The reason why parent may (?) be misled is that networking stacks are not merely simple pieces of logic. Certainly there are some general rules... these are the very easy parts to program. In RFC terminology, these are the "shall" parts. Every packet that doesn't meet those rules, "should" (theoretically) be tossed.
Where networking stacks get really ugly is not the mandatory, properly implemented, "shall" parts, it's the "may" parts, and the "shall" parts that have been misused, abused, or ignored, and the burden of handling pieces which *cannot* be broken into easily testable discrete units.
For an analogy about why networking stacks are so complex for something that seems deceptively simple, it's fairly easy for most people on/. to recognize a normal, well formed, email address of foo@example.com. That's quite easy to write a short regular expression for. (Go try it, I'll wait). All done? If your regular expression is under 200 characters or so, it's broken on a vast number of perfectly valid email addresses, or allowing through a vast number of illegal addresses. Here's a *still* partially broken regular expression (it can't handle inline comments) for, oh, "simply validating email addresses" based on the defined logic found in only *one* RFC. Not so simple with email, and it's not so simple with networking.
Oh, and by the way, networking code has to perform at absolute breakneck speeds, so you need to optimize the heck out of implementing those "simple" rules, too... and while it's certainly *possible* to validate every single bit, in every single packet, against a protocol's defined ruleset, I can pretty much guarantee you that you will rapidly find yourself consuming 100% of an average desktop's CPU before a 10Mbs stream can be completely checked, let alone used.
Couple of problems:
1. You wanted a custom build, of your preferred tools, rather than using the existing tools. This is a problem with *any* "Enterprise Level" scenario.
2. Enterprise apps tend to use directory services (not dumb databases kludged to act like directories), wherever desirable. MySQL is great for lean databases, but it's not really suitable for a 6 million user mail architecture with possibly hundreds of mail servers. That's why you were being steered away from a slower database technology into a faster directory technology.
3. It's quite possible that you are one of the very few users who wanted to run a mail server architecture based on a slower *database* technology metaphor, rather than a faster *directory* technology metaphor.
To answer the inevitable objections:
1. Yes. LDAP is basically a front-end to a back-end "database". Unlike conventional database queries, however, LDAP is optimized to run like a bat out of hell on the equivalent of "SELECT" statements, often at the cost of "UPDATE" and "DELETE" statements, "JOIN", etc.
2. MySQL does not, and simply cannot, scale, in the same way that LDAP servers do. In an optimized enterprise, each LDAP server handles only the branch nodes it needs to think about. It's somewhat analogous to BIND, in that you can have 80 different LDAP servers in 80 different locations, each only storing the subset of records needed *for that location*. Ever try syncing 80 MySQL servers, while only having each node store its requred records, with pointers to branches further up? Good luck. This is why enterprise-level email usually requires enterprise-level directories.
3. Yes, you'd be duplicating records if you were storing user information in both directories, and MySQL databases. One of them should probably be removed, or split at a reasonable boundary. Hardcore LDAP users know that the boundary is on "what changes a lot", so things like a user's First name, which rarely ever changes, probably shouldn't be in a database designed to update that user name hundreds of times per second. Likewise with that user's current mail server, etc...Things that can change hundreds of time per second or so, well, keep that in a database designed for much slower use/access/queries, but with much more frequent update ability. Things that rarely change, or only change locally, keep them in a fast, small, and efficient directory.
Buyers stay on eBay because there's a big selection of stuff and they can find what they want
I'm currently working for a company that opened their auction site only 4 months ago, and yet we already have 40,000 active listings at any given point in time. Not a ton of stuff, (lots of drop-ship and storefronts), but at least it's better than the many, many, roadkill sites out there that look like they just "gave up".
Sellers won't defect unless another site offers them some other savings...
Most of our prices are lower than ebay, if only by a penny, (and the first image is free... stupid ebay has all of those "photo" icons). For "power-users", we have a price plan that allows up to 1000 listings per year for $419, or 42 cents a piece, plus a flat 5.25% final value fee on any auctions (none of this "how much do I have to pay" silliness).
