next topic »

[Reply #15 on:]

A second serial port and 2x RS-485

______
Rob

[Reply #16 on:]

I am going to fix the poly fuse issue and use a proper USB power controller chip. I have about 100 rescued from a skip.

[Reply #17 on:]

what would your first hardware ad-on be?

An Arduino.  :-)
Because I don't want by 1GHz-class CPU having to bit-bang I2C (400kHz) out some GPIO pins.

First prize to whoever reconfigures a Rasberry-pi or similar board to be an "ethernet shield" for an Arduino!
(that may be the best way to get advanced networking features like IPV6,  IPSEC, SSL, etc.  And it's cheaper than the existing ethernet shield...)

[Reply #18 on:]

Because I don't want by 1GHz-class CPU having to bit-bang I2C (400kHz) out some GPIO pins.

You will be disappointed if you did.
I have just done some quick tests to see how fast you can toggle a pin on an off with a simple loop. You can only go at 4MHz and it is very glitchie due to interrupts popping off all over the place.

[Reply #19 on:]

what would your first hardware ad-on be?

First prize to whoever reconfigures a Rasberry-pi or similar board to be an "ethernet shield" for an Arduino!
(that may be the best way to get advanced networking features like IPV6,  IPSEC, SSL, etc.  And it's cheaper than the existing ethernet shield...)

Main usage I see is to connect it to arduino and take the ethernet and sd load to the raspberrie.
Sd can take up to 250 ms and downloading a file is taking long. Moreover starting Ethernet with DHCP on my arduino in  my setup takes several seconds. If these delays happen "at a bad moment" my robot is blindfolded and will not respond properly. For a 1000+ euro robot having 2 100 watt engines driving blades at 3000rpm..... ... I prefer to be on the safe side.
Using the Rapsberry to read/write the serial data from/to arduino; Arduino could send my logging information to raspberry. Raspberrie can process the input (get out the business data) and service the Ethernet and do the SD logging.
At the same time Ethernet actions could be translated to Serial commands send to the Arduino. The nice thing about this is that you can test with the shield or with the serial monitor (so not 2 systems but 1)
This setup would take quite some work from Arduino. It would also make your website/webserver config lots easier as it runs on linux on the raspberry. And finally it would free up some pins for other usage. Maybe I can go back to the UNO 

This would not make it an "Ethernet shield" but a "Smart Serial shield capable of SD and Ethernet"
In short "don't expect me to win the first prize."
Best regards
Jantje

I have been able to order my raspberry today. Expected to ship in 4 weeks.

[Reply #20 on:]

I have been playing around with the GPIO pins. For me those are the most exciting aspect of the pi. Guess I will eventually connect some nice ADCs vie I2C.

Yeap I agree with that. For some like me who likes hacking hardware the GPIO are the RPi's most attractive feature :-)

The APC IO looks a great board, but there's no info on what GPIO is got. Connector J3 is a mystery, perhaps that's got GPIO or I2C.

question to you guys: If you could mod the pi (or if you will mod the pi) what will/would your first hardware ad-on be?

Already done my first, second and third addon board for the Pi: Prototyping Board. BBC like I/O User Port. MIDI Interface.
Just waiting for my PCB's to arrive and then my old soldering iron won't know what's hit it :-)

[Reply #21 on:]

Even if J3 does have GPIO there can't be much of it. And the blurb says

Input and Output    
   HDMI
   VGA
   USB 2.0 (x4)
   Audio out / Mic in
   microSD Slot

No mention of GPIO in the IO section so it's not going to be a big selling point.

That said it's the first of these boards that interests me, although the mini-box looks pretty good as well.

______
Rob

[Reply #22 on:]

I am still waiting when I can get a PI

Not me: I'm officially giving up on it. I've had it with their "bleep you" attitude toward potential customers.  After seeing a posting the other day that said they were finally taking orders, I discovered that what they were actually doing was allowing some of the people who signed up months ago to backorder units for possible delivery in a month or so.  And still not a word about if (much less when) they're ever going to start shipping in quantities that would allow people to actually buy them for practical uses.

