Where is the Oracle Instant client for ARM ?
Still no download available for generic Linux / ARM systems :
http://www.oracle.com/technetwork/database/database-technologies/instant-client/downloads/index.html
What gives ?
60 publicly visible posts • joined 20 Nov 2007
It would be nice to see Intel with chips in the microcontroller + WiFi space where I think the IOT hometurf is really at. But somehow I find it hard to believe, as Intel has a prodigious past in burning just these kind of bridges. Intel does not have many microcontrollers to offer compared with the compition, and all things ARM (XScale) has been ditched long time a go.
I have gotten dozen of indecipherable messages when visiting central Beijing. Asking the locals to translate them, invariably turns red faces. But it is more bothersome that my phone becomes unavailable for legitimate traffic while it is hijacked by the fake GSM base stations.
The story doesn't say, but it may be reasonable to assume that the bits of Nokia that have been flogged to Microsoft and by extension Microsoft is not party to this truce. Thus making it an agreement between the surviving Finnish bits of Nokia where all the IPRs are to be found and HTC.
The USB standards group is to blame for not having created a standard USB serial interface specification, which in turns means you can't have a serial modem dongle work out of the box.
3GPP group, and various telcos for not specifying a default internet APN (empty string perhaps?) and operators desire for complicating the setup, requiring all sorts of bundled crapware on modem dongles.
Having a web server on the dongle gets around the setup problem, but not checking the Referer in the http request header is plain stupid in this case
I don't think any company big enough to have a proper legal department will touch this thing with a barge pole. The problem is with the licensing terms, which gives a royalty free license to some undisclosed IP. The caveat is that if you enter into any kind of legal dispute with Qualcomm (even on completely unrelated areas), you loose the rights to utilize said undisclosed IP. No serious player in the mobile business would agree to such a legal hobble.
My previous droid was an HTC One X which was rather good when I bought it, but had some problems:
1) Did not wan't to accept US T-Mobile sim card (prepaid when visiting US)
2) Firmware update broke Bluetooth A2DP for my Sony car stereo
3) Later Firmware update broke WiFi channel 13 support (regional setting)
4) Finally the GPS antenna came loose and navigation stopped working (well known issue)
At this point I was not so much inclined to buy another HTC product, and no amount of marketing I am afraid would make me change my mind. In addition to this the bilingual text input was a complete mess as "å" (Norwegian and) would always be corrected to "a" (English), with no thought given to how far apart they are on the keyboard, and that the likelihood of mistyping is virtually nil.
Perhaps with the exception of floating point performance, it seems unlikely that highly tuned programs will run very much faster on this. If on the other hand you are running a crapload of bloatware, all of the listed performance tweaks should translate into higher overall performance.
The venerable Swiss crypto maker Crypt AG was little match for the subversive skills of the NSA when they wanted a back door into Iranian diplomatic communications. I have a suspicion SIM card makers are even more of a pushover when the men in black come knocking with polite requests for copies of embedded crypto material embedded in SIM cards as they leave the factory floor.
This way, the NSA won't even have to break A5/3 when doing intercepts in foreign territories
This weeks BlackHat conference is firing a volley of shots at GSM. (Break A5/1 - intercept calls, set up IMSI catchers) Some of these faults are due to GSM phones not properly authenticating the network they connect to. Unfortunately as long as 3G phones rely on GSM, they will be equally vulnerable. Moving 3G networks to 2G frequencies is a step in the right direction for patching 'the Achilles heel of 3G security.'
From BeeLine GPS User’s Guide:
17 GPS Module Limitations
The GPS Receiver firmware contains an algorithm that allows either the speed limit (515 m/s) or altitude limit (18,000 m) to be exceeded, but not both. This allows the receiver to be used, for example, in high altitude (research balloon) applications.
So as long as the speed stays below 515m/s you should get good high altitude readings.
