Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?

Comment Re:Dealing with IP theft in China (Score 2) 68 uncle (anthony pickford) who used patent law to protect against dangerous copy-cat medicines [which were killing people - literally].

I would not use patent law to protect prospective customers against a lethal machine. Even in China, building a product that kills its user is an inherently limiting prospect.

yeah, in china, because our clients are State-Sponsored factories, it's their business that would be disrupted and threatened. so we just give them the name of the product, and let them go beat them up. or whatever it is that they do in china to people who kill end-users.

i mentioned this in the original post: the use of patent law (by my uncle) to track down the criminals was extremely innovative. remember that they had absolutely no idea who was supplying the killing-drugs, and the only people who would know was Customs and Excise. but they weren't telling. so they sued a random company for patent infringement that they *suspected* of importing the killing-drugs, *knowing* that they would lose, but during the pre-trial they made a "Discovery Request" to the Inland Revenue.

  they lost the case, but they then sued the Inland Revenue when the Inland Revenue refused to comply with the request. they lost that case, appealed, took it to the High Court, lost that case, appealed to the House of Lords, where my uncle was able to make a presentation and got the law changed. it's now part of UK Law that a 3rd party *must* comply with a "Discovery Request". prior to that case, there was no precedent: any 3rd party could refuse to comply.

  he was an incredible man, my uncle. very intelligent, very very angry and intolerant, but he was amazingly good at his job. of 36 changes to UK Law over the past 50 years, *EIGHTEEN* of those changes were down to him. including the modifications to the Dangerous Dogs Act.

Comment Re:What do you have against..... (Score 1) 68

it's a hang-over from when i had RSI

I knew a guy ten or fifteen years ago that had had the same problem, were you a Quake player? If so, was your in-game and internet name "crash"? I've lost track of all those guys.

ha - no :) i was however an avid "descent" and "descent 2" player - i got some 386s and 486s, set up a network in my house and invited friends round. yes that's descent and descent 2 the *proprietary* version, not the free software version.

then i got into "dark reign", and even invented the "zombies-in-the-underground-transport" thing that ended up on the "hints and tips" because it was so successful at creating mayhem: i would go after the artillery because they had a significant latency on load-and-fire which couldn't be stopped, so they had a habit of blowing each other up and everything around them whenever one of the suicide zombies popped up in their midst.

but no - it wasn't the gaming, it was the 16 hour days doing reverse-engineering and packet analysis in cold temperatures and only being 65kg (i'm 6ft 1) that really did it.

Comment Re:Patents and open hardware licenses (Score 2) 68

Are you willing to release some sort of board using the EOMA standard under an open hardware licence (e.g. CERN's or TAPR's) that gives legal protection to people making (open) derivative products?

love to!

While I can understand why you would like to keep control of the standard with patents, that approach may make the EOMA standard unusable for open hardware projects.

i'm glad you said "may", because i would have to contradict you, and i don't like doing that. we thought this through: the patents will come with an automatic royalty-free grant for any product that's FSF-Endorseable, written in stone into the papers associated publicly with the patent. yes, i'm keenly aware that there aren't that many FSF-Endorsed products: it's something i want to encourage.

If not, using the EOMA standard in my projects isn't worth living in fear of a lawsuit, and I can't in good conscience pass them on to other people with this hidden risk.

it's not hidden: i've mentioned it a number of times. actually, every time this issue comes up i've answered it, that there's been a plan in place even as the patents were being conceived. i really *am* a free software advocate, not a free software band-wagon-jumper.

If you'd like to make money off of people making money using your standard for closed hardware I have no issue with that. If you are worried about compatibility and hardware damage, however, some sort of (copyrighted) certification logo may be much friendlier to open source community.

well... we'll have to look at that as-and-when it happens. if a product's scope is small, i honestly don't think we'll be particularly concerned. and, also, remember: the CPU Cards are where we'll most likely find that people are having to buy these off-the-shelf from mass-volume manufacturers. it's mainly the mass-volume manufacturers that we're concerned about. a few engineers developing alpha-grade hardware and selling 100 to 1,000 units, knowing full well that they and their clients are taking a risk plugging a card into a board? yeahh, i'm not so worried.

but, a competitor company that's making a million units a week and they're incompatible and short-circuit existing customer's products??? that's bad.

