Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×

Comment Re:Pi in deployment (careful with max OC) (Score 1) 233

Which is not completely, though largely irrelevant to the question of using them as small PCs or embedded controllers.

I'm sure the Ouya will be able to be used as a small PC.

Ouya has - on a quick glance - no onboard accessable peripheral interfaces.

But what percentage of Raspberry Pi owners will actually use these interfaces? I'd say well under 1%. The vast majority simply don't care.

Comment Re:USB problems continue on RasPi (Score 1) 233

Recent revisions of the RasPi board have improved the power problems substantially.

Not really. They've just fixed the idiotic voltage regulator error with the LAN9512 chip (see http://www.raspberrypi.org/phpBB3/viewtopic.php?f=29&t=14489) and removed the USB polyfuses. The instability problems the Pi has with power supplies which output a low but still within USB specification voltage, such as 4.8 volts, are still present.

The USB revision means that certain devices can now be successfully powered directly from the USB ports on the Pi, whereas before they would malfunction if this was attempted. However, an unfortunate side effect of this change is that merely hotplugging a USB device into a revised Pi can be enough to reset or crash it.

Power problems aren't the only kind on the RasPi though. Probably the biggest problem is the broken USB handling. The USB controller on that particular Broadcom SoC device has severe limitations and makes invalid assumptions about the Linux kernel, and that cannot be fixed because it's hardwired in the device silicon. As a result, even the latest board revisions have major USB problems that make the board unusable for certain applications.

Spot on. The Raspberry Pi Foundation still believe this problem can be fixed though, judging from the comment Foundation representative jamesh made in this thread: http://www.raspberrypi.org/phpBB3/viewtopic.php?f=24&t=23053. I think he's kidding himself, especially when it's considered that the USB issues jamesh apparently considers solved have just been worked around, not truly fixed.

Comment Re:hardware vs software (Score 1) 233

Only the early Pi's were built in China. The newest models are manufactured in a Sony factory in the UK, and I believe that very few if any are still manufactured in China.

You are wrong. A very significant number are currently made in at least two factories in China. Look at what Raspberry Pi Foundation representatives jamesh and liz say in these threads: http://www.raspberrypi.org/phpBB3/viewtopic.php?f=62&t=20388 and http://www.raspberrypi.org/phpBB3/viewtopic.php?f=24&t=23248 and http://raspberrypi.org/phpBB3/viewtopic.php?f=63&t=24057

In fact, the Pi Foundation has stated numerous times that they wanted the Pi to be manufactured in the UK from the start, but at the time this was impossible when considering the necessary quality and fixed selling price with little margin. Thanks to the Sony deal they are now able to do this.

Yes, but have they ever said they wanted to manufacture in the UK exclusively? They may have done, but I haven't heard this.

I have a first revision Pi from China, and it works flawlessly. I even applied the highest stock overclock (1GHz) and it's totally stable without any additional cooling.

But have you stress tested it for at least 24 hours by generating a load which is both CPU and GPU heavy? This is essential, you won''t know whether you have a stable configuration or not until you do this. I get the feeling that many people who think they have a "totally stable" Pi actually don't. Some crashes or other malfunctions don't occur until you have been properly stress testing for hours. Also, watch out for overclocking induced SD card corruption.

Also, note that many of the user problems are due to sub-standard power supplies or using USB devices which draw too much power from the USB ports.

Some of these "sub-standard" power supplies that have problems powering the Pi aren't actually sub-standard, i.e. they maintain the USB standard 5 volts ±5% (4.75 to 5.25 volts) when measured at both of the test points on the Pi's PCB while it is under full load. The only reasonable conclusion that can be drawn from this fact is that it's the design of the Raspberry Pi itself which is sub-standard, which shouldn't come as a surprise at this point.

Remember, the Pi does not come with one, so the user has to supply one or buy a third party power supply. The quality of these vary widely. I purchased one that was pre-tested with the Pi and offered directly by the UK distributor and like I mentioned, I've had zero problems.

Unfortunately even this hasn't guaranteed success. I know of multiple cases where power supplies bought as part of a Raspberry Pi bundle have been unable to properly power it.

Comment Re:Pi in deployment (careful with max OC) (Score 1) 233

Overclocking isn't just about heat (I added a heatsink and the CPU runs cool). The jump from 950MHz to 1Ghz is a very steep one (it suddenly bumps up all the other system clocks by a large amount) and this can make it unstable, corrupting the filesystem. 950 seems to be reliable.

Having the CPU clocked at anything above 840 MHz and/or the GPU/core at anything above 275 MHz has resulted in massive filesystem corruption on the SD card for me, but RAM clocked at 600 MHz along with CPU at 840 MHz and GPU/core at 275 MHz seems fine. Everyone seems to be experiencing something different though.

