mbps FTTC product, the buffering disappeared.
"Netflix may need to build data delivery centers closer to consumers to alleviate these problems."
From a recent Reg story, it's clear Netflix *ARE* building out distributed delivery nodes - colocating with ISPs where they can, Akamai-style, peering directly with the rest so users are always served by local nodes rather than stressing a backbone.
I've just finished re-watching 24 day 2 from Netflix. The TV's built-in Netflix app can start playing an HD stream as faster than a DVD player can get in gear (Entanet FTTC, ~ 60 Mbps); even over Three's 3G to an iPad, though, Netflix generally manages to deliver decent streaming quality with about 3 brief streaming glitches out of about 15 hours of streaming.
No idea about Lovefilm, but I'm very impressed indeed with Netflix. Having been cynical about Net video for a long time, I'm stunned by the service they deliver now. I can see why even Sky are moving in that direction now, with 'Anytime+'; with decent bandwidth (i.e. not the over-contended wet string peddled by the likes of TalkTalk and Virgin) it really does work very nicely.
At the very least, there should be a pro rata refund: if I pay £30/month (i.e. £1/day) for broadband, and it's out of order for 3 days, I should get a £3 credit/refund, since I haven't actually received the service on those days.
Back when I was getting FTTC installed, BT screwed up and failed to install first time, then failed to show up for the second appointment. When making the original appointment, it was emphasised to me that if *I* failed to be there, I'd be charged £85+VAT ... apparently, though, BT failing to show up doesn't oblige them to pay me, though it really should do.
I wouldn't be at all troubled by this wiping out the bargain-bucket "ISPs", cutting every corner known to man; the world would be a better place without them, particularly the draconian traffic limits, shaping and port blocks with massive oversubscription.
What I want? Simplicity and honesty.
A nice clear price/contract. No more "unlimited (but only while standing on one leg, excludes use with applications we don't like and use between 9 am and 5pm)": cap it at 10 Mb if they must, but that 10 Mb contract should mean you can actually use that 10 Mb, whether it's for your smartphone downloading a few kilobytes of email every hour or two over a month or streaming a brief burst of HD video to your giant TV from a connected tablet.
Giffgaff's absurd contortions over "unlimited (we can't really afford to do it, but we want to pretend we can by adding other silly limits instead) data" really irritated me, as a long-standing customer. O2 charge them by the megabyte on a wholesale level - so pass that on. Don't claim "unlimited" then have to ham-string it to stop anyone being able to use "too much" - and don't slap arbitrary restrictions on what sort of device those bytes are going into, either.
No more discrimination between tablets, handsets, dongles or personal hotspots: a byte's a byte. No discrimination between applications, either: a VOIP call or video stream might not work well for technical reasons (latency, jitter etc or plain old speed limitations), but they shouldn't be allowed to attempt to restrict that.
Oh, and BAN sim locking (if I buy a handset with a contract to pay £20/month for two years, yes, insist on getting the money - but don't restrict the handset as a ham-fisted way of trying to force me to keep coughing up) and tethering restrictions (Apple/Giffgaff, I'm glaring at you here!)
0.1% still beats 0.0%!
So, right now the Dutch government is getting about 0.1% tax on the money flowing through. Not a huge wedge, but for just acting as a passive conduit, not a bad deal either. Ban this arrangement, they will get 0.0% instead: a net loss.
From a business point of view, they have a bunch of customers, paying them a little bit of money, so they make a little profit on the deal they wouldn't otherwise. Get greedy and try to charge 20%, they'll just lose the "sale" and get 0 instead - but tighten things a bit, they might get 1% or 2%, so everyone would be fairly happy: 1% of those trillions would make a heck of a big difference to the government's coffers, but losing the 0.1% wouldn't benefit either party.
Tried it - nice idea, poor implementation
I tried it recently - pretty poor call quality compared to Skype etc (using WiFi on a fast connection). The price for 08xx calls is very nice, but the lack of incoming call support along with using your mobile number rather than landline for caller ID are both irritating to me.
Allow me to answer calls to my landline on the mobile app (as I think Vonage does now in some form?), present my landline not mobile number for caller ID, get the call quality up, I'll be happy.