Some further study in the gospel truth of google searches, reveals
http://www.thefintels.com/aer/gpstelemetry.htm
Meaning only one of the limits need apply, and by judicious choice of GPS, and keeping speed well below mach 2 it should be possible to track the craft, even above 60000'
I see a potential problem that could ruin the mission. Civilian GPS units have built in hardware limitations, and won't track if the speed exceeds 999 mph, or altitude goes above 60000'. This to prevent the ilks of Kim Jong-il etc. from building a cheap missile guidance system.
You can stack 8 GPUs to a motherboard (In the form of a standard 5890 card) - And you can stack 2 motherboards side by side in a 3U unit. The proposed rack would hold 22 3U units.
Then to make it easy to program, you write a OpenCL wrapper API for the whole cluster. - and divvy out the work through OpenCL to the individual computers - together with a slice of dataset to chew on.
It isn't quite a homerun though:
0.816 petaflops @ 66KW (1500W per machine)
It would need custom motherboards to fit in the rack, but one could start work today using off the shelf components, and by the time ATI moves one step down from 40nm, the power and size issues will take care of themselves.
Where do I collect my prize?
@AC 13:01
It will be interesting to see how they compare to the 58xx. I have two of those cards in my rig, and use them for rendering stuff like this: http://www.youtube.com/watch?v=6w6vo9t62c0 or compute rainbow tables etc. (or just plain video and image processing)
Has anyone notice that the DX11 instruction set contains the "canonical NSA instruction" (bit population count) - I expect various 3 letter government agencies to stock up on these cards as well now.
I bought the 5850 card for for playing around with in Linux. Very impressive stuff. Unlike nvidia the chip has a full integer instruction set, very useful if you want to compute rainbow tables. The ATI cards are ahead of nvidia in when it comes to speed on the GSM cracking effort.
For floating point work it is brilliant. The author mention mandelbrot, he should get his hands on my mandelbulb renderer. For sure he will waste more than just 30 minutes looking at that twisted old knob.
I am no M$ fanboy, but judging from the speed and compactness (small amount of data transfered) it seems that Redmond got something right in the exchange protocol.
The alternative SyncML - ( made by OMA, whose standard specifications are more akin to statements of intent ) - is just incredibly slow and chatty. That is if you can manage to implement it - I guess big G may have just given up on trying to implement SyncML - even with the full "spec" in front of them, and found it easier to reverse engineer a proprietary Microsoft protocol.
Gmail is an excellent way for me and my miss to keep contacts and calendar in sync on our S60 phones.
Most www animtians are flash these days, hardly an improvement. But the GIF patents expired in 2007, so there is little impetus now to pay the royalties that unisys demanded. 2012 / 2017 is the expiration years for mp3 patents, till then you will be giving alms to Fraunhofer for every single mp3 capable device you buy. Your cost for convenience on a personal level may be negligible, but for HW producers that churn out units by the millions with paper thin margins, pennies and cent to Fraunhofer accounts takes a good sized chunk out of their margins.
What Fraunhofer doesn't seem to get, is that the road from a Chinese manufacturing plant to the EU market is long and arduous. It would make more sense to ensure that products shipped by the container load to EU are fully licensed. Enforcing licensing requirements on the trade show floor, where much of what is being showed is just prototypes anyways, adds more hurdles in bringing product to market, and could very well result in fewer overall unit royalties. This coupled with the ill will it generates, could bring about more uptake of royalty free alternatives, such as ogg.
I know of at least 1 Chinese company that got a visit from Spain's customs / police during MWC 2009 - even though they had their papers and licensing in order. Either Fraunhofer or Spanish officials are getting sloppy, more of this it and it could very well trigger the necessary backlash to see a real uptake of the ogg format. ( remake of the burn all gifs campaign , 2012 / 2017 is still a rather long wait. )
After 1 week of using the G1 I have almost been sold. I did expect it to be an half baked attempt at an iPhone killer. But it seems that what has happened is that Google set aim for the iPhone, missed woefully on that target, but hit windows mobile right between the eyes instead. As a phone, it completely trounces windows mobile (except if you have an absolute need to talk with MS exchange)
Fear is what I hear in Ballmers voice.
