Req: (Free) Embedded Platforms for Education

How dare you! Media studies graduates serve a very useful function in society and without their invaluable burger flipping skills then there would be famine in my household.

Must...Control...Fist...Of...Death...

"Tom Lucas" wrote in message news: snipped-for-privacy@iris.uk.clara.net...

Well you're talking to a vegetarian so...

...ah, but my nephew is a thespian... but hang on, he works in a restaurant by night.

I don't think I've proved anything - oh well!

P

In article , Peter Dickerson writes

The US rarely has the one true definition.

If you are referring to C the definition is ISO. In fact it took ANSI about a year to catch up with the rest of the world.

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
/\/\/ chris@phaedsys.org      www.phaedsys.org \/\/\
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/

Chris, I think you should introduce a bit more transparancy into your posts. Statements like those above should be followed by something like this:

| | DISCLAMER / FULL DISCLOSURE: I make my living by competing with | Free & Open Source Software, and thus my opinions may reflect a | certain amount of bias towards the kind of products that I sell. |

--
Guy Macon

On Jul 19, 4:09 am, "Tom Lucas"

In USAian parlance [as far as I have learned! remember I am an alien too] grad school means

I also have a lot of trouble explaining to USAians that a "public school" is a privately operated school, and that it is not what they call a "private school" but something closer to a "grammar school".

Only derived work is bound by the same license terms. If you build an application using Linux as your OS you're free to do with that application whatever you want. You have to provide source code for all the changes you made to the OS itself, but then it's in your own best interest to have those changes in mainline anyway.

Code that comes with no license at all basically means that you can't use it at all. It has to be explicitly placed in the public domain, or you can't do anything with it. Source code from silicon vendors is often a horrible example as far as licensing is concerned. What good is sample code that states in its header that all rights are reserved, without a license telling me what I'm allowed to do? What makes this situation worse is example code for silicon vendor X's chip written by compiler manufacturer Y, often with headers stating that you can only use this code within Y's development environment.

FreeRTOS is licensed under a modified GPL (using exceptions as defined in GPL §7) to allow use of the FreeRTOS code in commercial applications without having to open the code to the actual application that merely uses FreeRTOS through the defined API. eCos comes with a similar GPL + exception licenses since it was relicensed in 2002.

I couldn't find anything definitive on uc/OS-II licensing. The website says that you should contact Micrium for licensing information, the source code download is equally sparse on details. The website further says that educational use doesn't require a license, but I wouldn't know for sure what I'm allowed to do with it and what not.

You told us the GPL was complex? ... at least I'm allowed to read it.

I was almost expecting that reply. Could we agree that lots of people (except you) use networking gear that's running Linux, even if they couldn't care less about FOSS?

Heh, just another reason to avoid anything less than 32 bit ;)

Well, of course someone has to write that software - but in most cases these people benefit from the work others have done. Chances are they're using a lot more free software they didn't write than what they're going to write themself. Isn't software great? Once written everyone can use it, if you give them right to do so... And if you're doing it right, noone's going to take that code and run with it, making money from your work without giving back.

Best Regards,

Dominic Rath

Ask them. I am not being sarcastic. Not all software is or should be in the public domain. There are companies like silicon companies that invest a a lot in the development and support software of the products they sell. They are doing that because without some form of return they cannot afford to invest in new technology of take many of the risks on low return from products. They don't want to give their development costs away.

The choice is obvious comply or don't use the code.

The real question is if I want to make a living developing innovative software and the software is put in the public domain for free how do I eat? Does it become a question of doing the innovation or doing something else? If that is the case then development tools are doomed to eventually all look like what GCC has become, old technology.

The software technology we develop is for sale it is the only thing that we do. Our customers share in the cost of this innovation by spending a few man days equivalent for a copy of it. The companies and individuals who determine that a copy of the new technology will benefit them more than they pay for it are our customers.

Call or email Jean Labrosse directly for licensing information on licenses if you want to use uc/OS-II some something other than educational use. Make a choice after you have all the relevant information you need.