I can think of quite a few things I'd be happy to delegate to a machine were it possible and practical. Do you really enjoy doing laundry? I'd be happy just dropping dirty garments in an automated laundry basket and waking up to them clean tomorrow. Cooking dinner - yes, I do sometimes enjoy whipping up something tasty, but other times I just feel hungry: a Star Trek replicator would be great then (and indeed that's pretty much what home delivered takeaways offer now, and seem to make plenty of money at it).
If you're rich enough to have staff to take care of your laundry, cleaning, routine cooking (and cleaning up afterwards, which I'd very happily automate away!), your commute (again, I'd love a taxi/chauffeur user-experience with my car: get in, state your destination, get on with some reading/sleep/eat etc), yes, you have little to gain - but the rest of us will certainly benefit.
Tablets ... earlier today, I was going through email and making some schema changes on a remote database server. The email side, I find I DO often use a tablet already; the schema changes, a tablet would have been fine for since I was using a web interface (legacy app, the one SQL Server instance I still babysit; the rest's MySQL).
Everything I've done on this laptop today could quite easily have been done on a tablet right now, if I wanted; I spent a while SSHed into some colocated servers, moving data around, which would be slightly less convenient without multiple side-by-side windows, but nothing major - and nothing that couldn't be fixed quite easily in time. There are already developers with similar setups to mine making increasing use of tablets quite happily.
Re: "Plugging in" is perhaps the wrong image
This is weird: I can certainly understand Netflix wanting settlement-free peering, rather than having to pay for transit into heavy customers like TWC; offering Akamai-style colocated edge cache nodes also makes a lot of sense for both parties.
Are TW currently charging Netflix for transit? That would explain their resistance to moving to settlement-free peering or edge caching; "better service for your customers who use our services" versus "revenue stream" is all too easy a choice for most cable outfits!
Re: Appalling
My experience - customer since the age of eight, when it got me a cuddly toy squirrel; family, customers since my great-uncle was a branch manager - is very similar. As soon as BoS collided with Halifax, the wreckage started decaying: features started getting dropped from online banking, things became less reliable...
Even now my mother's credit card balance isn't visible in their online banking fiasco! Mine is, oddly (though I never use it, keeping it around just in case of 'emergencies'; Amex gives me spending rewards, Barclaycard works well enough for the places that don't take Amex).
I was actually bitten by this outage: my own Amex payment went through (from my BoS account) in about an hour, my mother's BoS card payment was relegated to the next-day route, as was a small transfer to my brother's Royal Bank of Scotland account (paying him for some tickets he'd bought me). Until reading this article I didn't know it was a service outage ... which says something about the quality of BoS communications.
Re: @ac
"Sky already have Easynet's network which is an actual collection of cables and things. O2 don't actually own anything except the DSLAMs (they rent capacity off other companies like BT) and a few core routers so where's the synergy?"
Migrate those Be DSLAMs onto Sky/Easynet's backhaul network, boosting capacity as needed on the way? I got the impression there was at least some actual network behind Be; O2 referred to using some of that for their mobile backhaul too, which was one complication with spinning out the ISP business again.
I get the impression Sky's bottleneck is mainly the connection between BT's equipment and their own network: BT charge the earth for that bit (partly so they can make the per-line charge look cheap); Be using their own DSLAMs bypasses that. Someone posted (on another Reg story I think) that Sky have been running out of ports on their own kit and using BT ADSL ports instead, which also gives a worse service in various ways.
My gut feeling is that Be+Sky/Easynet could give an improved service. Just pooling DSLAM ports would help a bit (the exchanges Sky have run out of ports on probably aren't exchanges Be are out of ports on too), and just combining links to give a single 200 Mbps hop instead of two separate 100 Mbps ones will give slightly better peak performance (peak on one link won't exactly coincide with peaks on the other), even before using the greater volume to get better prices.
My brother went with Sky at home ('too good to be true' offer when he switched the TV service over from Vermin); I went with an Entanet reseller (Vivaciti) for home (FTTC) and Be (ADSL - no FTTC on that exchange) for the office. He regrets it, I don't regret either of mine.