That is, they would have to pay the sellers a commission on each auction to attract them...
Interesting idea, but notice that the "free" venues such as craigslist can only support such a model by keeping bandwith *way* low. The economics of defection is certainly a tough nut to crack, though. We're giving away vacations to purchasers. *shrug*
...to create parametric search system
We're about 80% done on that one. We made our entry page look quite similar to... uhm... some other searching tool.
..dramatically different infrastructure could greatly reduce server and bandwidth cost..
Most definitely. Now that wikipedia has crunched through the hard work of "how can LAMP scale into a top 20 site?", it's fairly easy to see how it can be done with pages that are dynamic *to the second*, without needing anywhere near ebay's level of infrastructure.
Perhaps others have better ideas...
We're working on it....:-)
For example:
-Free items list for free in our "free stuff" area. No limit on listings, items, etc. It's simple, and cheap.
-We're really trying to pay attention to UI. I don't think very many people would even feel like defending eBay's UI, but we have meetings about color schemes, table vs. CSS tradeoffs, etc.
-Did I mention the first photo is free?;D
-No reserve fees.
-No minimum/starting bid fees.
-No "Buy It Now" fees.
-Sellers can choose to offer any payment method they want.
-Everyone in the company is customer service first, coder/designer/whatever second. Contrast to "we don't care, we already have your money" sites, of which there are many.
-RSS feeds for "stores" are pre-built for store users.
(etc. etc. etc.)...
Basically, we're trying to make something like...iBay(?), or what Apple would do if they did an auction/classified site. Offer simpler features, with better UI, resulting in a better end-user experience.
Correction: There is erosion with similar *characteristics* to water erosion on the sphinx. This doesn't mean the sphinx predates the surrounding desert climate. 1). It could be water erosion, brought about by channeling/bringing water to the sphinx in a ritualistic fashion, even during desert times. 2). For those of you who have never lived in a desert, well, many deserts flood, and have severe damage, but only every 50/100/500 years. They are devastating enough to rip whole stone/concrete/etc. structures apart. 3). It could be erosion *similar to*, but not *indicitive of* water erosion, or even a combination of the two.
(Oh, wait, look, there are canals on mars! There *must* be people there!:-) )
Real science is not about drawing simple conclusions based on what we know, but about finding many possible conclusions based on what we currently know, and being ready to dismiss as "bad" any conclusions that are based on absurd prepositions, if we can find simpler prepositions.
Since most of the "old sphinx" tripe comes from such reputable folks as alien researchers, astrologists, and similarly accedited (which is to say: not at all) types, I find it generally more amusing (and of intellectual value) to point a finger at the general clown show and laugh. When a real scientist steps in (and they have), they're worth a listen, but they don't make similar absurd and bizarre "statements of fact"... because doing such a thing, quite frankly, isn't science. It's mere showmanship.
Yes, it seems like the sphinx is only 3,500 years old. Yes, there is odd weathering, that looks like water weathering. That doesn't mean that the sphinx is much older, or teleported to us from another planet, it means that we have a currently dated as 3,500 year old object, which requires new research, to figure out why it shows weathering signs of being dated older. To state it is much older *because* of unusual weathering... well, you might as well start telling your next-door neighbor about the anal probe the greys gave you.
SAN FRANCISCO
Title: Bay Area Electronic Crimes Task Force
Address: 345 Spear Street
San Francisco, CA 94105
Phone: 415/744-9026
Fax: 415/744-9051
Contact: SA Erik Enos
Office #: 415/744-9026
Alternate: SA Susan Broad
Office #: 415/744-9026
Email: eenos@usss.treas.gov
From:
http://www.ectaskforce.org/regional_locations.htm
I suggest you investigate the earlier versions of the Oompa Loompas, before they were made a more politically correct green.
