Tuesday, September 3. 2013
Since Elop took the reins of Nokia the share price has been slashed and divided by 4, today Microsoft is essentially buying Nokia and nominating Elop to run both the existing Microsoft Mobile Business and the Nokia business.
Nokia Communicator 9000 my first "smart phone", when Nokia was still relevant and interesting!
Incidentally rumours indicate that Stephen Elop might be the successor of Balmers.
So ruining a company is supposed to make you the right candidate to run another one ? talk about second chances...
Or maybe it when like this:
In reality it is probably just an instrumentation of what happened, nevertheless I believe the Nokia shareholders have good cause to sue jointly Elop and Microsoft, but I guess they won't and will just be happy to salvage at least a quarter on the dollar from the whole mess.
Monday, July 8. 2013
To be exact I mean : Move yourself where you find opportunities, even if you are a person that happens to be over 30, but as a meme it does not work that well.
And remember who invented the internet, personal computers, smartphones, etc ....
So what do Berlin, Paris, Amsterdam, Brussels, Barcelona, Dubai, Hong Kong, Sao Paulo, ... have in common ?
A vibrant set of creative startups, high standard of living at reasonable cost (although it might imply exchanging a large house in the suburbs and a car for a studio and metro in the center of the city).
If you've got ideas, energy and creativity are not adverse to learn a new language look around, and see what you can bring to the conversation.
Friday, May 3. 2013
Of course I feel for the co-founders, it is certainly depressing for them, as they are apparently living in the US and might have been in line for some "silver lining" (even if Orange has bought 100% of the company already).
This said Privacy laws in Europe and in the US are very different, so are a few other important legal aspect relative to free speech, accessibility by minors, method of copyright enforcement. Since Yahoo wanted to buy them to offer an alternative to youtube it would be quite illogical from their point of view not to move the whole shebang to the US. First of all as an US company they would have to apply US laws anyway, and even if they would try to keep the subsidiary at arm's length any potential US studio partner would insist on it.
In conclusion, if Yahoo would truly believe that Dailymotion is a good investment and a good fit with the rest of the Yahoo services, the proposal of the French government was quite reasonable, a 50/50 share would have enabled Yahoo to merge the user data base (applying safe haven etc....)
It would have enabled them to create a "richer" user experience that they could have used to build up as a beachhead to significantly grow their presence in Europe. The fact that this was a deal breaker implies that they wanted the power to do things that the current managers would not want to do. (in this case not the current owner, Orange would obviously be happy to have a largish cash influx).
Will this discourage foreign investors, probably in some cases, but in most cases the issues will be very different and the main attraction for an investors is not really what the government does or not, but are there enough people wanting to work in a startup ? and from this point of view, I think it was the right thing to do.
I hope it will encourage foreign investors to really think about the right way to handle an European strategy, and not just hope that they can dive in, and run off with the client/user list and some piece of code.
Tuesday, April 30. 2013
In unlikely places interesting things happen, the UMSA the largest state university in Bolivia has started a Master in Free Open Source Software Development in part to make sure that the local government becomes able (if not necessarily willing) to apply its own law that ask them to prefer Open Source solutions when ever possible.
It is an interesting development, to be honest I believe it will mostly help these students to find work abroad, the three main issues for IT in Bolivia are:
Nevertheless, kudos to the community & staff at UMSA.
Friday, April 26. 2013
The genivi alliance does interesting things with Free Libre Open Source software, the basic idea is that the car makers and their component providers looked at what happened in the server market, and thought that since they all need more or less the same infrastructure before they can try to differentiate, why not cooperate competitively...
So let's load a Linux in all the cars, make sure that it is isolated from really scary stuff (who might be managed by some sort of linux too but we ain't telling you, and you probably do not really want your ABS to query yahoo weather before deciding to quick in or not). Then let's define a common "language" that application developers can use to access car status, sound systems, in car radio, first responder systems, RDS, etc...
Then let's hope that many, many developers will provide interesting apps and add value to the car.
One of the first thing that happened obviously is that many providers offered some sort of android stack on top of Linux so that all the nice applications that are so popular on the Mobiles and Tablets will be "ported" to the car. It is technically cool, but in reality the wrong idea. - first the human interfaces should definitively not be the same, the driver and the passenger positions are different, the way the driver is sharing his or her attention between the car, the IVI and the road is not adapted to "mobile phone type interfaces" or crash texting would not be something to worry about. - second there is a high risk of "ennui", both from the user side and from the provider side, since now 500 000 applications run on your IVI why are you not buying ? (panicky voice) we will not invest more if you do not buy right now ! (stern voice), well I already have all the applications I really want on my phone, and if you do not add something that is really useful in a car I will not bother.
In practice there are at least three issues: - you keep your car at least 3 years (probably 6 now "sabes : es la crisis"), and you hope to sell it for good money (unfortunately "accessories" do not add that much value to your car), so expect that the IVI will be always "old tech" compared to mobile phones, even if NVIDIA is doing some interesting things with their tegra IVI boards. But bus, interfaces, etc.. evolve also, and I'm not sure that the same reasons that made PC hard to "upgrade" will make cars just as hard to upgrade. - the owner buy the IVI, but s/he is driving so it's the passengers who really enjoy the machine. - and as said before the human interface has to be reevaluated.