I just wish we could get past the "100* (* actually 0.01, but we lie, because the ASA don't understand technology or numbers) Mbps for only 5p* (* plus oxygen usage surcharge, first-born child and a weekly kneecapping) per month!!!" nonsense. Be actually deliver pretty good speeds; Sky/Easynet apparently did until choking up the network, and having admitted there's a problem it sounds like they will do a better job fixing it than the worst bottom-feeder ISPs, who would just say "yep, we cram 100,000 customers on one piece of damp string ... what do you expect for 50p a month?"
Like another comment here says, people who wouldn't even think of buying Tesco Value bacon or cornflakes still jump at "Crappy broadband, half the bandwidth for 10% less money!" every time. Sad.
4% up, with a dead competitor?!
To me, only getting a 4% boost when their biggest competitor imploded sounds pretty disappointing at best!
I've never been impressed with Dixons/Currys/PCWorld, but I suppose they must be doing something well to stay afloat this long, and they can occasionally be handy for very urgent trivial needs like a new cable or toner cartridge. Amazon's next-day delivery beats that these days unless you really need something RTFN though...
Re: Mis-read
Maybe I'm too cynical, but I'm inclined to suspect it IS about "inadequate" payments ... the journalists and their police sources broke the law. Even while the story itself was breaking, we kept having reports along the lines of "someone has been arrested ... police haven't officially said who, but we bunged a plod some cash to find out that it was (this guy)" - exactly the crime people were being arrested for, but apparently they got away with it this time!
Did we really need a "public inquiry", rather than arresting and prosecuting the culprits like for any other crime? What's the point in new laws, when the acts that caused the problem were already illegal and the culprits known to the police for it?! Enforcement, people!
Re: No virgin, no fibre
I have FTTC in Perth (Craigie, ex-Vermin customer). It did take three engineer visits (first one showed up with the wrong paperwork and went away again; second was a no-show; third came in, hooked it all up no problem with a bit of tech-to-tech chat, no problem). The bandwidth is pretty good - about 65M down, 15 up on actual tests, haven't checked the line stats directly yet - and the service (Entanet) is much, much better than I ever had from Vermin on what was supposedly an equally-fast link.
Skipping upgrading cabinets irritates me - in fact, the engineer who actually installed my FTTC noted his own cabinet round the corner is much further down the list to be done - as does the misuse of "fibre" for this and cable modem service, which will no doubt come back to bite them when promoting genuine FTTP services...
I do wish BT were more transparent about the rollout plans. Yes, they can change - but saying "March ... no, maybe June ... OK, October it is" is worse than useless. I know they release more information to ISPs than directly to the public, I wish they'd change that at least!
Re: Actually
I use 5 GHz for exactly that reason - but have 2.4 GHz enabled on the access point too, because my mobile handset (like most, if not all) only supports that band, ditto my old (relegated to occasional use) laptop and the bedroom satellite box.
Mine's actually the only 5 GHz presence I've detected, against half a dozen or more 2.4 GHz nets fighting each other. On the bright side, moving nearly all my traffic up to the higher band means I'm contributing very little to that mess of interference, as well as experiencing very little myself.
It's a shame 5 GHz support isn't more widespread, though; a nice surprise that the Nexus 4 apparently does support it.
Where am I?
I live in Scotland and have a little blog which happens to be on a server in the US (Virginia, to be precise). Obviously what I do online is bound by Scottish laws, since I'm physically here, and the server to US/VA laws, because it's there - but the minute my website gets visited by somebody in France, suddenly I need to comply with French laws, even if I've never visited or interacted with France in any way?
If France gets authority over me that way, does North Korea, too? Can Kim Jong Ug demand the extradition of El Reg's people for insulting the Supremely Misplaced-Marbleness, or the Saudi/Iranian theocracies stamp out any non-Islamic website? The notion is absurd.
Now, the French government might just be able to get the US/California government to persuade Twitter to help them investigate what may have been a crime within French jurisdiction - in the same way they could get my local police to come round and interview me if I seemed to have relevant knowledge of a crime in France.
At the very least, jurisdiction should require manually *instigating* interaction with that location - for example, composing and addressing an email to someone there, or placing a phonecall, not merely auto-responding (like a webserver) or forwarding (like a mailing list).
Re: And this is why the DeadPhone will fail
Fortunately, this flaw seems to be BB7-specific (and earlier): BB10 finally drops this absurd flaw, switching to the same approach iThings and Android devices have used all along.