I am not sure what the last sentence means in the context of the paragraph.

Walter..

That was a direct reply to what Chris said about non-FOSS software that is still free and could be used in a teaching context. They are of course free to license their software in which way they want, but source code that comes without any license telling me what to do or not is useless. Having asked silicon vendors on both technical and legal matters I found it hard enough to get answers to technical things, let alone on legal stuff. I doubt it would be much of a problem to put a LICENSE[.txt if needed] into the archive that tells me what they want me to do with that software and what not.

The problem is that a user who writes software for some product using example code that is bound by these license terms is locked into using Y's development environment. For a user who's aware of this problem, the recommendation to look at non-FOSS code that's bound by such restrictions isn't much of an option.

Again, my point was just that including information on the licensing somewhere easily accessible isn't asking too much. An exceptionally good example is FreeRTOS. The "License and Warranty" page clearly lists your options for licensing the software, and explains the implications of these choices.

The "it" that you could do "right" is "give them [the] right", i.e. the licensing, and choosing which license to use. I was referring to the difference between placing software in the public domain and using a free software license that asks users of your code to give others the same rights they were given.

There's certainly a place for both free and proprietary software. But I replied to someone who stated that FOSS wouldn't be the right choice in the OP's teaching context - which in my opinion is just plain wrong. I pointed out the shortcomings of some of the alternatives he offered, and tried to show that his concerns about FOSS ain't that much of a problem in the OP's context.

Regards,

Dominic

In article , Guy Macon writes

So should you.

You are wrong. I supply free software.... I have written FOSS in the past.

I am no more biased than any FOSS devotee.

I get really fed up with this sort of stupidity where it is assumed that any FOSS devotee is a saint and has exclusive rights to an unbiased opinion and anyone involved in commercial SW is always biased and distorting the truth. It just isn't true.

I have seen, in several places, from industry commentators (who don't supply any software free or otherwise) comments like:-

?...the open-source community needs to get over its overweening sense of superiority and messianic inevitability; the alternative is just good enough that if it doesn't get its act together, open source may find itself the subject of retrospectives like "Remember Unix?"?

On the whole I have found FOSS Devotees have a far more warped view than anyone else.

From personal experience most FOSS Devotees should end their posts with

| DISCLAMER / FULL DISCLOSURE: ] I am a FOSS Devotee and | therefore have a clouded view of reality. I believe ANYTHING from the | FOSS camp and NOTHING from anyone involved in Professional SW

I am no more biased than any FOSS devotee

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
/\/\/ chris@phaedsys.org      www.phaedsys.org \/\/\
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/

The problem is that both the silicon companies and companies like Micrium have negotiate licence agreements based on requirements from both sides. Many silicon companies waive license fees for their customers. What they don't want is to finance the development for knock off competitors, licensing sets the bar a little higher for their competitors and levels the competitive playing field.

I stand by my comment. There may be many reasons that why code is tied to specific tools. It is common for tools to utilize product information that is not going to be publicly available. Silicon technical marketing plans may very well be incorporated into tools when they are developed. The code generation tools are responsible for generating code for the silicon including dealing with the underlying silicon issues. Silicon companies and parterning tool vendors work together to support customers often making joint marketing presentations and both participate in engineering reviews with common customers.

I followed the links on the page you referenced and eventually reached the following in OpenRTOS section

" To obtain more information, discuss your individual support requirements, or enquire about purchase options and packages, please contact . . ."

Lots of licensing open questions in this document. That of course is not your point that public licensing information is not that clear.

The arguments that have been used for FOSS need to be critically tested as well. Technically they don't come out that well, the licence examples with the exception of GPL and one or two others are a mess. Commercial licences are usually set up as contracts between two entities.

FOSS does have problems in a teaching environment. The two most serious problems are it often does not represent current best practices and secondly it generally was not written with education in mind.