Then there is the competition from the phone, there are so many people who have already replaced their in car GPS by a mobile phone, our content is already on or accessible through the mobile phone.
So the secret is probably to create an interplay between Cars, Phones and Cloud based services where each element offers the best integration point possible, and the applications are split into two or three parts, each "cpu" doing it's part.
My car knows how fast I'm driving, my phone knows what I'm planning to do in the next couple of hours (because it combines my private agenda with the cloud based corporate one) and the cloud know that I might or might not miss my plane, train, metro.
Friday, April 26. 2013
Once upon a time, I was doing 3D stuff, for real :
Copper sculpture I made when I was about 18 years old, I loved the "19th century" aesthetics long before steampunk became fashionable . And it paid my studies..
I studied Computer Science because it was new, interesting and cool, cutting pieces of copper, hammering them and soldering them together was my "serious" activity.
So now that I see a way to link the two sides of my brain into one activity (and now it is not the left and the right one, you need creativity and analytics in both professions) I find it really interesting.
But what is obvious is that there is a very long road from where we are now to something that works as well as A bunch of hammers. You can try Blender Sculpt Mode, but it involves quite a high cost of entry, and for 3D printer to pass the "look it prints something" stage we need to build a full workflow that is intuitive and accessible to DIY/Cosplayer/Artists/...
But the people who got exited by stuff like this Heathkit H8 computer kit, are the ones who build the PC revolution and the people who started with an acoustic modem created the Internet as we know it today.
So 3D printers are the first building blocks of the technology that is attracting todays dreamer.
Even if it will not be as ubiquitous as the PC or it's bastard child the phablet , since you can spend 2 to 8 hours a day "looking at moving images and texts", but you cannot spend the same time producing new objects. But it has the potential to move us from a society that consume massive amount of identical objects into a society that consume less objects in way more variations.
More relevant objects done for you might be one of the main drivers to reallocate manufacturing capacities close to the consumers, and produce products that people actually want to keep. So it gives hope for shorter transport, smarter logistics, longer product lifespan all in all not greenwashed massive production of "sustainable products", but greener production through a better product life cycle linked to higher quality and higher "desirability".
Friday, April 26. 2013
Recently I've read some comment by Larry Ellison : Zdnet Article about Oracle's T5 & M5 SPARC servers Now I have a special fondness for the SPARC architecture, for many years the sun workstations where my "dream machines", and I also believe that no industry should depend on only one provider. So although ARM and AMD are giving Intel some run for their money the more high end alternative there are, the more incentive for innovation there is.
But the main argument Oracle is offering to support the SPARC architecture is that they put database query and Java accelerators onto the chips.
Ellison summed up by saying that
"Software in silicon is redefining enterprise computing",
And he is certainly right, but the conclusion might not be in his favor. Because the fact that Java is faster is interesting only for people who prefer java over C++ or php or phyton or something more exotic ... The real potential win is related to the database query optimization, and this might generate some paranoia at HP or IBM who will suspect Oracle to optimize specifically for the SPARC and forget about generic optimizations that could run "anywhere", and weaker database sales will not help hardware sales...
But the real competition will come from a completely different direction.
Here enters a completely different player the center for Harvard's middle eastern studies department is not really the place where you would expect IT innovation, but Todd Mostak a master student did a clever hack he created a specific parallel database called MapD and used a GPU to crunch his data. see Using GPU for fast database crunching
So we have an industry that made cheap massively parallel processors in very large quantities (thank you gamers), and enough people having not very much money but a very good idea of what exactly needs to be optimized, and the open source movement that makes it easy to share this kind of inventions.
So the competition might not be HP or IBM, but somebody like supermicro with an added NVIDIA Kepler GPU...
GPU's are used to do graphics obviously, but also a lot of number crunching (see Top 500 computers in the world) and now database crunching (and it's not really "new" people have been offloading some of this kind of tasks for quite some time, but having a first "open source data base" built around the concept is interesting). I'm waiting for the GPU accelerator for (sky|my)SQL postgress and redis ...
What might be non trivial are the kernel tools to manage the sharing of the GPU among various processes, but there is a very interesting presentation at a not so recent Usenix: about TimeGraph that aims to help there; Unfortunately the link to the source code does not work anymore and it seems that some collaboration with a closed source software provider might be an issue.. But the ideas are already there, and that's what counts.
Friday, April 26. 2013
When you buy some land parcel, and do not use it after some time it gets full of wild grass, various trash appears and it starts to look quite derelict...
And this is true also for "virtual" real estate, so having my own domain is a cool way to keep an easy to remember stable email address..., but it starts to look like a responsibility.
Therefore I made the first step, as any self respecting geek, I used advanced procrastination and am working on selecting a blogging CMS.
I could obviously use the http rewrite capacities of my hosting provider to mv www.sinz.fr to