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.

 

I had a discussion about evaluation licenses last week.  Many commercial tools, particularly compilers have free evaluation versions that are normally time or size limited.  Sometimes you even get a choice.

 

However what many people do not realise is that these licenses usually say you can not produce commercial or saleable products with the tool. That said, certainly for compilers, the manufacturers are never actually going to enforce this part of the license.  It does not hurt them and it gets their tool used besides source code never gets smaller. So eventually, the reasoning is, the user will have to buy the full version of the tool anyway rather than port to a different compiler.

 

However, licensing is not just for the obviously commercial tools.   Free software also has licenses of some sort implied or explicit.  There are many versions of licenses about.  Several full compilers and RTOS that are free to download and use are chargeable for commercial use.  Then there are the GPL and GCC Open Source licenses. At the moment there are some severe differences of opinion in the Open Source community on the new version of GPL.

 

What a great many people do not realise is that whilst you can use the GCC compilers for commercial use without making the source of your application available this does not apply to the library which is a separate entity hence the various GPLL licenses.  The problem is there are many libraries about. Some are GPL and some GPLL.  I suspect a great number of people using GCC are also using libraries that do require the source of the application they are part of to be open source.  So check your open source library licenses.  You could be in for a nasty surprise.     Whilst commercial companies are not going to enforce the no commercial use on their eval tools it only needs one FOSS Devotee to blow the whistle and you might, legally, have to make the source for your current products available to all including your competitors. So check the licenses on all the free software you use specifically the GCC libraries.

 

Some years ago a company, now gone, did a Linux/GCC License tool that would dig out all the licenses of the various Open Source components and organise them for you and let you see the requirements of using them.  I think things are a little simpler now as but you do need to check and remember not all software sources that you can download freely Open Source or otherwise, has a “free to use in anyway” license. In fact most does have a license of some sort.

 

The MISRA teams are getting busy. There is a MISRA Forum on the 20th November in Coventry where, if we are lucky MISRA-C++ will be launched. Also there should be an announcement of work starting on MISRA-C3.   The question is should we reference C95 or C99? Compiler or relevant tool companies can email me with any thoughts on this.  Is there going to be a mass move to supporting 9899:1999?

 

Before we get to the MISRA forum in November we have ESS 07 in Birmingham 17-18 October, larger than last years and larger than 2005 etc. The UK embedded market is moving up again.  The event is worth a visit just for ideas and getting out of the box for a day. As usual stop by for a coffee and a chat (even doughnuts again). Hopefully there will be a little more decorum and in some quarters it will look less like the FHM Show … and that is from someone who is known for not being PC!

 

Being Politically Correct can be and often is counter productive. In a similar vein a new law that came into force in Germany the end of August.  It outlaws tools used for hacking into computers.   http://www.networkworld.com/news/2007/081307-german-antihacker-law-could-backfire.html?nlhtsec=0813securityalert2&  The new law now defines hacking as penetrating the security on a computer and gaining access to the data. There is no requirement to actually copy or change the data.

 

Hackers are apparently defined as any individual or group that intentionally creates, spreads, or purchases hacker tools designed for illegal purposes. Now the problem is “creates, spreads or purchases hacker tools designed for illegal purposes“ Well most network administrators usually have a whole battery of these tools to test their own systems. Not to mention companies developing security software.  They could all face up to 10 years in prison for these offences. 

 

Now what happens if one of your IT/network people goes to Germany with his laptop. For some reason the airport security people look at his laptop and discover his “hacking” tools. With airport security as it is anything could happen these days.  Computer Security people in Germany must be wondering what to do.  It is like banning guns and realising you forgot to put in a dispensation for the armed forces, police, and intelligence services etc.  Of course the criminals will ignore the law anyway.

 

Mind you things are getting worse all round.  Trutex who make school uniforms want fit GPS trackers in their uniforms… strangely the kids don’t like it. It is much the same as the majority of people not wanting road charging because it will mean the government can track cars [people]. There seems to be a lot more control about and it is getting worse.

 

Meanwhile the US has now started using UAVs and robots in battle field. They have cheerfully stated the have managed to kill some suspect terrorists with one… it is the word “suspect” that bothers me. Knowing, personally,  how good the US military are at identifying targets.  It is not the technology but the [miss] use that other people put it to. That really worries me.

 

Whilst we are on miss-use and licenses etc I saw an interesting item by former Boston University Law School professors, James Bessen and Michael Meurer. They looked at patents and the amount of money they made for the patent holders balanced against the cost of defending those patents. Amusingly they found that in the US more money was spent on patent litigation than was made from those patents!  This excluded the pharmaceutical industry where I assume the profits outweighed the costs. The reason why, they concluded, is that US patents are often imprecise.  More to the point I understand that US patent offices grant patents far more easily than the European offices do. Also there are various tests used in litigation such as “obviousness”. Clearly that one was drawn up by the legal profession!