Of course, the BB-worshippers who point to the single point of failure as being a "security feature" (instead of a pure encrypted tunnel from handset to mail server, like I have, you put an extra step in the middle and they think this 'helps'!) will need to think up a new line...
Short version:
A top-down framework for entrepreneurship. Sadly, they still haven't figured out that this is much the same oxymoron as centralised autonomy...
The real tragedy is that we get to foot the bill for their latest half-baked pipe dream - again.
Wrong destination...
I thought he'd be a perfect fit for running Bernard Matthews... except their turkeys sell better.
It's an odd departure in itself, though departing just after Win8's launch rather than before does make sense. Management guru, though? Certainly with hindsight he'd have been better off going out just after Windows *7* launched, not 8.
On the other hand ... Kevin Mitnick's security consultancy is so well known because he eventually screwed up and got caught, not because he was infallible and got away with it all...
Bungling bundling
The bundling has always irritated me for various reasons; I'd much rather pay £240 for the handset plus £10 a month over the year than £30/month for a year for a "free" handset. I think things have improved lately with far more "SIM only" offerings from most operators and more SIM-free handsets, but still not good.
At least one EU country bans SIM-locking, as I recall, another bans handset subsidies; I can't see any reason why we shouldn't do the same.
As for price changes: fixed term contract should mean fixed. If I'm committing to paying £10/month for two years, they should be committing to the same. If I don't get to phone them up and "notify" them that the payment is going *down* 10%, why should they be entitled to put it up that way? An indefinite contract, where either side can switch to another tariff freely, is fine: but if I lock in, so do they.
Pay for .. what, exactly?
So, when I buy something, which was made in China to an American design (paying with my credit card from an American company) on a website hosted in Dublin and managed by a company in Luxembourg using software from Washington, what exactly should the UK government get paid for, over and above the 20% surcharge it already slaps on top of almost all my purchases? The roads used to get it to me? No, the courier company already pays for that several times over in fuel and other taxes. My broadband connection? No, I pay for that myself - plus that 20% surcharge again, out of income already taxed to fund whatever the government feels like squandering on.
Sadly, the irony that most of the people complaining about Amazon in Luxembourg paying taxes in Luxembourg rather than in the UK were also complaining about Vodafone Luxembourg not paying taxes in the UK rather than Luxembourg is probably lost on such people...
Re: Crappy FTP services
Their FTP service is worse than that: it doesn't even exist in the first place! The only way to get FTP there is to set up a server and install an FTP daemon yourself (or pay to use somebody else's), at which point it's not really much of an Amazon service any more than any other colo/VPS provider.
Are you thinking of S3 as an HTTP (as opposed to FTP) download host? If so, you're really supposed to front it with the CloudFront CDN for fast static downloads (though I do know a lot of people use S3 directly, because it's actually a fairly good service for that, rock-solid reliability, albeit not the fastest). Both S3 and CloudFront certainly *do* support download resumption (partial GET requests with a specified byte-range) - what they don't support, though, is FTP: it's HTTP(S) only.
ELB = SPF
Relying on a single ELB seems to be a point of weakness, unfortunately.
Given the media emphasis in the design, I would expect Amazon are using their own CloudFront CDN for streaming, rather than ELB - CloudFront being a simpler, more distributed and more robust setup for various reasons.
If I were setting up a big Amazon operation like Netflix, at the very least I'd want to spread across 2+ ELB instances - we've already seen ELB fail in the past, it's bound to fail again in future.
In general, I find *some* Amazon services (S3, CloudFront) to be very solid and well designed for robustness, while others seem much more vulnerable: I've had problematic EC2/EBS outages in the past, but the static assets from CF+S3 never had a glitch.
Specify destination, not route
I'm reminded of the old quote apparently from Henry Ford, that if he'd asked his customers what they wanted they'd have said they wanted a faster horse. Probably an urban legend, but exactly the mentality that seems to prevail with tendering.
If you get a taxi, you don't usually say "Go along Hill Road, take the third left, carry on until you pass the supermarket then turn right" - you say you want to go to East Road. The taxi driver's supposed to know the route and anything relevant. The government, though, would phone up the taxi office and say it wanted a navy blue four-wheeled vehicle running on petrol with five seats and a CD player...
