Well, if the GPL wasn't a bullshit license which states that you're subject to the GPL if you even use GPL software in your project, this wouldn't be a problem.

Since it doesn't say that, I'm sure you'll agree that this is a problem.

Like all licences, GPL constrains how you may used the licensed thing. All you have to do is satisfy those terms and conditions and you're fine.

Not sure I'd agree with a cheaper ARM11 ... more like a cheaper Cortex-A8!

Absolutely agreed that A5 targets ARM9 and ARM11 users though. ARM makes that clear. All other things being equal, I'd want a Cortex-A5 instead of any of those. ARM9 is trusty but limited at the high end. ARM11 is kind of awkward; never quite took over from ARM9, and given Cortex I doubt it'll ever catch on all that much more. ThumbEE (on Cortex-A) is way better than Jazelle (on ARM9/ARM11); it works for any JIT-oriented runtime, not just Java. Thumb2 makes me think of pure Thumb2 userspace, for the code (and icache!!) space savings. If I were designing smartphone chips, I'd like that ability to stick in additional A5 cores, without huge changes when customers need more CPU horsepower. Ditto if I were designing any other kind of ARM-based SoC product.

Another point is that Cortex-A5 is I think what Xilinx will use in their new ARM-equipped FPGAs, due next year or so. That's going to take another segment of the market by storm ... they're not giving up their PPC-equipped options, but are working more closely with ARM IP, and to evolve the ARM interconnects. Seems like a win all around; less work to integrate the various IP products, and Xilinx will be able to align some power reduction efforts to better match what most developers want: ARM applications driving low powered semicustom designs. In contrast, the Cortex-M1, long available in soft core packages from e.g. Altera and Actel, is a pretty low end beast ... for microcontrollers, not applications processsors.

Didn't the AMD LANCE Ethernet controllers (including Am7990) do per-packet DMA into host memory, and ring buffers? The Linux LANCE driver has a 1993 copyright, and I'm fairly sure the chip was earlier than that. At the time, ISTR it was one of the few nicely designed Ethernet chips in existence.

So if your quick scan is correct, then either AMD should have been sued back then ... or there's this thing called estoppel which really ought to block this suit. Unless maybe AMD licensed the patent? Though I also seem to recall that estoppel doesn't always apply in patent cases like it does elsewhere.

Get an effing clue. Consent is the line for rape, not violence. If the rapist issues any kind of threat -- social too, not just physical with a "weapon" other than hands -- then it's rape. By your bogus definition, if the victim is in sufficient fear for her (or his) life that they believe fighting back is not an option ... then it was consensual non-rape sex.

Teh stupid ... it burns ...

I sort of agree with the sentiment that releasing someone may be the problem ... but how can you know when/if it will be? There's a really basic problem with the so-called "justice" system in the US, in that it no longer attempts to rehabilitate people. And punishments are so late (I read yesterday about a trial finishing FOUR YEARS after the crime!!) and disproportionate to the offence (on top of crap like "big" drug players getting sweet deals while "small" ones get their lives ruined), that expecting sanity is unreasonable.

Prisons are not just there as punishments; they are there to keep dangerous people separated from the rest of society.

That's a serious misconception. Today, they're fundamentally about punishment ("sit in this overcrowded and dangerous hellhole for a few years"), and secondarily about segregation ... and in effect, also secondarily training about how to become a repeat offender.

Of course, if prisons had effective rehabiliation efforts then the repeat-offender training would become a non-issue. And there'd be a lot less of this "throw everyone (but mostly minorities) in prison, and never let them out" crap. But the prison-industrial complex wouldn't be so profitable then either.

If you've ever had to find data sheets for chips given schematics (or sometimes just a board) you'll see another version of this. Google returns lots of results ... with over 95% of them being chip brokers or third parties that somehow never seem to have those documents in their for-pay database of data sheets.

It gets very hard to find, say, a page with the current vendor of those chips (after three or four buyouts).

Maybe this company can make *those* numbers work better....

We're not sure this is the right design to be encouraging given that it wasn't in HTML 4.01.

I'm not sure that is the right thought process to be applying, given that HTML 5 is supposed to extend HTML 4.01 ... regardless of the specific feature in question. One hopes that's just a really rushed/broken edit artifact, not a real reflection of what they think.

I could believe many of their comments are appropriate, but it's worrisome to see one like that escaping orbit.

It's not nothing. The TTY module has lost another very talented maintainer.

Fixed that for you.

Previous maintainer was ISTR Russell King, who does enough other stuff that he never had time to overhaul the TTY stack. It's needed such an overhaul pretty much since it was first written. Most developers who've looked at TTY have broken out in hives, either before or while running away. Which has, previously, been the only sane reaction from anyone who didn't fancy cleaning up those Augeaen stables.

At some level, it was clearly just Time For Alan To Move On ... since most of the overhaul is done, and he's smart enough not to want to get stuck with TTY forever.