The tools that have been written specifically for education have done a lot better job of educating students. Some examples over time have been WATfor, UCSD pascal, XINU, most of Wirth's languages, Andy Tanenbaum's offerings. The development of these products has essentially ceased with the demand for zero cost no strings attached tools especially for undergraduate fundamental education. There is a reason that schools that have continued to internally develop educational tools in the area are sought after.

Regards

Walter Banks

-- Byte Craft Limited

snipped-for-privacy@bytecraft.com

I have just been looking back through the posts on this thread and realized that there is a fundamental thing missing in the conversation. That is an education requirements list for tools. The initial post stated the requirement was no cost no strings attached and that is where the conversation has centred.

The original poster has gone on to post a curriculum for the course. Lets take a shot at extending the curriculum to a lab requirements list and match that set of requirements to what is available and what is needed.

Walter..

----------------------------------------------

snipped-for-privacy@lirmm.fr wrote:

You are right the goal isn't clear. I realized this as i started to detail my response. I can say that if the focus is on Software Engineering, then the details of which hardware platform used in the lab is less important than the software development concepts taught. At this point, I would only suggest that different platforms are available (processor like

8051, intel x86 family, ARM.) for the lab assignments. Which department is this program in? As an example, I have a Masters in Electrical Engineering (Software Option). Embedded systems is certainly in that border area between computer science and electrical engineering. Some more details might be helpful.

I am glad to see someone else who thinks the option of building an embedded system with an underlying OS should not easily be overlooked. There are many choices as you point out. A good software engineer should have a basic understanding of when not to use an OS, but the general case, especially with todays hardware, is that a Operating System should be part of the product solution.

yes, Please tell us more about the program. Ed Prochak

Chris Hills posted news: snipped-for-privacy@phaedsys.demon.co.uk with Subject field entry: "Re: Req: (Free) Embedded Platforms for Education", which Guy Macon responded to with news:EvWdnRLFMajRFgLbRVn snipped-for-privacy@giganews.com with Subject field entry: "FOSS Bashing", which Chris Hills responded to with news: snipped-for-privacy@phaedsys.demon.co.uk with Subject field entry: "Professional SW Bashing".

Some pe |--------------------------------------------------------------------------| |" In article , Guy Macon | | writes | |[..] | |>| | |>| DISCLAMER / FULL DISCLOSURE: I make my living by competing with | |>| Free & Open Source Software, and thus my opinions may reflect a | |>| certain amount of bias towards the kind of products that I sell. | |>|" | |--------------------------------------------------------------------------|

What would that achieve? Anyone could easily check it. Should he mention his phone numbers in every post as well? Persons' livelihoods can completely depend on "Free & Open Source Software" (unlike Chris Hill's source of income) even if these persons "Free & Open Source Software" activities depend on competing against other (or even the same) "Free & Open Source Software" (for example Red Hat, Incorporated competing against Canonical Limited (which is a company responsible for Ubuntu)).

|--------------------------------------------------------------------------| |" | | You are wrong." | |--------------------------------------------------------------------------|

Guy Macon is probably a person and is definitely not a Boolean value. Is it untrue that Chris Hills does not compete against "Free & Open Source Software"?

|--------------------------------------------------------------------------| |" I supply free software.... I have written FOSS in the | | past." | |--------------------------------------------------------------------------|

Chris Hills had seemed to be unaware of The Open Source Definition in May 2007 judging from his participation in the general email list of the Association of C & C++ Users. Perhaps Chris Hills wrote Free Open-Source Software in the past, but I have reason to doubt that he has learnt yet what the meaning of the term Free Open-Source Software is, in which case I am unsure whether his claim is true. Granted, Berkeley Software Distributions had been open-source before the term was coined so Chris Hills's claim could be true.

|--------------------------------------------------------------------------| |" | | I am no more biased than any FOSS devotee. | | | | I get really fed up with this sort of stupidity where it is assumed that | | any FOSS devotee is a saint and has exclusive rights to an unbiased | | opinion and anyone involved in commercial SW is always biased and | | distorting the truth. It just isn't true." | |--------------------------------------------------------------------------|

