These are my own personal views and those of my company Phaedrus Systems. www.phaedsys.org which is where the full version of this column resides under the Technical Papers button.

 

  It’s that time of year again. ESS is bigger than last year and the year before. http://www.edaexhibitions.com/ess/  The conference is getting larger too.  We will be there of course (with the usual coffee and doughnuts) to network and discuss.  A day at the main UK embedded event will be time well spent even if you don’t make the conference.
 
The MISRA-C team is starting on Version Three for release in 2010.   As some of the long standing members are retiring we have some open slots in the MISRA-C working group. Membership requires attending two day working meetings six times a year, and some “home work” between meetings. If you have a background in critical systems in any of the following fields: automotive, rail, nuclear, aerospace, medical or communications please send me an email with your relevant experience to register your interest.  Relevant academic input would also be of interest. Your company will be required to sign the MISRA working group agreement.  
 
On a side note we contacted most of the embedded compiler companies and we are very likely to base MISRA-C3 on a subset of C99 rather than C95. This is because most compilers are now “moving towards C99” and are based on a C99 system. Any compiler or source code verification tool maker we have not contacted might like to give me a call and I can put you on our liaison list. We do occasionally ask a wider audience for opinion during MISRA-C development. There is no point in writing a coding standard if the tools are not able to support it. Something the international ISO C panel should have thought of 8 years ago.
 
The current ISO C standard is 9899:1999 but it has three Technical Corrigendum on it. Although the standard costs money the TC’s are free to download. However for the majority of you the Committee Draft of the standard with the three TC’s rolled in (effectively C07) can be downloaded from  http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1256.pdf

I should stress that this is a Committee DRAFT and is not legally an ISO standard. Those of you who need an official copy of the standard will still have to buy one.  Amongst all the administrative stuff there are other useful documents on the ISO C web site. http://www.open-std.org/jtc1/sc22/wg14/
 
There are some points to clarify. Firstly there is currently no UK/BSI C panel. BSI suspended the UK C panel and the last convenor was stopped under circumstances I can’t comment on as things are still on going.  There were no UK representatives at the April meeting.  I hope to be able to bring an update on the sad state of affairs in the UK C panel before the end of the year.  Though, it does appear that globally MISRA-C is more than making up for the lack UK-BSI/ISO C involvement over the last year and a half.
 
The following are cut from the official notes on the discussion at the end of the ISO C meeting in April (they have two a year) on revising the current C99 version.

 
The Convenor introduced the topic and asked if anyone was prepared to speak against revising. Willem asked for justification for a revision, particularly with respect to backward compatibility. The current language needs cleaning up, not just extending. Embedded systems need less, not more. Should remove floating point and complex arithmetic.
 
Clark's biggest concern about opening up the standard for revision is the shortage of progress that Microsoft has made toward C99 compatibility. What's the point in extending the language if existing vendors haven't caught up with ten year old stuff? PJ spoke in Microsoft's defense; there is no customer demand for some features. We should consider removing some poorly adopted features.
 
There are three classes of things in C: things that have been there forever, work, well, and fully understand. There are some portions that have been added that tidy up loose ends (e.g. _Bool and VLAs). There are other things that just aren't worth it. There is a burning need for a revision based on concurrency requirements from multi-core processors, to fix up some known weaknesses, and to adopt common existing practice.

 
The discussion continued but the interesting points are that the ISO working group have finally acknowledged that the current C99 has not been and is not ever likely to be fully implemented by tools vendors. So we could see a new ISO C standard with things removed!
 
One of the main concerns, and in some places paranoia, in the world is security.  There are threats from criminals, hackers, terrorists and shadows.   Security becoming more of an issue in software and electronic systems and these days any serious bank raid is more likely to be electronic rather than shotguns and balaclavas. The bigger perceived threat is terrorists. Of course one mans terrorist is another man’s freedom fighter and in The Great Game the people we armed and supplied in the last decade are the enemy in this. 
 
In the past few months I have mentioned a couple of times incidents involving US companies refusing to supply (European) equipment to the UK citing counter terrorist export laws This included an NXP ARM7 dev kit!.  In the last couple of weeks I got involved in an interesting situation where a US based tech support desk refused to supply some one with header files for one of their chips. The reason was:-
 
Our system has detected that you may be on the Denied or Restricted Parties List. By order of the United States Government, ********* Inc. is prohibited from exporting or providing services of any kind to any party identified on a denied or restricted parties list.
 
 
Who was and where was this dangerous person?  A student in Moscow! Now you might think using something like hot mail or gmail would get around this…  Well “Dmitri” was using Gmail!  Whilst I queried the UK rep of the company Dimitri solved his problem by downloading, from a non US we site, the current evaluation version of the compiler, not made by the US Company, with the all header files he required. The files in question are legally freely, in all senses, available from several sources outside the US! In fact the header files as far as I know don’t even originate from the US.  Also I have no idea who is on this “Denied or Restricted Parties List”.
 
I am not going to name the company or publish the email because it is clear that this is US government export law and it would be unfair to name a specific company in this as they are just trying to stay on the right side of their own government. As far as I can see ALL US companies will have the same constraints. I think it is just that some are applying them more zealously and blindly than others. Though we all know what happened with Matrix Churchill whey they tried exporting “oil pipes” even with the full knowledge of the Secret Intelligence Service. The Customs people nailed them.
 
To compound the issue Microsoft AKA Spawn of Satan has deservedly, many would say, lost it’s appeal with the EU over its near monopoly in some areas of it’s market.  However if you stop to look and realise this will actually affect other US companies such as Apple and Google  you can see that transatlantic high tech dealings are due for an interesting time.  Particularly with the changes in US patent laws that have been going though the administration. Anyone involved in technology in the US should start here:
  http://www.eweek.com/article2/0,1759,2181469,00.asp?kc=EWKNLGOV091207STR1
 
So you need to look carefully at what you source from the US at the current time.  Not only might you have problems getting support but you might find there are strings attached on whom you can supply your technology to if you develop using it using US technology.  Also you need to keep an eye on the changes with the US and EU patents.
 
This can affect small companies as much, if not more than some of the larger ones.  In this modern world where all points on the internet are the same place and couriers able to deliver to the other side of the world in 48 hours large companies have people and resources to deal wit these problems.  Most small companies and consultants don’t.  Relying on common sense is no help these days.
 
Last month I mentioned the new German laws on hacking would make many tools network administrators use illegal as they are “hackers” tools.  This was brought home when a   Swedish security researcher made public government and Fortune 500 e-mail passwords in late August using The Onion Router, or ToR, exit nodes outfitted with a packet sniffer to catch the unencrypted e-mail messages and passwords.  ToR is a free tool distributed by the Electronic Frontier Foundation that's supposed to allow users to communicate anonymously.  The question is: are the authors of this tool and indeed the EFF now “hackers” under the new German Law?   However this incident also highlights how insecure email can be if not encrypted.   However encrypted email will also attract attention too…
 
 
There have been some reports that India is the place to be and it appears that not only are fewer Indian graduates leaving some are starting to return to India.  Though, just as the Indian high tech industry is maturing it appears that they are having the same problem of China and Taiwan under cutting them for software.  UK engineers need to realise that they are in a Gobal market.
 
And finally I got this email recently from, as far as I can tell, India:

 
message: I'm Graduate(B.Tech) in Electronics & Communications Engineering.. Presently I'm working as an Embedded engineer. Can u guide me in a right way how to pave my path towards a bright future in embedded domain

If any one has the answer for paving the way to a bright future in embedded email me! I am dying to know the answer and can I have an exclusive UK distribution for it?