35 posts • joined Tuesday 16th June 2009 21:59 GMT
Tetra, being digital...
As I understand it, TETRA is a digital system, therefore any interference WON'T be voices coming out of poorly-shielded HiFi (that'd be some predecessor analogue system). I have a vague recollection that the radio-interface for TETRA was spawned off of early GSM specs, so you'd expect any interference to audio equipment to be in the form of buzzes and beeps not-dissimilar from GSM....
Some machines already employ an intermittent motor on the card insert mechanism
presumably to thwart skimmers.
The card judders as it is absorbed into the slot. This would make readng the magstripe challenging to say the least.
Well it should work
I've successfully loaded BBC computer programs off a (non-Apple) MP3 player
Any connection to the Raspberry Pi Raspbian "Wheezy" based on Debian?
Cyclist threat from inattentive drivers
A couple of months ago I saw a cyclist (in a cycle lane) almost rammed from behind by a car doing 35-40mph. The car swerved out into the road only just in time. I was driving in opposite direction at time... just as I was thinking "what happened there" I saw that the approaching driver had a phone in their hands in the middle of the steering wheel - and despite nearly killing the cyclist just moments ago they were STILL fiddling with it.
'Averages' tell you very little ... need a histogram
In my world, the "average" duration of a call may well have halved in the past few years ... mostly because every other call is now some chancer from India trying to help me reclaim my PPI or chase insurance payouts for accidents I haven't had - and 4 times out of 5 it's a silent call anyway.
The last thing we need...
" the changes include tax breaks for companies that patent their products"
... is more incentives for big companies to file large volumes of largely-vacuous patents.
Too many patents are just a total waste of time to search (and prohibitively expensive to contest) for small companies or individuals trying to develop some true innovation.
Simplicity of Billing and visibility of minutes/data-allowance used and remaining
They force direct-debit, paper-free billing - but make it very awkward to get to your bill on the website. Every bill (for the past year at least) should listed on a single web-page, downloadable by a straightforward right-click "save as".
It should also be possible to EASILY sign up to a PDF bill by email, sent automatically, monthly.
Voda and/or Orange require you to jump through multiple hoops to download bills, multiple steps forwards and backwards if you want to get several month's-worth in one go, websites that fail miserably if you start trying to simplify things by spawning multiple tabs, proprietary not-quite-PDF browser-plugins that won't save the document easily, email-it-to-me things that don't work, spurious choices of "full" vs "complete" vs "4-page" bill, click-to-proceed buttons which are located beyond the default border of the popup window (needs resizing even to see the button).... ARRRGGGGHHHHHH!
I second the call to get rid of silly tethering restrictions. If you pay a fair price for the data, it really shouldn't matter whether you're tethering or what device you've popped the SIM card in for this hour or that.
When you've got 3-4 data contracts (different networks; redundancy/backup!), 3 data-dongles, 2 'main' voice phones (different networks), an Android tablet (at present, strictly for data: apps/web on-the-go), and a few more live PAYG phones on-hand (you never know), and a few loose SIM-cards sitting on the Rubiks cube next to the monitor ... it does get a bit confusing remembering what you can do with which.
What I'd rate them on / what I'd like to see
1. No unreasonable underhand charges (35p per minute or part thereof for Answerphone-access [Orange] - 10-12p/min is tolerable, but should really come out of inclusive minutes on predominantly voice-contract; £15 step-charge for going a few MB over a multi-GB monthly data allowance on a data-contract [Vodafone])
2. Option to have text notification of imminent exceeding of data-bundles sent to a phone/device other than the device at the centre of the contract (e.g. to a phone handset rather than the data-dongle)
3. 24hr support for network-related issues
4. Acknowledgement that network faults can and do occur. Three and Orange have no specfic means to report faults or suspected faults, and they do not acknowledge such issues (masts down etc) on their websites.
5. Transparency about what's permitted or incurs additional charges on data packages (e.g. Skype)
6. For data connections, not just raw throughput, but ping-times [Three consistently quicker than Voda in my experience] and propensity to mangle (proxy) data connections [in my experience Voda causes spurious delays and random http 504 errors (4+ years now) and web pages which just intermittently never get returned (including some banking sites)].