Seriously.
http://www.roalddahlfans.com/books/charoompa.php
Only allowing the top 20 in this case would be like only allowing the top 20 default *nix distros on stock hardware in a "best OS" prize.
Just imagine what life would be like for geeks if we could only choose the best "pre-fab" systems. No tuning. No alternate systems for some parts. No replacement of software or hardware. Just the top 20 "out of the box" systems.
If I was in charge, I'd run *all* entrants, award top prize to the best overall, but also keep the best traction group, best collision avoidance group, best navigation group (etc.) in my rolodex, and divide the problems out.
Here's how it works. They can't exclude you, or charge you more if you only have a *proclivity* to genetic-related disease.
However, they are allowed to discriminate against pre-existing conditions... even if you didn't know about them.
So, as the genome is mapped, and causitive genes are found, they can drop you for a disease you may not get for 50 years. (If you live that long).
Privacy... yeah, right. Privacy and protection for those with a "non-deterministic" genome, maybe. Those who have causitive genes for later-life diseases are SOL.
Dear Ada:
So, I have this really ugly couch stuffed with 7 tons of down feathers, so it's now comfortable for everyone. To put it anywhere, I need a crane, and a team of 3 movers.
I don't mind though, because everyone *I* know who uses it for their guestbooks and web-to-mail forms and other wep-app-101 tutorial tasks is happy.
Funny thing, though, people who care about speed and performance and portability keep complaining about it.
What does is matter if it's big, heavy, ugly, and unportable... it's comfy!
What can I tell these poor souls who don't enjoy my couch, and want high-speed lightweight benches, or standing-room-only environments?
Is there a new wave of technology sweeping the industry, where cool new killer-apps are popping up like mad, because of new language idioms? I think not.
Can Java solve any single problem that wasn't solvable before? Of course not. It doesn't put new magic into the box, it's just another kind of magic.
This survey seems to be a measure of the buzzword bingo managers who are *still* not realizing the simple fact that: fads don't fix problems. I thought most of them dot-bombed in the 90's, but I guess we have some stragglers.
It doesn't matter what the fad language is, I can still do twice the work, in half the time, by cobbling together bash, perl, php, C, and whatever else is optimal to solve a problem. Any language which is slower to write in, or requires more design time, or stupid ROSE diagrams, ER committees, or 24 hour code reviews for 100 lines, is a *massive waste* of time.
Customers don't pay for design, theory, or somebody's philosophy, they pay for solutions. Now that hardware is cheaper than coders, design matters less than ever, as long as it can scale across boxes.
(Amusing note: I had a contract job where I distilled 300 lines of perl to one line of bash... the spec was re-written to not be language specific, and boom, I had replaced the perl bloat, where they re-implemented bash functions in perl).
Caveat: I get paid for solutions, not buzzword compliance. I don't do gigs where I waste hours on "modeling charts for the extremely daft".
What is so fragile about our 'net that one single organization can stop it? (conversational)
Why place eggs in one (censorware) basket (or even ten, or twenty), instead of hundreds (watermarked data containers, laundry systems, radio packets, pigeon-net, new encryption, new data hiding)?
Heck, if we can't easily stop it with obvious foes (China), what about more insipid ones? (DMCA)
I'm going to stop whining now, and keep coding.
I'm working on proto-coding various shared-secret (yeah, I know) based byte-use methods, so gathering an assembled message may be in 50 website gifs, an mp3 file, an ISO image, etc. etc.
Example: Byte 1297 of this porn image, Byte 16 of that slashdot header, Byte 1027,7189,1081 of a specific NT distro....so it's functionally impossible to censor all possible sources until a "reader key" is compromised (so, single-use keys are best). It's a WW2(1? prior?) transmission technology (words 197, 213,618,541 in a local newspaper), only the number of available sources to hide data has increased exponentially....
Even in Red Lobster's website... (ponder that for a second...)
