What is benefit of buying a Pi rather than Linux Partition.

The point is that if you have a linux partiton, and you let your kids play with it. 

1) Your kids can still do   dd if=/dev/zero of=/dev/sda  and wipe your data. (even windows data)

2) you only have one computer.  So you can't read the documentation on your normal computer while playing with the Pi

3) You can buy multiple (cheap) SD cards for the Pi.  Boot off different cards for different purposes

4) You don't need to spend time shrinking the windows partition before installation.

I have a laptop with a dual-boot, Linux and Windows. I leave it in Windows because it's used by me, my wife, and my preschooler for various things they're used to doing in Windows (email & websurfing, and kid's videos on YouTube- haven't got the latter working properly in Debian yet).  Switching the OS requires rebooting which takes an annoyingly long time; sleep/wake is much faster, so it gets left in Windows.  My other PC has video editing software which I use regularly (got to keep sending cute kid videos to grandparents) and the software isn't available on Linux. I guess I could have a dual boot but there's the same annoyance about reboot time, and having got the system working I don't really want to mess with it.

The laptop uses about 25 watts and the desktop is 90 watts (not including the monitor), so it would be nice to have a Linux PC I can leave on all the time, doing data aquisition etc, at just a few watts it draws less than many A/V appliances take in standby.

I have 2 PC's and a laptop and they are all dual boot and I would only boot into windows to play a game or do my accounts. Otherwise it's Linux all the way. Not only do I e-mail, surf the net, watch Youtube and edit videos I also author DVDs, compose music, practise music, mix tracks, edit audio, edit photo's, manage large catalouges of images, build websites, create and publish documents (as in desktop publishing), build servers, write letters, mail merge, create spreadsheets and play games. To buy the software to do all this on a wondows machine would cost me thousands and would need to be upgraded/replaced every 4 years as the built in redundancy makes it obsolete. My friend from where I sit it seems you've pitched your tent in the wrong camp.

As for the R Pi I can't wait to get hold of one and put it to the test.

BTW if you haven't got youtube to work it's probably because you don't have flash installed.

Tony writes:

> ... compose music, practise music, mix tracks, edit audio ...

I have lots of MIDI gear here, and one of the many (far too many) applications I'm planning for Raspberry Pi is a networked Soundfont player and synth to embed in one of my MIDI master keyboards.

It wouldn't surprise me if MIDI projects based on the Pi become quite popular.  Musicians rarely have much money, but $25-$35 could come out of a lucky weekend's busking.

Tony writes:

> ... compose music, practise music, mix tracks, edit audio ...

I have lots of MIDI gear here, and one of the many (far too many) applications I'm planning for Raspberry Pi is a networked Soundfont player and synth to embed in one of my MIDI master keyboards.

It wouldn't surprise me if MIDI projects based on the Pi become quite popular.  Musicians rarely have much money, but $25-$35 could come out of a lucky weekend's busking.

I wouldn't describe myself as a musician I just dabble because the tools are are freely available, I make pleasant noises to run as background music but I see the use of a R Pi as a synth. Plug in your midi via usb through Jack and let something like Qsynth pipe out the audio BUT I need one to test it on although I'm sure it will work.

@Tony: That's how I play Soundfonts currently on my Intel multicore boxes, with Qsynth as the front end and fluidsynth as the backend doing the actual playing through Jack.  The kernel has to run in realtime mode though for Jack to work properly and not suffer audio overruns and poor MIDI response latency, which I forsee will be a problem with the much slower and single-core ARM of the Pi.  Realtime kernel is probably not the way to go on the Pi, as it barely works adequately on much beefier machines with multiple cores.

Instead, a better plan would be to integrate a dedicated DSP microcontroller (the STM32F4-Discovery board is a possible candidate) with the Pi at hardware level, and let each of them do what they're best at, the micro at DSP and realtime response and audio playback, and the Pi for high-level non-realtime duties and overall control.