7. A commitment from telecos to resolve longstanding issues (see 6) rather than constantly deny them or fob folks off with try updating this, that, or the other, try in a different location, try standing on your head etc etc.
8. Admission from the phone co when there's some fundamental problem, e.g. which is stalling data to dial-up modem speeds (40kbps) for much of the evening for months on end, despite excellent signal strength [Orange].
9. Greater flexibility with contracts - e.g. more shorter-term/flexible options rather than 2-year contracts (which have to be paid in full if you quit early) even when there's no appreciable device-subsidy.
The physics is easy...
... it's the tech to achieve it that's tricky.
For the physical configuration you need to layer the colour sub-pixels in a stack on top of one another (not adjacent to each other like an LCD).
At the bottom of the stack is a white reflector, then above that you layer cyan, magneta and yellow pixels which can individually change from colourless transparent to their saturated C/M/Y.
This setup will get you close to the brightness (in principle) of a printed page.
All that's required is to design a technology to produce such a configuration.
Since each transmission signal spans (almost) the 8MHz bandwidth, comprised of (almost) 8k carriers for TV SFNs, 'ghosting'/reflections/echoes or other distant trasmitters effectively become notches in the received frequency spectrum domain, only severely affecting a handful of the 8k carriers - which the data redundancy and error-correction flywheels over. The symbol rate is of the order 1ms (off the top of my head; this closely related to the period of the frequency-intervals between the carriers so that the Fourier transform is nicely orthogonal) so as long as echoes/reflections/distant transmissions are delayed by rather less than 1ms (the safety "guard interval" probably being about 1/4 to 1/8th ms), then intersymbol interference won't be a problem and all will be hunky-dory.
Re: Hoarders unite!
I have two BBC Master 128 computers, a single 5.25inch disk drive, and twin Cumana disk drive. All my old 5.25" disks, several unopened boxes of 5.25" disks. AMX mouse, joystick. All fully functioning. Plus a BBC-style "cassette lead" (yes it does work to load programs off an mp3 player!).
If a BBC micro won't start, one of first things to try is to see how far the start-up beeps get - this is some diagnostic. Remove and re-seat the keyboard cable-connector. Possibly remove and re-seat the OS ROMs.
If you've got a mod, check the wiring for that - I carefully extracted a non-standard video mod from one of mine as that was causing trouble.
A BBC to PC RS232 adapter cable with 15 -> 5volt-dropper resistors wired inside the plug on the signal-lines from the PC...
Probably every mobile phone I've ever owned (even though the first three I stopped using as they no-longer worked properly) - batteries and/or keyboard worn out.
I've got a JVC not-a-Walkman which sounds horribly tinny (capacitors expired?) and LCD screen (for clock/radio) long-since non-functioning... but it's the only thing left that will (sort-of) play tapes.
... and no end of other "rubbish".
Vodafone mobile broadband has seemingly killed youtube vids for the past few days
I'm seeing a "An error occurred. Please try again later", accompanied by a fake analogue-noise "not tuned in"-kind-of image.
If I connect via Orange instead, then the video works (although it judders to a halt every 10 seconds because they don't have the bandwidth)
Vids on the BBC website and elsewhere work just fine.
I wonder if Virgin Media use the same flaky web-traffic-proxying system that Voda use, and it had a bad update recently...?
My pet hate for misleading/deceptive Google ads
are the big green graphic "download" buttons which appear typically on the download pages for legit shareware, but are nothing to do with the page-author's content but are in fact an 'advert' which links you to software/malware unknown...
The Fuji minilab machines are absolutely superb - the print-making machine is phenomenal in the image-quality and precision (their minilab film-scanner is pretty poor by comparison, and aliasses film-grain - but probably didn't matter too much as it was basically legacy-support). Kodak got into the seemingly got into the digital minilab market late, with a joint-venture product, that gave lousy prints (from a multi-exposure LCD shutter mask) compared to Fuji's precision laser scanner technology. Pity.
I did work-experience at Kodak/Wealdstone R&D in 1990 - the stuff they were doing then with HDTV telecine (well ahead of its time) and digital image enhancement (mostly contrast-management from what I remember) was impressive... but perhaps too far ahead of its time in the market.
I still like/d Kodak professional colour negative films... but haven't bought much recently as I've gone digital (albeit a relatively late-adopter).
I'll understand why they failed, but still be sorry to see them go.