To quote a genuine example from a tender a few years ago: "This web application must be written in Visual Basic or .Net". It went on for pages like that, mandating NTLM authentication, two contradictory requirements for the top section of the pages (two documents specified totally different naming and logo requirements, something to do with an ongoing reorg that left it unclear quite which branding actually applied). They did fail to specify a requirement about URL structures which later caused problems, though.
(Oh, to cap it all, after setting up the architecture astronaut job on their hefty cluster of servers, it turned out to run faster on Amazon's smallest EC2 instance - even when that was over a WAN to Ireland, against using their on-site cluster over the LAN. Plus replacing a big chunk with PHP, because there was never actually a need for .Net involvement in the first place...)
Posted in Windows Vista woes killed MS Pinball
Re: But it's not 64-bit so you can't have it?
Raymond explained this himself in the article's comments section - at that point in development, their Setup infrastructure couldn't cope with 32-bit exe files getting mixed in! (A bonkers limitation, IMO, and apparently one they fixed later, but I can understand them not wanting to put that effort in just to keep a freebie game working.)
Knowing Microsoft got suckered like this (comment-free source) I feel slightly better about having to kludge around a mass of undocumented outsourced Java spaghetti at work. Only slightly, though.
Karma?
I'm not too convinced of the merits of this patent either - but when a big company plonks down billions to buy and gut a whole company, sacking a load of staff just so they can asset-strip the patents to use as ammunition in a patent-war, there is a certain karmic balance when one of those patents falls apart afterwards.
Compete with products, not lawyers! (Which of course goes for most if not all of the current phone outfits, I know...)
The real flaw
Really, it's the whole evil concept of a "speech crime". Barring actual threats - which were always a different crime anyway - or breaches of national security, or similar - why should ANY mere words be crimes in themselves!? The whole concept is flawed.
If "offensive speech" is out, that's pretty much every religion banned (since they each teach the others are wrong/going to hell etc). Of course, scrapping the whole stupid law would require politicians to admit they were wrong...
Refusing to leave, resisting arrest?!
Her actions are completely crazy here. OK, trying to buy a bunch of iPhones to re-sell - nothing wrong with that, though Apple are free to refuse to sell if they want. Refusing to leave the store when kicked out, though, then resisting arrest for 15 minutes confronted with one then two actual cops!? That's just nuts.
Language barrier or not, "GTFO of this shop" is easy enough to understand, as is having armed men in uniform putting handcuffs on you - WTF did she think that bit was, an over-accurate stripogram? It was obvious she had to leave; refusing to go even after the police get involved is unbelievably stupid however you look at it.
Cultural differences don't really help her case either: the Chinese government isn't exactly noted for tolerating people standing up to authority.
Re: >:(
It seems what Foxconn said was "the law says you have to go and be assessed in Huizhou". Now the arbitration committee has ruled that yes, the law does indeed say he has to go and be assessed in Huizhou.
What did people expect Foxconn to do, turn round and say "no problem, we'll just get the law changed so you don't have to go there for assessment"?!
Foxconn *haven't* told him to "get lost" - they've kept paying a salary on top of medical costs - but Chinese law (and, Reuters seem to indicate, their insurance policy) requires an assessment by the government disability office in Huizhou.
It's not as if he's expected to travel to another country: it's only 43 miles away! I'd be expected to travel further than that if I needed specialist medical treatment right here in the UK (the hospital near my home is a minor one, the one 20 miles away my research is in doesn't cover all specialties: a lot of cardiac stuff gets sent 50-60 miles south).
"BT have never had a monopoly on fibre, in fact they are helping break Virgins."
Virgin still aren't offering a genuine "fibre" service - which FTTP is - they just offer the same old copper coax cable modem service, but with faster (DOCSIS 3.0) modulation.
That's the problem with letting BT and Virgin get away with lying about their current services being "fibre" because there's fibre beyond the 100m or so of copper that goes to your house (which, of course, cable modems have always had, ever since they were half-megabit services) - when BT offer a genuine fibre service which really does connect fibre to your home, it gets confused with the fake-fibre services.
Re: Hard to see the need..