I've spent more than 2 decades working in technology start-ups, and many more years as an end-user, so I understand the growing pains and glitches that arise in bringing something new to market.  But this nonsense of "Don't call us: we'll call you. Someday. Maybe.." is beyond the pale.

[Reply #23 on:]

its really quickly falling into my vaporware pile, While I dont have decade or even years (yet) in production, just my handful of months in it, there seems to be a lack of commitment in producing the units.

[Reply #24 on:]

You must remember that this is just five people who are amatures at production. World wide intrest has taken them aback somewhat.
You can't just turn on a tap and have product come out. This sort of volume can take up to a year to plan. I have worked for companies that do this sort of volume and more and it is not a smooth process.
At the moment they are trying to meet the demand on a first come first served basis, so there is a queu, one way of handling it is to have a waiting list - registration of interest - and then an invertation to order.
How would you handle 100,000 individual orders on day one, when you only had 10,000 boards in the pipe line?

[Reply #25 on:]

For what it's worth, mine arrived yesterday. I was one of those that 'expressed interest' on launch day (being too late to actually order then), and I got to place my actual order about three weeks ago.

No real idea what I'm going to do with it yet. Not sure whether to try and learn python, or just stick with C. Or just stuff XBMC on it and have a tiny media player.

[Reply #26 on:]

Or just stuff XBMC on it and have a tiny media player.

That is one of the major selling points to me, I have been looking at e-bay every now and then trying to get a media centre for cheap. So when I saw it had HDMI and supported XBMC I was very pleased, except I could have had a media centre months ago.

[Reply #27 on:]

How would you handle 100,000 individual orders on day one, when you only had 10,000 boards in the pipe line?

I would make sure people knew they were signing up for a maybe oneday item instead of selling them in every shop as if they existed, its a Tucker

I mean I go to their website, I see no status, no availability information, no plans, I do see that they think they are awesome and we should vote for them in not one but TWO web awards thingies and that they are going to maker fair to show off a product that is currently unavailable in the markets in which it was released with no word of IF it will hit the states.

Meanwhile other big players are gearing up to make similar products in a similar price range, and potential customers like me are already spending more for other pico systems available in quantity now.

[Reply #28 on:]

You must remember that this is just five people who are amatures at production.

Why? With all the out-of-work, retired, and bored single engineers and production management people out there, why are these 5 amateurs trying to go it alone, instead of recruiting some volunteers with a clue? One would expect that an engineer from a fabless semiconductor company that sells lots of bleeding-edge chips to startups would already know from watching other people's fiascos how essential that "boring detail work" is.

How would you handle 100,000 individual orders on day one, when you only had 10,000 boards in the pipe line?

I'll second most of what Osgeld wrote (although I wouldn't be quite as harsh). Instead of blogging about how totally kewl it is to unexpectedly be a celebrity, I'd spend the time putting together a webpage with:

a. A semi-realtime counter (updated at least daily) of how many people are queued up to buy,

b. a table/chart showing how many boards we plan to ship, and when, and

c. a weekly-or-so update indicating whether we're on track to meet those goals.

That wouldn't satisfy everybody, of course, but it should satisfy enough that you could blow off the rest with a clear conscience.  And come off looking like you're trying to satisfy, and learning how to do it, instead of just playing.

At a total cost that's probably less than the taxi ride from the airport to a Maker Faire.

[Reply #29 on:]

The point is they didn't. Do you want to do better, you are free to try. For the record I think the made a total farce of it as well. I think they were steam rollered into going faster than they planed to. Mainly because the spent too much time stoking up publicity and not enough time testing the board. There are some quite serious problems with the USB and the power.

At a total cost that's probably less than the taxi ride from the airport to a Maker Faire.

I don't know. Have you ever taken a taxi in New York, I was taken aback I can tell you.