@ A.C. lies, lies and network coverage maps
> the problem i have is i don't know if any of the other networks are any better, its not like you can trust the coverage maps...
The first mobile I had, over 10 years ago, was a secondhand Bosch 509e. It had a network-scan function, and would show you the "bars" of signal for each network it found.
Several newer phones I've had will do a scan, but they don't show you the signal strengths.
Of course with the right AT commands it might be possible to hack a USB- or Bluetooth-connected mobile or dongle to read out the information?
The devices mentioned may not interfere with each other, but I bet you can't listen to any FM radio (especially radio3 or 4 which tend to have lower modulation levels) without the audio being overlayed with all sorts of buzzes and whistles (unless you live right next to the transmitter). AM radio is almost certainly also shot. You'll struggle to hear any aircraft comms within your room if you had a suitable radio too.
You might want to check whether the "Rugby" (now Anthorn) clock is actually synchronising - I recently investigated a clock which had stopped syncing and found the problem was that it was sited within a metre of an electrically noisy BT Homehub (wi-fi, not PLT) power brick.
Proper sample, controls?
Randomised for left/right pockets?
People could quite conceivably walk asymmetrically naturally and/or through carrying a bag on one side, leading to mismatched left/right bone density.
Controls to eliminate mechanical (and/or thermal) effects, such as phone squeezing on leg and causing poor local blood circulation?
And in other news, pockets full of phones and wallets also "remove" body-hair locally...
Spam levels way down
A couple of days ago I checked the serverlogs: over 1 week, the incoming mail was down by a factor of 8 compared to the same week last year. 280-ish emails compared to over 2000 per week the previous year. All but a dozen or two of those are spam.
Over the past year or so I've found that my low-tech procmail filter, honed over almost a decade, has needed less and less update-work (now only once every couple of months) to keep the passed spam down to a minimal level.
Would it be premature to call Peak Spam?
I got added to a spam list after buying from an online cycle retailer
About 3 years ago I bought cycle spares (probably a chain and or sprocket set) from a uk online retailer (Spa Cycles IIRC) using a unique-for-them personal email address. A month or two later I started receiving spam to that address - so clearly my email address leaked somewhere. Fortunately there wasn't any related fraudulent card activity.
Gently moving customers on
3-5 years ago Orange started signing up pay-as-you-go customers to a "free evening calls offer" where you got 600 minutes of Orange-to-Orange evening calls each month that you topped up by 10 pounds or more. Most PAYG customers who had it are finding that this offer is now expiring (a few months either side of now), so will be looking at other options. It could be a different PAYG tarriff, but in my case, I went to a SIM-only contract (I reckon I'll be making more calls, and more daytime calls, for much the same monthly spend). This also means that the other 2 (or maybe three?) Orange PAYG SIMs I have kicking around on different tarriffs will probably begin to fall out of any use.
Rising or falling numbers of PAYG customers is probably almost meaningless, given you get SIMs bundled with non-contract phones etc etc.
Voice Activity Detection
And another thing: the phone transmits far less if the phone is in a silent or near-speech-free environment. Only when you start speaking does it transmit at full-pelt. If the "user" was silent for the test-protocol as suggested by Robert Carnegiem then this wouldn't be a realistic test anyway.
If you want to see how this works, listen to the interference made by the phone when you are silent versus talking (but take care that the phone microphone doesn't "hear" the audio interference as this will create a feedback loop and upset the experiment).
For those of you without programme-guide...
eg with WInTV NOVA S USB2, the BBC1 HD hasn't got it's name field correct, and the programme name is marked as the ID, ie 6941. It's on the same transponder as BBC HD channel, so with probably be found one channel up from BBC HD.
JPEG vary enormously anyway
JPEG specifies the encoded data stream, and how to decoded. Exactly what psychovisual model you use to reduce the data is up to you. The very best modern JPEG encoding today (Adobe Photoshop "Save for Web" is pretty good) gets a far better ratio of real quality to byte-size than the coding of 10 years ago.
Even if the new algorithm is better, that's not to say that a lot of the images on the web couldn't be made smaller and/or better by re-coding from the uncompressed source using a newer JPEG encoding algorithm.
In my experience of 3G, its the compression steps applied by the mobile operator which SLOW DOWN down the loading. That and the 100ms+ latency.