Ironically, a couple of decades ago they were sitting there with literally the keys to the realm in their hands, and they threw them away. Back in the late 80's they introduced the Sun386i workstation, featuring (drumroll..) Intel's 386 processor and a 386 port of SunOS. This was a proper preemptive multitasking OS with 32-bit virtual memory and a decent GUI, far ahead of Windows 2.x at the time. Not only that, it also had a functioning DOS emulator, allowing the machine to run MS-DOS programs. By focusing on x86, and selling SunOS/x86 for $50 or so they could have become the Microsoft of today.

The Sun386i product line was what got Sun onto huge quantities of financial market desktops, and got Sun beyond Engineering/Server markets in a major way.

And they were set to be the first to market with (drumroll) the Sun486i workstation, which worked even better. In fact it out-performed the first SPARC generation... can't have that! They invested in SPARC to get founder Andy Bechtolsheim to come back! (He wanted to design CPU chips, and that wasn't really practical at a company making primo commodity-based systems.)

But, they weren't interested in playing the massive volumes with razor thin margins game of the PC world, thinking that the unix workstation market was insulated from the PC market. After all, PC's were for chumps running 1-2-3 and Wordperfect.

But the Sun386i was a workstation, not a PC. The big apps were CAD tools and financial analysis packages. One reason it was popular at customer sites was however that if you had one, you didn't need TWO honking big pieces of computer hardware at your desk. The same one could handle all that PC stuff (which you needed regularly) as well as the hefty stuff (which you needed constantly).

The real issues with x86 were political ... sometimes masquerading as strategic. It was developed on the East cost, not the west. Keeping Andy; not having to deal with the fact that the engineering culture on the west coast was aggressively blind to a lot of issues. Wanting to see themselves as Sun Gods. Even the desire to avoid investment in DOS/Windows compatibility, despite the customer demand for it.

So they introduced their own hardware, SPARC, and discontinued SunOS/x86.

They gave the Sun386i product line a nice lingering death, though, then more or less excised it from their corporate histories. That all the wood behind one arrow buzz-phrase, widely used inside Sun for a while, was all about getting rid of non-SPARC product lines. And stifling dissent.

Another factor was that the Sun386i products had a different -- and more Apple-influenced -- design approach. Maybe it was realistic to focus on higher margins for a while. But the level of internal censorchip it took to ignore everything the '386i stood for (and Sun itself once stood for) ... was intensely damaging over the long term. A lot of upper level Sun engineers and managers internalized those battles so deeply they just kept blinders on.

I think Schwartz is a closet hippie (witness the pony-tail). He snuck into Sun pretending to be an MBA-bearing preppie...

Actually he came as part of the purchase of "Lighthouse Design" (?) which wrote some GUI tools that Sun liked enough to buy (and then obviously to kill). He was in Engineering ... but didn't exactly do anything except politicking. (IMNSHO.)

In various places, truth is not an absolute defense against libel.

One of those places, it seems, is Massachussetts. Perhaps there's a common heritage in English law.

Well, actually the netbook makers such as Asus are trying to move towards ARM-based machines with Linux so that they can reach much lower price points.

And much longer battery life. Power efficiency with ARM is a lot better than on x86.

When do they start suing the Intel Corporation or Acer (one of whom had coined the term IIRC)...

Ding! That's the sound of your head ringing after being struck by a cluebat.

The whole point is that Psion coined, and trademarked, the term several years ago. Neither Intel nor Acer coined the term, despite what the bogus article summary says.

In fact I even own one of the Psion netbooks. It's actually not particularly old; maybe four years. I have older machines that are still in active use. If the Linux support for this one were in mainline, I'd use it more.

That's why multiple sources are the best. Whenever sources disagree, the more reliable sources are trusted over less reliable sources.

Not really. They don't like primary sources in regards to current Technology, for one example ... where the hierarchy tends to go (1) specifications are the primary sources; (2) comments from people involved in the specification development are secondary sources, and may have some biases but may also provide useful explanation that's not immediately clear; (3) trade rags publish articles written by people who aren't competent to participate directly, and these are tertiary sources. Wikipedia strongly prefers tertiary sources, which in these cases are the least reliable. Write an article based on primary sources, and it gets flagged as needing references. But hey, there may not be any ... and if there are, there's no way they're as reliable as the primary sources.

That's almost the same point as in the article by Jaron Lanier. He's the primary source about himself. The fact that an article about him is more about a myth than about the real Jaron ... indicates a problem.

At some level you could claim this is an illustration of the need for domain-specific ontologies ... a notion which Wikpedia doesn't currently endorse. It's one-size-fits-all, and they use a methodology better suited towards history than technology. Moreover, a method that's not well geared towards good history ... since it puts tertiary sources on a pedestal that is entirely inappropriate for current topics.