Agreed, of course it is not true. Did anyone in this discussion indicate otherwise? Is it possible for no one to be both a "FOSS devotee" and "one involved in commercial SW" at the same time?

|--------------------------------------------------------------------------| |"[..] | | | | On the whole I have found FOSS Devotees have a far more warped view than | | anyone else. | | | | From personal experience most FOSS Devotees should end their posts with | | | |[..]" | |--------------------------------------------------------------------------|

Are you aware of "FOSS Devotees"?

Sincerely, Colin Paul Gloster

I have to say that "FOSS" remains a vague term for me. It doesn't carry exact and specific details in my mind. It seems to me to simply be a huge rug under which all manner of good and evil is swept. So anything I'll say here must take my own lack of precision on this term into account.

The GPL is quite specific, in contrast to my muddled picture of FOSS

-- as you point out.

When you talk about "commercial licenses are usually set up as contracts between two entities," what are you meaning to suggest, here? It's been my experience this is, by default, the well worn legal ground of this kind of commercial business, generally. So there is no new information here, unless you intend it to be read in some particular way I'm missing. Could you elaborate about why you made this comment in the context of both FOSS (which is vague) and GPL (which is not?)

"Best practices" is not a single, decisive point, though, Walter. For one thing, developing a thorough and well-worn (tested) curriculum requires ages of serious effort. It simply takes a lot of time to do, and test. Some folks even refine their testing, in some subjects such as mathematics and reading, so that the __order__ of the questions is itself tested for how it helps correlate positively.

We haven't yet heard back from the OP about many questions which remain, but there is also a huge deciding difference here between the case where they wish to simply teach to use as a library an existing operating system, selecting one or more commercial systems for that, and where they wish to actually teach to build one. Personally, I think that embedded graduate students should actually build one. It deepens the understanding like little else can. But if the OP's desire is to teach familiarity with commonly found commercial operating systems, then there really is no question at all.

"Best practices" is also vague, Walter. What's "best practices" in general purpose commercial operating systems, such as Vista for example, has very very little to do with what is "best practices" in the embedded world. I can, without risking too much, say that "embedded is not Windows." In fact, the skills required are exactly as the OP has in fact said, that there exist enduring concepts, long known, long tested, with clear and meaningful values in embedded and that these enduring concepts are what should be taught. That the OP can then be better enabled for understanding the myriad complexity of the real world by being fitted well with the key and central concepts involved.

Finally, it's not so much the particular piece of software at the school, but the teaching staff which makes a school world class and best able to prepare their students. You want students who can think for themselves and push the envelope and create the next "best practices." That gets back to understanding well the enduring concepts. This is true for both embedded and otherwise.

XINU and Minix. And they are excellent for what they bite off.

Yes, although I'd have to admit that these are not necessarily where you'd learn to understand the internals of Vista, for example. But for embedded grad students, I'd imagine that XINU and Minix are each excellent in their way. I actually believe that any embedded student should have fully read, page by page, the first XINU book and have developed their own operating system from scratch without using the source code available for XINU. It's not hard to do.

If I understand this last sentence, you are right. It's fairly obvious that the leading schools will be somewhat heavily involved in research. If they don't have a healthy research program going on, it's probably not a top school.

Jon

gouaich wrote: ...

Hi,

First of all some course examples:

CTEC1630 Computer Systems Design Home Page:

ECE 476, Designing with Microcontrollers, Spring 2007:

Advanced Microcontroller Systems on a Programmable Chip:

Microcomputer Project Laboratory - Spring 2007:

Autonomous Robotics EECS/BIOL 375/475:

EENG 350 Robot Details:

EE 227 Mobile Robot Details

-

Please do not encourage use of PIC16F84(A) - it is outdated and costs more than better chips according to

. Use PIC18f1220 or 18F4550(has USB possibility) instead - or at least PIC16f628A, PIC16f648A, PIC16f684A or PIC16F877A.