I already see the evidence
I use one of the AVs mentioned (and not one that gets any regular bashing from the folks on here) and am becoming increasingly aware that I receive obviously malicious email which the AV doesn't recognise as such until 5-24 hours after I received the email.
Then again, seeing as my (personal) server receives one spam every 6 minutes on average, 24/7 I probably get to see more malicious stuff sooner than most....
I'd probably see a lot more malicious stuff pass my AV if I didn't already have my own generic filters in force upstream which sends 97% of incoming mail to /dev/null
Hmmm... infra-red remotes?
It's a shame you've not posted any more details. As a technology, pulsing LEDs rapidly to communicate data is clearly nothing new, although the application of using domestic lighting as the source most certainly is. Reading between the lines, with the bit about eliminating everything from the spectrum except blue, I assume you're meaning filtering the receiver. This makes sense if using 'white' LEDs, which are actually a blue LED with an integral yellow phosphor. We could reasonably assume that the phosphor is "slow"; therefore by filtering out "all but the blue" (i.e. the phosphor emission) the receiver will just see the direct blue LED light and modulation and achieve hence maximum data rate.
What was the uplink channel?
Bottles... pressurised... CO2
"12,500 soda bottles ... pressurized by dropping 12 grams of dry ice into each one and then sealing it. After the carbon dioxide reverts by to a gaseous state, the ultra-light container is virtually unbreakable." (page 2 of article)
Maybe, but won't the CO2 gradually escape over many months? After all, even an unopened bottle of fizzy pop seems to have lost quite a bit of its fizz by its Best Before date.
Keep a sense of perspective...
In the free RAM plot (page 4) you're showing a range of improvement of up to 20MB in 300MB. It's going to be a rare edge-case for which this makes any real-world practical difference.
Worse, the variation in registry sizes (page 5) you show is 170kB in 47MB, barely over 1 part in 300 (0.36%). Which I can't believe will make a blind bit of difference.
Now, for comparison, show us the performance of the original system before you'd disabled anti-virus. Depending on your AV, I wouldn't be surprised if that used 100MB+ of RAM, increased boot-times by 30% or more, slowed Windows Explorer browsing of directories with 1000's of small files by a factor or 2 or 3... I'm not suggesting Joe average shouldn't use AV, but using a resource-light product makes a far bigger difference than any of the "tune-up" software you've looked at.
Disabling Window's Fast Find or whatever they call their disk-indexing service is normally one of the first things I do with a new system - which makes things more spritely and gives the poor hard disk a break.
All very well having the low-level hardware in place
And all credit to the RF engineers etc.,
But what's the point if the mobile operators then stick flaky http cacheing / image compression on top which frequently leads to web-page load-times measured in the tens of seconds (even with a basic 1.5Mbps "bulk data" throughput)... and corrupts ZIP downloads to boot?
(Yes, I'm a frustrated mobile-broadband user)
Why the rush to get rid of FM?
The plan, according to the report is to put ultra-local stations on the existing FM band as a temporary measure.
What is to happen to these ultra-local stations after the temporary period? DAB works with "bundles" (ie multiplexes) of stations and is inherently unsuited to ad-hoc geographical configurations of small-time local stations.
Of course, the centralised infrastructure required for DAB broadcast networks suits this government's command-and-control mentality.
If they're occupying the FM band with local stations, they can't re-sell the spectrum for other purposes (and the ultra-local stations won't have the budget to pay huge licenses).
And while I acknowledge that there is demand for more stations in the major conurbations -London in particular- than FM has space for, why should the rest of the country where the FM band is not overcrowded have their existing stations forced off FM purely on a government whim?
The graph on page 93 of the report showing uptake of DAB seems to have the labels transposed on the extrapolated curves for optimistic ("organic growth" 56% by 2015) and very-optimistic ("drive to digital" 68% by 2015).
From the actual data shown, I think more reasonable interpretation is that the market for DAB is saturating and may just asymptotically approach 30% share of radio listening by 2015.
The report marvels how many DAB digital radios now consume less power than a low energy lightbulb. That would be about 11 watts. Plenty of analog radios with a small speaker run for tens of hours on a couple of puny AA batteries, or for several hundred hours on earphones (about 25milliwatts).
Fail, fail, fail, and fail again.