I was surprised to see Cambridge and Oxford on the list too - and Perth, where I am, where BT has already rolled out FTTC to the only BT exchange (ESPER) and Virgin have most of the place wired for the so-called "fibre" coax service too - which was supposed to be 120 Mbps, though the website only mentions 100 Mbps again now. Even before FTTC hit, ADSL speeds seem to be pretty good, both BT and unbundled operators (both C&W/Bulldog and Sky/Easynet, as well as TalkTalk/AOL). Where's the funding need?! Maybe there are gaps in the surrounding countryside, but hard to believe those are the most pressing.
Perth in particular is just 20 miles from the (larger) Dundee, whose central exchange ('Dundee Steeple'/ESMAI) still has no FTTC installation date. (It's also the exchange which serves both my brother's home connection and my office...) It lags far behind Perth already ... so it's Perth that gets extra funding to improve further?!
Given the level of competence so far...
... the case will collapse after it comes out the government had actually screwed up the targetting entirely and been spying on his homeless illegal immigrant namesake Kim Dotorg instead...
Sherlock, because he'd be bright enough to check his own immigration records before snooping. (Actually, he probably wouldn't bother, but he'd certainly know he was meant to!)
Re: Where AMD can win
"They have abandoned ideological ties to x86. They're still committed to what makes money, but they have no embarrassment over punting ARM."
It's ironic to see this giving AMD an edge now, considering how Intel tried to leave x86 behind while plugging Itanium - only to have that same AMD come along and eat their lunch by offering x86-64! Back then, of course, the small server market really needed "x86, but 64 bit", so they could run and port their legacy x86 environments and applications over; these days, for mobile and server applications, x86 isn't even advantageous let alone essential.
A shame Microsoft tried to give x86 an artificial edge in Windows 8, by restricting ARM to the feeble subset 'RT' rather than a full-blown port. (Yes, regular Windows-ARM software would still be hard to find - but that's no reason to ban it outright!)
Re: Microsoft enterprise price increases steepened.
It's an interesting reversal - for a long time, they had a tight grip on the *bottom* of the market (which at the time meant PCs) while they fought against the entrenched higher end (NetWare, various Unixes, mainframes etc) - using the revenue from all those Windows and Office sales to push their server products in.
They've got a pretty solid hold on the enterprise end right now, I think, but squeezing hard in a bad economic climate seems short-sighted, particularly when they have new competition from the likes of Google Apps and other hosted services... interesting times ahead!
Re: 100 meeeeellion?
For accessibility purposes, "disability" can include being colour-blind (which affects a substantial number of people to some extent, but wouldn't generally preclude driving). It wouldn't surprise me if one in six had at least some minor impairment of that sort.
At first glance the "disabled access to driving licence renewals" might seem funny, but my grandfather is disabled (knee injuries from a car crash in 1981 mean he can't walk far, particularly now he's 88!), but can still drive - indeed, living outside a city a mile from the nearest bus stop, he *needs* to drive. Now, mandating that the driving licence site work with Braille readers would indeed be a bit silly (like the train lavatories with a little yellow light, "door locked when lit" - in Braille as well as print), but most aspects of accessibility are still very much relevant.
If it means fewer Flash-heavy, light grey on dark grey tiny-print websites that don't work properly on modern browsers, I'm all in favour - though I thought government websites had to be disabled-friendly anyway?!
Effective lock-in
We have a $$$$ annual campus-wide licence for Oracle. For a lot of things, Oracle doesn't make much sense - but since one particular instance runs a little task called "payroll", changing it is something to be approached with extreme caution...
It's possible to migrate individual products off, though - it may not show an immediate saving, but as soon as you hit the point you could switch to a cheaper Oracle offering it'll pay off. Down to a single (virtual) CPU, for example? (Payroll's certainly an important job, but doesn't exactly need vast computing resources to run!)
Short-term, the 'path of least resistance' is easy: just cough up another six figures for another year and leave everything as-is. Each time it comes up for renewal, though, I feel the urge to whittle away at the reliance on it.
More importantly, though, in years gone by it was "either we use Oracle which we've already paid for, or pay for (other db)"; these days, it's "either we use Oracle, which doesn't cost extra *for now*, or MySQL/Postgresql/other which is free anyway". Much easier - and much less use for Oracle. (Back in the 90s, we had a cross-platform user account sync server which used Oracle as the database for exactly that reason - would it now? No chance.)