Of course, this is probably *already* being done, so any links to existing sources would be helpful.:-) The hardest nut to crack (so far) seems to be gathering an accurate profile of the non-random destination user (to not trigger "abnormal behavior"... "hey, send me your browsing history, ji-chang, I like your p0rn"!), quickly find words/bytes/bits (or even chawmps, playtes, varying sizes to increase obsfuscation) in their "typical" stream to use, and, of course (as with all shared secrets) to transfer the key itself (pigeon-net! spaces/tabs in an HTML page (each space=0, tab=1)! and so on).
"i bet very few OSX users even know they have it(1), let alone use it (2)"
1. No. The *nix geeks regularly evangelize on every support line, in every tip site, and every magazine. They know it's there, even Apple pounds it into them.
2. Yes. The fear is there.
Apropos of this, I suggest one of the most important things these students can learn is that it's okay to do things that break, that wedge, that "go boom" in funny ways, because they can actually fix it themselves with OSS, or at least, "add eyeballs" to the problem!
Don't forget that some of Mac OS X's PPP issues were resolved by a teenager...
-Bop
Re:Um, you've never lived in New York, have you?
on
Step 2, Groceries
·
· Score: 1
Nice to see sanctimony hasn't gone out of style.
You might try shopping for a family rather than just for yourself...
Me, I'd consider a service like this even living by myself -- I value my time enough that the nearly 2 hours a week I'd spend getting food your way is a bigger price to pay than a $10 delivery charge for a week's worth of groceries.
Hm. Your family foods are being defined as a quantifiable "service". With children and elderly defined by cost.
My family is not a cost (or revenue) center. Good food is much more important than cheap food, even if it's web delivered.
If you can't pack dinner for a family of five, get in shape! It's only 8 pounds or so!
Is lugging 8 pounds 4 blocks more important than your family?
Good food matters more to me than the convenience.
Drunk driving is not deadly, unless you've had so much to drink you die at the wheel while waiting for a stoplight.
Running into people, crashing cars, etc, is much more deadly, but it's not the same thing, is it?
Until the US can intelligently think about such silly 'slippery slope' arguments, every thing that could have fatal consequences, based on correlative, not causative, statistics, is up for grabs.
Let's apply that same enlightened thinking of 'drunk driving is deadly" to other items, shall we?
Linux/BSD users are (or will be) crackers, and should be in jail.
People who express anti-government opinions now are future terrorists, and should be in jail.
If you have a single illegal mp3, you are probably a future multi-billion-dollar-pirate, and should be in jail.
By all means, punish those who commit crimes against others, but to arrest someone *only* for drunk driving is to arrest them for future-crime, for the "potential" and "likelihood" of harming others.
I live in AZ, where 'drunk driving' is one (1) beer before driving. And I use good computers, which tend to kill less people than cars, and I own a handgun.
Three strikes against me on future-crime, I guess.
Certainly, not all countries need (or can afford) the decentralized information economy and infrastructure that the first world shares, but I think there's some possibilities for making a real impact.
They don't have the money to waste on absurdities like 1 gig drives, 486 CPU's, video cards with more than 128K of RAM, or all of the much more excessive crap the first world uses to do absurdly simple, basic, tasks, like send messages or analyze information.
That doesn't mean, however, that effective tools, that can benefit impoverished and starving peoples, don't need to be written and then run on simple (cheap) computers in the few places where power and electricity do exist, or that alternate infrastructure technologies (solar power, cheap radio transmission, etc) can't be deployed. They don't need "lickable", they don't need 3D, they'll do just fine with monochrome characters on 2 VT100's hooked into a 386 if it saves and improves lives.
Some Examples:
1. Route planning software for food/medicine deliveries in countries with unreliable transportation systems.
2. Accounting software to reduce the amount of graft and corruption involved in monetary aid. (If bribes are required, as least they can be tracked, budgeted for in the future, and properly expensed.)