Comment Re:Call me old-fashioned... (Score 2) 233

This isn't true. The Raspberry Pi is currently being manufactured Sony's UK factory in Pencoed, Wales AND in China.

It appears that the number of Raspberry Pi's shipped faulty appears to have significantly increased in the last month or so with one report on the forums from someone who claims that they recently received three faulty Pi's out of ten and another from a person who claims both of theirs are faulty. These reports have been described by Raspberry Pi Foundation representative who posts on their forums as jamesh as "statistically insignificant". This is the same person who incorrectly claimed in August that the Raspberry Pi was still in beta.

There have also been recent reports of poor packaging with the damaged PCB unsecured and rattling around inside, solder bridges between header pins causing constant rebooting as well as multiple instances where PCB holes have been inconsistently filled with solder. See here: http://www.raspberrypi.org/phpBB3/viewtopic.php?f=28&t=22473 and http://www.raspberrypi.org/phpBB3/viewtopic.php?f=23&t=24571

Comment Half-baked, like the Pi itself (Score 0) 97

This is rather pointless, for all the reasons others have previously stated. But this is really laughable, Eben Upton of the Raspberry Pi Foundation thinks they can wait until 2015 to produce a successor to the Raspberry Pi. Even more amusingly he intends to still be selling the Raspberry Pi Model B in 2020, seven years from now: http://www.itpro.co.uk/644701/raspberry-pi-founder-has-plans-for-a-sequel-in-2015

Comment Re:Wait, what?? (Score 1) 91

The Raspberry Pi as it is today, i.e. underpowered and broken, will never achieve much success in the education sector. The number of Pi's sold to schools so far has been absolutely insignificant. Lack of hardware hasn't been a real problem in those areas for many years anyway and its never been the main problem. The main problem has been and still is insufficient numbers of appropriately skilled teachers. The Raspberry Pi can't solve that.

Comment Re:Availability. (Score 2) 125

It's the newer model with 512MB RAM and screw holes (whoever fucked that up in the first place... it's mind boggling)... but it's not made in the UK. Not sure what that's about, I thought they were supposed to be now.

Some of them are, some of them aren't. There have been multiple reports of the recently Chinese manufactured Pi's having questionable soldering and reliability problems. See here: http://www.raspberrypi.org/phpBB3/viewtopic.php?f=28&t=22473 and http://www.raspberrypi.org/phpBB3/viewtopic.php?f=23&t=24571

Comment Pointless change, more or less (Score 2) 125

Judging from their forums, there's a significant number of people who think removing the Ethernet controller/USB hub chip is going to solve or at least substantially reduce the Raspberry Pi's problems with USB. It won't. Unfortunately, but predictably it doesn't look like the Foundation have done anything to correct this misconception. Isochronous transfers which audio interfaces, webcams etc. rely upon won't be affected much, if at all by this change, i.e. they will still be utterly broken. If you get one of these types of devices working at all with your Pi you should consider yourself lucky.
Bug

Submission + - Faulty Raspberry Pi's

rephlex writes: Along with the ameliorated but still ongoing problems with USB on the Raspberry Pi it appears that the number of Raspberry Pi's shipped faulty appears to have significantly increased in the last month or so with one report on the forums from someone who claims that they recently received three faulty Pi's out of ten and another from a person who claims both of theirs are faulty. These reports have been described by Raspberry Pi Foundation representative who posts on their forums as jamesh as "statistically insignificant". This is the same person who incorrectly claimed in August that the Raspberry Pi was still in beta. There have also been recent reports of poor packaging with the damaged PCB unsecured and rattling around inside, solder bridges between header pins causing constant rebooting as well as multiple instances where PCB holes have been inconsistently filled with solder. See here:

http://www.raspberrypi.org/phpBB3/viewtopic.php?f=28&t=22473
http://www.raspberrypi.org/phpBB3/viewtopic.php?f=23&t=24571

Comment Re:Why is Slashdot to Hostile to Raspberry Pi? (Score 1) 74

Yet you use the quote to argue that they have claimed the Pi is open. The quote has no merit in that regard.

The quote refers to something that was eventually done, so what else could have been meant other than "completely open access to the hardware" was ultimately achieved? Especially considering that this article wasn't the last time someone connected with the Raspberry Pi Foundation has made this sort of misleading claim, see http://www.raspberrypi.org/archives/2221

It's not Open, get over it and move on. Use the board, don't use the board, use one of the totally open ones out there if you want. But don't make claims on the Foundation's behalf, then complain that they aren't matching them.

What makes you think I care? Whether or not the Raspberry Pi is open isn't something that particularly interests me. But claiming it is when it clearly isn't is concerning.

Slashdot Top Deals

What is research but a blind date with knowledge? -- Will Harvey

Working...