Speaking as a developer, the Linux kernel is just brill for mobile phones. The problem lies with a number of manufacturers that confuse "Linux phone" with "smart phone". There are a number of Linux phones out in the market place, but very few third party applications for such phones. That stems from the almost complete lack of standardized APIs for the cellular space.
However this doesn't mean that it will always be this way, efforts are well underway for instance in and around the LiMo foundation to supply these standardized APIs that can transform linux into a viable smart phone platform. But of course LiMo is a in a race with several contenders ( iPhone, Android etc) to determine which ecosystem of application , developers and devices will be the dominating one in the years ahead. Thankfully, it won't be Symbian.
Carriers are carried away (bad pun) by a Linux pipe dream. Android, and iPhone have been a clarion call - They can no longer control the users experience in their little walled gardens, and thus they risk being commoditized and made into dumb pipes.
But the fact of the matter is that there is no credible Linux platform for application development today. Android mind you is just a glorified java phone.
GTK has prooved to be to heavy to run on small ARM devices, thus it has been bitched by the OpenMoko effort. The LiMo foundation is still sticking to it, but will get badly burned when they fail to provide a secure sandboxed 3rd party application environment.
So as a Smartfone platform Linux is quite lethargic, but that is not to say that it is dead in the water. It is still a very good OS to run a phone on, especially compared to e.x. what Qualcomm has to offer. The Qualcomm platform is so decrepit that there is now talk about running a user space Linux on top of it. Not that this will turn it into a "Smartphone" platform that the carriers can control, but it should make development easier.
Having seen 2 dead & crying iPhones on the cross already, I realize that they are not the sturdiest of devices. There is no way average device lifetime is even close to 2 years, much less 3. What will the punters replace the broken iPhone with when they are tied into a 3 year data plan? Androids ?
I agree this would be nice, but eee supports the huawei E220 modem flawlessly. If you have some clue about what 3G setup you need, the wizard will have you up and running in less than a minute.
My 701 + E220 works so well for El Reggin on my commute, that finding reasons to upgrade to 901 is hard. Integrated 3G and a bigger screen would perhaps be reason enough though.
Pure 4G IP networks (like WiMAX) can run VoIP on top of an all you can eat IP plan. ( Mayby some QoS qoutas will be thrown in ). Charging termination fees on such a setup is impractical, and (I hope) will force other operators to abandon them altogether. Competing with something that has zero marginal cost is rather awkward.
Data roaming charges are broken beyond repair, and if I have to make any extended stays in EU countries, I try to buy a prepaid SIM and use HSDPA / GPRS and make my calls using SIP or Skype.
For Instance, in Italy you can get a TIM SIM for 10 EUR - charge it to 30 EUR, and then activate "Maxi Alice 100" this gives you 100 hours of full Internet access anywhere in Italy ( apn: ibox.tim.it )
In Sweden you can get a Telia SIM, charge it, and Internet traffic is capped at 69 SEK a day. However you must first register the SIM to a Swedish address. Ask the dealer to help you out, customer service is expensive and confusing
The trouble is that researching all the loops you have to jump through is time consuming, and should not have to be repeated over and over again by anyone wishing to roam cheaply. This information should be in a wiki somewhere.
Paris, well since she seems rather cheap
From what rumors I picked up at MWC it seems that the DVB-H pushers will move traditional access control mechanism from satellite TV smart cards and stick them in the cellphone SIM. There are a couple of problems with this approach though.
1) Punters will not want pay a full subscription for a service they use 5 minutes every 3 months.
2) Watching TV while roaming is out.
For this reasons DVB-H will only work in Germany, where consumers have enjoyed free satellite TV since the inception. Thus asking punters to pay a subscription fee is out of the question, and DVB-H will have a modest chance of seeing any uptake.
Many smaller handset manufacturer balks at the entire 3G licensing scheme, and are looking for ways out. A SIP capable WiMax / WiFi handset with a decent battery life could bring the 3G house of patents tumbling down as well.
This would also leave IMS ( http://en.wikipedia.org/wiki/IP_Multimedia_Subsystem ) a smoldering ruin, which can only be a good thing for consumers. Future networks should be an open IP architecture, and not a fragmented jumble of carrier approved services.