3. Weather prediction software to help in crop planning.
4. Water management software to help prevent over-usage, which creates (artificial) droughts.
5. Simple communications hardware/software to relay information over low-tech, low-bandwidth, links. You don't need an email client->TCP/IP->DSL to send "Tornado coming, evacuate", you can do just fine with anything that does 300 (or even less) baud, like shortwave radio to a TTY, or, with human operators, 'net weather service->voice->shortwave radios work, too.:-)
6. Inventory management software for food supplies, medical supplies.
7. Simple message aggregation systems (remember why UUCP batched email was a good thing?) to speed up communications... (Village 197 needs penicillin brought in by jeep. Village 123 road out, needs 'dozer or shovels). Much faster than human messenger systems, and without the absurd overhead of the current crap we use.
8. Water testing/sanitation issues/road building/disease statistics/famine information/etc. aggregation systems, so the minimal resources available can be focused on the neediest areas.
9. Software to enable translation (and working translators) into *many* more languages than the computer world currently supports.
I'm surprised by the cynicism in many of the posts, which smack of first world arrogance. "Oh yeah, we'll send them our current technology, and they can't use it, because they lack our current infrastructure..." They need the computing ideas and technology of the 40's to 80's, not 2002's. Right now, many places are trying to survive on the same level of technology (or worse) than the US had in the late 1700's.
Come to think of it, the CPU cycles we *waste* on lickable interfaces, skinnable GUIs, looking for aliens, ripping crappy joke songs, or adding transparency to everything, could be better applied to *many* issues greater than the ones the first world faces.... with the CPU I use on a single protien fold sim, I could probably calculate *all* possible routes for a single food shipment to a village in Ghana.
Apple had a campaign centered around a current desktop being equivalent to what used to be a "supercomputer". If every third world country could now afford a 70's era supercomputer, they could be doing many of the basic tasks we all take for granted now. For only 10 grand, *any* nation in the world can afford the equivalent of two Cray-1's, and accomplish *all* of the computing tasks outlined above, with only two power cords, and one phone line. It's not about the infrastructure to send colored IM text to the next village, it's about extremely simple needs to help solve issues too complex to be done on paper.
It doesn't take a computer on every villager's "desktop", it takes computing power in the very few places that can benefit from it. Why would it need to be deployed in the same way? We already learned these lessons in the 60's, the first computing power to help the public was shared by very few people. It didn't look "pretty", it wasn't "fun", but that wasn't the point.
Addressing another point, those same Apple machines cannot be shipped to some countries, because they're considered powerful enough to be weapons. Again, we *also* learned this lesson in the 60's: The military wants first crack at all of the available CPU cycles, and will use them. Despots will use them t enrich their power... that's a side-effect, side problem, but as long as the CPU power is kept low enough, it'd still be legal.
non-MS?
Nope. They don't give a damn who the vendor is.
Not even the "non-MS desktop" ideology.
What they want is an easy to use computer. MS made apple's life hard by making their OS/apps "harder" to use with Macs, and vice versa.
Users don't care about about evil companies, when MS was found guilty, users didn't flee...
-Bop
The "voters" in most early democracies were the land"owners", those who "owned" the conduits (roads), those who "owned" the knowledge... the original voters were those withan "ownership" stake, not those who had any benefits.
Those who lived on the land, built the roads, or created the knowledge itself had no votes.
Slashdot Mirror
The reason why parent may (?) be misled is that networking stacks are not merely simple pieces of logic. Certainly there are some general rules... these are the very easy parts to program. In RFC terminology, these are the "shall" parts. Every packet that doesn't meet those rules, "should" (theoretically) be tossed.
/. to recognize a normal, well formed, email address of foo@example.com. That's quite easy to write a short regular expression for. (Go try it, I'll wait). All done? If your regular expression is under 200 characters or so, it's broken on a vast number of perfectly valid email addresses, or allowing through a vast number of illegal addresses. Here's a *still* partially broken regular expression (it can't handle inline comments) for, oh, "simply validating email addresses" based on the defined logic found in only *one* RFC. Not so simple with email, and it's not so simple with networking.