Starting with PICmicro controllers intro, first steps, tips, links, etc. (C) 2002 Wouter van Ooijen ( snipped-for-privacy@voti.nl)

PIClist:

PICList Projects:

Open source language JAL (GPL):

"...Jal is a high-level language for a number of Microchip (TM) PIC microcontrollers (16c84, 16f84, 12c508, 12c509, 16F877 [and 18f452, 252,

242] - found in codegen.c)...":

Open source (dis)assembers:

-

ARM processors has a long it history and are very used and most ARMs are supported by Linux:

Quote: "...The ARM design was started in 1983 as a development project at Acorn Computers Ltd. ... The core has remained largely the same size throughout these changes. ARM2 had 30,000 transistors, while the ARM6 grew to only 35,000. ..."

Welcome to the ARM microcontroller Wiki!

ARM controller overview:

-

Microcontroller independent:

-

Evaluation/development boards & links:

Please look at OpenMoko. The software is called OpenMoko - and the it runs among others on the open hardware; Neo1973:

From october 2007 is a newer version available:

ARMs:

Samsung S3C2440A (ARM9) Board, RS232, USB, Ethernet, IDE, SD/MMC:

Atmel Introduces the World's Lowest Power 32-bit Flash MCU With Ethernet and USB On-the-Go

citat: "... The AT32UC3A0512 and AT32UC3A1512, the first devices available, deliver 80 Dhrystone MIPS (DMIPS) at 66 MHz and consume only 40 mA at 3.3V. ... The EVK1100 evaluation kit provides Ethernet and USB interfaces, along with many other serial communications ports such as SPI, TWI and USARTS. A 20x4 character LCD and the expansion connector allow advanced product evaluation and prototyping activities. Availability and Pricing. The AT32UC3A0512, with EBI, is available in a

144-pin QFP package and the AT32UC3A1512, without EBI, is available in a 100-pin QFP package. Pricing starts at US $8.16 and US $7.43 for 10,000 unit quantities, respectively. ..."

The SR4 Autonomous Mobile Robot:

Citat: "...As an Educational Platform...The SR4 features the Linux operating system, Java, the Medallion single board computer with an ARM microprocessor..."

Glenn

"gouaich" skrev i meddelandet news: snipped-for-privacy@x35g2000prf.googlegroups.com...

If you want to participate in the Atmel University program, you can register at:

Atmel will from time to time supply qualified Universities with free set of kits, typically a dozen or so.

The AVR is a good example of an 8 bit RISC processor which will be able to run small RTOS like

There is also the AVR32 which will run Linux and the AVR32 Gateway kit is dirt cheap. (< $100 in single qty,and university discounts probably apply)

Atmel also have a range of ARM processors, developed in Rousset, close to Aix-en-Provence. There are both ARM7 and ARM9 kits available allowing both RTOS and WinCE/Linux applications. One recent version (AT91CAP9) has a dedicated port for communication with an FPGA.

--
Best Regards,
Ulf Samuelsson
This is intended to be my personal opinion which may,
or may not be shared by my employer Atmel Nordic AB

Hi,

It might also be worth to have a look at FPGAs:

FPGA High Performance Computing Alliance: Technology:

20 March 2007 Green supercomputer is 'go' in Scotland: Quote: "...in the next two to three years - FPGA tech could really take off..."

e.g. BYU JHDL, Open Source FPGA CAD Tools:

8/31/2001, Lab to offer open-source Java-based FPGA tool:

This type of hardware could be very useful for robotics:

Elphel reconfigurable cameras:

Main page:

Imaging solutions with Free software and open hardware:

Just another FPGA+microcontroller example:

The Alya Project:

Quote: "...The Alya board mainly consists of a PIC16C64 ["C" - can only be flashed once (might only be UV eraseable)] microcontroller and an Altera 10K10 FPGA to handle the data transfer. Instead, I decided to use an Altera FPGA and just connect all pins to the FPGA. This is a good example how FPGAs simplify a design - instead of having to understand all the details of the devices involved, just connect them directly and implement the details later. This lead to a

208 pin device and, as I intended to use it only as a wiring, register and multiplexer box, the smallest FPGA available, a 10K10A, seemed to be sufficient. All information about the ATA interface can be found in the ATA spec. You can't download the spec itself, but there are countless drafts at that are as well. I'd recommend you download the ATA-2 draft, it has only 90 pages and contains everything you need for programming. ... Most people I spoke to thought the harddisk interface would be pretty complex and interfacing to it some kind of wizardry. It isn't. ..."

04/23/2007, FPGA tool startup 'rockets' for success:

?articleID=199200568

02/05/2007, FPGA tool bottleneck stalls HPC: ?articleID=197002705 Quote: "... Current FPGA synthesis, placement and routing tools are written for hardware designers, not software programmers simply trying to accelerate an algorithm. ... "We were casting about for ways to reduce heat in our data centers, and FPGA technology seemed like a good bet." ..."

-

FPGAs are among others used in high performance Layer 4-7 network equipment and harddisc (real hardware) RAIDs:

How to choose the right RAID level:

Quote: "...If a RAID offload or accelerator engine (chip, ASIC, FPGA) is being used..."

Designing a Breakthrough Line Card to Transport GbE Data Over Optical Networks:

Citat: "...The GbE muxponder card combines the Intel® IXF18102 and IXF30005 family of transport optical components, an Altera Stratix* 1S80 field-programmable gate array (FPGA)..."

kind regards,

Glenn

It is perfectly fine for a silicon vendor to apply a non-free license to code he supplies to users of his chips. But code that comes without any license is unusable, and from my experience getting information on legal topics is difficult if not impossible - support personell rarely wants to risk giving any definitive legal advice, and sales isn't interested in dealing with little or no revenue issues.

I never meant to imply that public domain would be better, I just wanted to acknowledge that there could be code that has no "license" but is still useable.

A good example in my opinion is the NicheLite TCP/IP stack for use with NXP LPC23xx/24xx uCs. Before the download link they're asking you to read the software license agreement, which clearly explains what you can, and what you can't do. Of course you can think what you want of those restrictions, but at least you know them.

I understand the reasons why such code is licensed the way it is, but that doesn't make the code any more useable if you don't want to risk being locked in with that one tool vendor.

I didn't care enough about the non-free variants of FreeRTOS, so I never went to look at the OpenRTOS pages, I only looked at the "License feature comparison" table, but yeah, you're right, the site really isn't that clear on the non-free variants' licensing. But then, apparently I have to acquire a license before I get the OpenRTOS code, and I can only guess that during the process of acquiring that license I can also have a look at it (the license). On top of that, I can use the available GPLed code in-house for whatever I want, it's only when it comes to distribution or when I need support that the licensing options matter.

That's why I'm strongly in favor of FSF approved free software licenses, especially those that are listed as GPL compatible. But of course in that case you could probably just use the GPL as is.

Yeah, we all know that Linux is obsolete because monolithic kernels suck, and microkernels are the next best thing (tm)... so much for "best practices" and reality. Seriously, I don't see where free software is any different from non-free software in that regard. The advantage of free software is that it allows in-depth comparison of different approaches.

Searching for XINU (I've heard the acronym before, but never had a closer look at it) brought me to the Wikipedia article that had a reference to "Embedded XINU", and the LICENSE file from that download seems like a very permissive free software license. No idea how a demand for "zero cost no strings attached tools" could have lead to the disappearance of such software.

Ulf Samuelsson wrote: ...

Hi Ulf

The AT91CAP9 looks very interesting. Is it possible to buy it now? If that is the case how much does it cost (e.g. 1/2 metal gates)? If it is not available yet, then aprox. when?

kind regards,

Glenn