An example of this kind of scheme is the USB standard; if you don't see the USB logo on a piece of hardware you know that you are taking a risk by plugging it into your computer. This has prevented catastrophes on the consumer side, but still allowed the open hardware community to experiment and share freely with each other and the world. While you could argue that allowing people to distribute non-compliant hardware risks fires and other safety problems, this hasn't really been a significant issue with USB.

well, that's because USB devices and USB chips come with built-in protection. here it's a different matter. if someone badly or even *deliberately* designs a card which is non-interoperable, and its use shorts out the power lines to ground, there's no over-voltage protection and chances are it'll damage something.

so i know what you mean, and chances are that we'll have to review things on a case-by-case basis and be encouraging to open hardware designers (i want them to succeed!), but we cannot risk letting it get into a free-for-all - it's simply too dangerous, and it would be irresponsible of us. remember, there's something called an "estoppel defense". if we "turn a blind eye" to an "open hardware" platform, and it gets copied and turns into a million-a-month product, and it turns out to have a hardware flaw, what then? the million-a-month company could claim "well you didn't contact the small company when they were doing 100 a month, so we don't have to pay either".

it's a tough one - but the idea is to use the patents to protect the *whole* EOMA community, *not* as a means to stifle innovation as they're presently used. please remember: i'm a free software advocate. i'm working with and from within the system; i'm not looking for ways to *exploit* the free software community, i'm looking for ways to make them relevant!

Comment Re:Define friendly (Score 1) 68

If you're going to be distributing your source anyway why does it matter?

Are they really going to do extra work to change the source significantly for free? It'll just be the same as whatever you gave them to load onto the hardware.

well, yes, they [pick any hardware engineer as an example] might well want - or need - to do extra work. many of these low-cost tablets for example have USB buses. that means printers, keyboards, specially-designed hardware devices, future products not currently supported by the linux kernel, DVB-T USB dongles, 3G dongles, GNURadio products, SDR Radio dongles with that low-cost RA-Link chipset and so on. ... if you haven't got the linux kernel headers and the original kernel config, you're f*****d - you can't guarantee that any .ko modules that you compile will actually work.... that's assuming that you can gain access to the system-level in the first place.

if the driver is _really_ out-of-date wrt the version of the kernel that's on there, such that you have to do a complete and full rebuild of the kernel, you're f****d unless you have the full build toolchain and know the boot process... and in some cases now have the key which allows you to sign the kernel so that it *can* be booted.

then also what happens if the bootloader (e.g. u-boot) doesn't do what you want? for example, it doesn't boot off of the USB-to-SATA device that your customer has asked you to connect up.

so it's actually really really important to have full source code and toolchain. without them, anyone wishing to make use of all this low-cost hardware in order to set up a business or remain competitive simply... can't even consider it.

Comment Re:What do you have against..... (Score 1) 68

After all, if you're old enough, you can claim that you were taught to write this way in half-uncial and that you can't teach an ancient dog new tricks. ;-)

cool! i'll have to remember that. mmm.... wait... first look up "half-uncial"...

ok, i get it :) naah i don't want to claim to be an "old dog" - i haven't the luxury: things are moving too fast to do that.

Comment Re:Slashvertisements (Score 2) 68

I am extremely interested in these (type of) products, the development of them, and any news related to them.

yeah. i mean - i'm a geek / software person. what the heck am i doing getting involved in hardware? ahh... because nobody else is - not with this kind of goal in mind, anyway, and certainly not with the same business aims.

but.... regarding the processor: we have to be quick! it's not going to be long before 40nm is outdated, and 28nm is "king". 28nm NREs and the whole verification process is... costly.

