Lack of bit field instructions in x86 instruction set because of patents ?

I believe that extra time spent in requirements and design does not actually increase the time required to deliver the product. In other words, each day spent (productively) in the requirements or design phases will reduce the time until completion of the project by at least one day, and often significantly more than that.

G.

Telnet still has its uses on devices which don't have the resources (CPU or memory) to implement SSH. The lack of security isn't an issue if such devices are kept on a restricted network (along with TFTP, SNMP etc).

Meh, many of the purely mechanical door locks of this type don't actually care what order you push the buttons in. So if you set a combination of 1234 then 3421 works just dandy.

G.

That is _extremely_ dangerous, particularly if you claim to sell bug-free products:

I have a good friend (Mike Abrash) who write an article about how he, during the development of the Quake game, chased a tiny visual glitch for a _long_ time:

This thing only turned up while running at full speed, withe optimized fp code.

He finally figured out that their PC vendor had sold them a white box machine with an overclocked cpu in it (maybe 100 instead 90 MHz?), and that this 11% out of spec cpu would pass every single BIOS and vendor test sw they had, except for the specific case of one particular sequence of hand-optimized asm code, doing fp multiplication on numbers with a specific pattern.

I.e. running 11% faster turned out to be the point where a few result bits from an FMUL could turn out wrong! :-(

Terje

The CPUs from FreeScale can be overclocked surprisingly high: the MC9S12 rated at 25MHz goes almost to 70MHz. It seems like the core can go even higher, the limitation is the lock range of the internal PLL. I wonder what is the reason for the clock margin to be that high.

Just the observation: in forums or usenet, people often try to compensate what they don't have in the real life :)

While ago I had PC AT 486DX40 which was overclocked to 50MHz. It worked perfectly well except for some specific places in "Doom" computer game, also from ID Software. When a semi-transparent monster came into the picture, the computer crashed immediately. The problem happened only in that game, always in that situation, and it was clearly related to overclocking.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Oh, yeah, suspended ceilings pretty much make locks irrelevent!

Had a boss (actually, boss's boss...) who of course started locking his door, as he had 'sensitive' material in it. However, he was a great practical joker, and he and one of my co-workers whose cube was outside his door had a running 'prank war' on.

First, we unlocked the door. We then piled an enter wall of printer paper boxes (full!) across his office and the door. Since it was near Holloween, we added fake cobwebs and plastic spiders.

Took him a hour to get the door open again, as there were three rows of printer boxes behind it!

Another time, we received a bunch of little electronic alarm clocks as a 'bonus.' We went into my bosses office, and hid them throughout the piles of stuff he had, all set to different times. The best were the three we put up in his overhead!

Charlie

It's funny, I had a similar problem in a 3d engine of mine, except I never found why on the set of seemingly unchanging inputs the same subsequent floating point calculations done over and over again yielded somewhat different results and that was visible. Since this reproduced on several different machines, I'm inclined to think that either there actually was a bug in my code or there was something wrong in the execution environment (Windows -- I don't think I tried that same code in DOS).

Alex

If you have to say that, something's very very wrong where you are or in the industry. I know first hand that it's hard to get good software engineers capable of doing low level coding/testing (device drivers, kernel, optimizing/asm, etc) and paying attention to that kind and level of detail. If applicants can't use a debugger or write a trivial program like you described, something's seriously wrong.

Alex

Heh, yeah, it was called trying to hire people to maintain legacy business applications written on 30+ year old minicomputer systems.

Fortunately I don't need to do that any more, but I have my doubts that the quality of random job applicants has improved much since.

G.

We had five-button "cipher" locks put on our lab doors in the late '70s. They then alarmed one of the lab doors to make it "more secure" (every day someone would barge through it, bringing security out of the woodwork). Security refused to put a cipher lock on the second door because doing so "would make the lab half as secure". BTW, the two doors were 20' apart on the same hallway.

In another building we had badge readers on the doors, yes, with drop ceilings, though they were pretty high so perhaps they wanted people to break legs getting over. ;-)

The ones the Air Force had at sites I worked at had metal shields over the lock panel (actually 5 rocker switches) to prevent that kind of thing.

So they _are_ synchronous.

No they don't. Unless you use next to no processors at all.

Phil

I don't think he has. I know that I can't do my initial solution sketches at a keyboard; I keep notebooks for that: I need to draw my buffers and vectors, and label their sizes. I need to sketch signal flow graphs. I need to be able to group dot-point ideas with multi-line braces, and connect those with cross-page arrows. I need to be able to break into symbolic mathematics, with all of the greek, subscripting, matrix notation and whatever other symbology goes with it. I would dearly *love* to be able to do that sort of thing at the keyboard with similar facility: then I'd be able to search through it. But I can't, at least that I know of.

LyX is on the right track, I think, but it's probably a bit constrained by existing LaTeX document styles, and doesn't (that I know of) have an interactive/keyboard interface to pict diagrams like it does for equations. It's quite a while since I used it. Add those few pieces together with some free-form page layout and you'd be heading towards an awsome creative tool.

Cheers,

That's *not* what he said.

Mike quotes you in this article:

"almost all programming can be viewed as an exercise in caching"

Indeed it can!

:^)

There is no need for near synchronous interrupts to be an insoluble problem.

if someting is not safe, don't do it.

More often like an exercise in stumbling.

John

Yes, but what he said isn't relevant to the discussion, either.