 

Google, Yahoo, Amazon US and seven other companies are being sued by a Texas patent-licensing company that accused them of infringing an invention that interprets and routes e-mail messages. Apparently Polaris say that the companies are using the patented technology without permission. The suite was filed in federal court in Marshall, Texas last month. Polaris, claiming it "will be greatly and irreparably harmed," asked the court for cash compensation and an order blocking the companies from using its patented technology.  Mind you closing down Google and Yahoo might improve the bandwidth on the net!  However it shows how much of a legal minefield it all is in the US.

 

There is some hope as there is move to change the US patent act to make it easier to challenge without an expensive litigation process and it will permit challenges at any time during the validity of the patent not just during the granting process.  It all sounds good until you get to the kicker. They are now coming into line with other countries that there will be a shift to first to file a patent rather than first to invent.

 

This leads me on to the fact that the US has software patents.  If the UK or EU gets software patents due to reciprocal agreements with the US the thousands of US software patents will apply here.  Now, unfortunately I am not a lawyer, but I think the only people who will make anything from software patents this side of the Atlantic are the lawyers. Especially all the US patents will have been filed before any EU ones and the US courts will now look at date of filing not invention…  I have a customer who says they have to defend a patent infringement involving software in the US.   The US patent is dated after they started producing the device in Europe.  However the US Company, who bought the patent from anther US Company at great expense, is going to court and has vastly more money than they do.  Are patents worth the money?  Only if the majority of the world uses the same system and at the moment they don’t.

 

Talking of using the same system a study by the IT Company Dimension Data across America, Europe and the Middle East found that 99.6 per cent of business workers use email compared with 80 per cent who use land lines and 76 per cent who use mobiles.  Presumably this was an email survey.  However it does point to the way business has changed from hardcopy letters and faxes. In fact most of the orders, quotes, documents etc I see are emailed I haven’t sent a fax in years.  The survey did not mention the number of Voice over IP, or internet phones are used though that number is growing too. Well it must be Skype has 220 million more people than it had 5 years ago.

 

However this reliance on the internet may not be as sensible as you might think. I know the internet is “bomb proof” but that is the network not individual nodes. During August 2007 a company I know had a main switch in their office go down and with it their email, order-processing system, Intranet and licence activation system.  It took then 84 hours to get it all back up. They were effectively shut for two days. Also in August Skype went down and for about four days 220 MILLION people lost their Voice over IP.  

 

Apparently the Skype problem was caused by a Microsoft update. No, it’s not Microsoft’s fault this time!  Millions of people have their updates set to automatic and one of them caused a very large number of Skype users to automatically logout and re-login almost simultaneously which hit a  glitch that had been in the Skype software from it’s inception in 2003. The overload brought down the entire system.

 

It reminds me of the USS Yorktown, one of the US Navies super Nuclear powered warships that is complexly computer controlled.  More than once one computer failing could shut down the entire ship apart from the reactor.   We need to take software testing seriously. But how do you test for 220 million people logging in simultaneously?

 

I understand that some 70% plus of internet traffic is now spam.  Again in the UK and I think Europe we have laws for opt in to mail lists where as the US where the majority of the spam comes from has an opt-out law. Though of course none of us in our right minds would open this sort of spam let alone reply to it.  In March I saw an email/news item that said at the weekend the internet was being closed down for reconfiguring.  The end of the week being 31st March-1st April…….  However it did get some people thinking it was not such a bad idea. Starting with disconnecting nodes where the majority of spam originates from. A small tactical nuclear device seemed the most popular method.

 

However the Japanese, (who else?) have seriously thought about re-configuring the net. The Communications Minister Yoshihide Suga said Japan will start researching and developing technology for a new generation of network to replace the Internet. The government would like to get the technology into commercial use by 2020. This seems serious as they have put 7.8 billion Yen aside in 2008 for this project.

 

OK so we have a better internet with more reliable systems less susceptible to breakdowns, viruses and spam… Given the Japanese preference for order it will be a fast and efficient system with controls to stop hackers, spam merchants and the like.  Will the we accept that level of control. More importantly control by whom, a single entity? National governments?  Governments want to tax ebay-businesses etc now.  Are we seeing the end of the last wild frontier as big business and governments take back the social experiment that was the internet?    In 2025 you may be able to amaze youngsters with tales of hackers and how wild the internet was when any one could have a web site with anything they like on it.  When Usenet was free and open, not a tightly controlled web forum. The question is who is going to implement this new system and the controls?

 

Just as we have had a survey to say that mobile phone masts are OK…. Another report actually says that the mobile phones are dangerous… see

http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2007/08/30/nmobile130.xml
This is based on a 25 studies over 6 years. This seems a damned sight stronger than the one which proved the masts were safe…. Unfortunately it shows the problem is not the cooking effect but one which causes a form of cancer without needing the heating effect.