Comment Re:Styli? (Score 2) 68

How much effort is there to get things to work with touch screens enhanced with ntrig or wacom technology?
I like touch screen technology, but at some point you do really need stylus close to a real pen.

oo. intriguing question. hmm, not sure i'm qualified to answer! my personal experience with wacom touchscreens has been limited to the Acer Travelmate C100, which i liked so much that when it fell to bits (i literally wore a hole through the shift key with my left fingernail for example) i bought a C112 as a replacement.

i found that it was quite easy - relatively speaking as far as being comfortable editing xorg.conf files - to set up the wacom drivers and configs etc. at the time i think they were serial devices, so there was a bit of a lag, but you got used to it.

then also i've seen someone wandering around in london victoria station with a tablet that was capacitive *and* had a stylus. i believe the stylus was set up to have the same impedance as a human, or something, but i'm sure i spotted some buttons on the side (like on the wacom electro-static styluses).

so, technically, it looks like it can be done.

Comment Re:Define friendly (Score 1) 68

Doesn't apply if they are hardware manufacturers building stuff for you. Keeping the GPL part is your responsibility.

it's Copyright Law: it's the license. so it's the responsibility of anyone who distributes binaries - be it the SoC vendor themselves, or the ODMs, or the factories, or the distributors, or the retailers, or you. you shift it, you agree to the license - period.

Unless they build extra units and sell/distribute them without complying to the GPL. Is that what he means by not being friendly?

yes. it's down to Copyright Law.

Comment Re:F/OSS-friendly ARM SoC manufacturers (Score 1) 68

why use the A10? because it's around $7.50 in very large volumes, that's why. all its competitors are around the $11 to $12 mark. the other reason is: allwinner are actually trying, within the best of their ability and understanding, to actually work with the free software community. some things they Grok, others they don't. it's challenging, but it's exciting.

So, basically, "screw the end-users as long as we can shave $5 off the price."

*lol* - well... not exactly. if it was AMLogic for example, who screwed us over with GPL violations, then i wouldn't give a flying fuck if their processor was $5 i *still* wouldn't give them the time of day.

and the main reason for that is because if we propagate GPL-violating source code, it exposes our customers to risk of lawsuits - primary *and* secondary Copyright violations. so, we just flatly refuse to work with GPL violators, now, and that's the end of it.

Comment Re:New SOC ? (Score 3, Informative) 68

What can you tell us about the new SOC?

well, people have worked out that it's using ICubeCorp's design.

I've heard that instead of a separate GPU that it contains more 3d type instructions as part of the native architecture.

yes. if you're familiar with MIPS64 and/or ARC (from synopsys) and/or Tensilica's DSPs, it's more like an extended instruction set that gets "farmed off" in some cases to entirely different engines, with their own pipelines. in this case, it's a suite of engines with their own hardware threads.

How long before we actually see these being produced?

from funding to actual silicon? as we're basically doing nothing more than "take some hard macros, add them, and put it through the verification tools" i.e. there is NO development (because the core has already been done, and proven), we've set a target of 8 months.

actually what we're probably going to do is simply add USB-3 host to the existing 65nm design, bump it up to 8 cores (from 2), and then have a *separate* team design a "USB3 Peripheral IC" with SATA-III, Gigabit Ethernet and so on. even if the two wafers end up on the same chip. what's great about that is that that USB3 peripheral IC is itself a great stand-alone product, so we're talking to a couple of teams (including some open hardware designers) about how to get this done.

And which compilers are they targeting for this new instruction set?

they have a compiler expert from SGI on-board, so they've ported Open64. and it has LLVM support as well. amazingly. open64 is an entirely different compiler, but a long time ago when it was started, they took the gcc front-end GPLv2 source code and used that - just wrote a different back-end. so, for the most part, the options are pretty much identical so you can do "EXPORT CC=..." and go from there.

(something to do with the architecture, Open64 is a better choice for them than beginning from gcc's back-end).

Slashdot Top Deals

The human mind ordinarily operates at only ten percent of its capacity -- the rest is overhead for the operating system.