Well we shall see what it can do once we actually get the devices. For myself I am not interested in projects that require bolting on additional hardware as the exciting thing for me about the R Pi is the low cost so if it can't do it then I'll shelve that idea and work on another.

Well that's the beauty of microcontrollers, they cost next to nothing --- the STM32F4-Discovery board costs under 10 pounds from Farnell UK, and in some countries they're giving them away free as promotion.

Also, the whole point of Raspberry Pi providing GPIOs, SPI, and I2C is so that people can interface other hardware to the board directly, without fear of destroying Dad's costly family PC.  Creating a generation of software engineers alone isn't enough.  The Pi provides those low-level hardware interfaces in order to spark interest in electronics too, so that hopefully we gain more hardware engineers and perhaps create more companies like ARM further down the road.

Of course it's not everybody's cup of tea, and I'm not saying that you should do it.  But the opportunity is there for those who are interested.  Also, that's the right way to handle realtime problems of this kind.

I wonder if it a coincedence that this months Elektor has begun a series in regard to a Linux Embedded board using the NXP Arm chips - not the same of course as the fruity Pi, but in a similar vein.

Point of the pi is, IMHO not the GPIO/SPI/I2C connections. The point is a cheap platform that is "indestructible" from a software point-of-view. This makes it possible for schools and parents to let their young children play with it.

If you want to play with SPI and stuff, get an arduino or http://www.bitwizard.nl/catalog/product_info.php?products_id=29

@Roger:  No.  People aren't going to stop playing with the Pi's GPIO/SPI/I2C just because you say so.

The point of the Pi is education.  If part of that point were not hardware education, then the board wouldn't provide any exposed GPIOs, SPI nor I2C at all.  Providing those connections added substantial design and layout pain and also added cost, and what's more it reduced the indestructibility of the board, so it was clearly done in answer to an important requirement or it would have been omitted.

It's pretty pointless to create a new generation of software engineers alone.  Someone has to create new hardware for their software to run on, and if you leave hardware expertise to other countries then you become subservient to them.

The Foundation has certainly done the right thing by trying to spark interest in hardware with this board, not only software.  Despite the low-level connections being limited, they are sufficient because more comprehensive hardware can easily be added externally.  And as I said earlier, that's the right way to do it except when your requirements are very minimal..

Morgaine.

I think the situation is a bit like the parable of the 10 blind men

describing an elephant based on the part they happen to touch,

where each has a completely different description.

Some people see the gpu as the most important feature, and

want to do home theater, but wonder why there isn't better

digital audio output and mpeg2 decode.

Some people see the low power as the most important feature,

and want to run headless web servers, but wonder why the gpu

exists at all, and would like to eliminate the hdmi and rca connectors,

but would like to add a SATA connector.

Some people see the small size and GPIO as the most important

features, and want to do robotics, but wonder why the motor control

is so limited, and wonder why there isn't any A/D or D/A conversion.

Some people want to teach software development, and see low

cost at the most important feature, and see the GPIO as just a

hazard, and wonder why there isn't more SDRAM, and don't

care at all how small it is.

Some people see the RCA video as a life saver, and want to

support in-dash displays, but have no use for HDMI.  Some people

want to use HDMI displays, but have no use for RCA output.

Some people see the 32-but cpu as overpowered compared to

the Arduino and want to underclock to save power.   Some see it

as underpowered compared to 64-bit multi-core alternatives and

want to overclock.

perhaps with sufficient volumes it will make economic sense to

differentiate the product at some point.

Very nicely put, coder27.

And the true engineer will of course see the Pi as a bundle of capabilities for a very nice price, and harness as many of those capabilities as possible with a view to creating the most effective target application.

This is true even in educational projects, and hopefully the Pi will spark enthusiasm also among those youngsters who are intrigued by hardware.  Some of those will then become all-rounded engineers, understanding both hardware and software, and  great things can come from that.

Morgaine.