Where networking stacks get really ugly is not the mandatory, properly implemented, "shall" parts, it's the "may" parts, and the "shall" parts that have been misused, abused, or ignored, and the burden of handling pieces which *cannot* be broken into easily testable discrete units.
For an analogy about why networking stacks are so complex for something that seems deceptively simple, it's fairly easy for most people on
Oh, and by the way, networking code has to perform at absolute breakneck speeds, so you need to optimize the heck out of implementing those "simple" rules, too... and while it's certainly *possible* to validate every single bit, in every single packet, against a protocol's defined ruleset, I can pretty much guarantee you that you will rapidly find yourself consuming 100% of an average desktop's CPU before a 10Mbs stream can be completely checked, let alone used.
Couple of problems:
..Things that can change hundreds of time per second or so, well, keep that in a database designed for much slower use/access/queries, but with much more frequent update ability. Things that rarely change, or only change locally, keep them in a fast, small, and efficient directory.
1. You wanted a custom build, of your preferred tools, rather than using the existing tools. This is a problem with *any* "Enterprise Level" scenario.
2. Enterprise apps tend to use directory services (not dumb databases kludged to act like directories), wherever desirable. MySQL is great for lean databases, but it's not really suitable for a 6 million user mail architecture with possibly hundreds of mail servers. That's why you were being steered away from a slower database technology into a faster directory technology.
3. It's quite possible that you are one of the very few users who wanted to run a mail server architecture based on a slower *database* technology metaphor, rather than a faster *directory* technology metaphor.
To answer the inevitable objections:
1. Yes. LDAP is basically a front-end to a back-end "database". Unlike conventional database queries, however, LDAP is optimized to run like a bat out of hell on the equivalent of "SELECT" statements, often at the cost of "UPDATE" and "DELETE" statements, "JOIN", etc.
2. MySQL does not, and simply cannot, scale, in the same way that LDAP servers do. In an optimized enterprise, each LDAP server handles only the branch nodes it needs to think about. It's somewhat analogous to BIND, in that you can have 80 different LDAP servers in 80 different locations, each only storing the subset of records needed *for that location*. Ever try syncing 80 MySQL servers, while only having each node store its requred records, with pointers to branches further up? Good luck. This is why enterprise-level email usually requires enterprise-level directories.
3. Yes, you'd be duplicating records if you were storing user information in both directories, and MySQL databases. One of them should probably be removed, or split at a reasonable boundary. Hardcore LDAP users know that the boundary is on "what changes a lot", so things like a user's First name, which rarely ever changes, probably shouldn't be in a database designed to update that user name hundreds of times per second. Likewise with that user's current mail server, etc.
Thanks, we'll take it!
Buyers stay on eBay because there's a big selection of stuff and they can find what they want
I'm currently working for a company that opened their auction site only 4 months ago, and yet we already have 40,000 active listings at any given point in time. Not a ton of stuff, (lots of drop-ship and storefronts), but at least it's better than the many, many, roadkill sites out there that look like they just "gave up".
Sellers won't defect unless another site offers them some other savings...
Most of our prices are lower than ebay, if only by a penny, (and the first image is free... stupid ebay has all of those "photo" icons). For "power-users", we have a price plan that allows up to 1000 listings per year for $419, or 42 cents a piece, plus a flat 5.25% final value fee on any auctions (none of this "how much do I have to pay" silliness).
That is, they would have to pay the sellers a commission on each auction to attract them...
Interesting idea, but notice that the "free" venues such as craigslist can only support such a model by keeping bandwith *way* low. The economics of defection is certainly a tough nut to crack, though. We're giving away vacations to purchasers. *shrug*
We're about 80% done on that one. We made our entry page look quite similar to... uhm... some other searching tool.
Most definitely. Now that wikipedia has crunched through the hard work of "how can LAMP scale into a top 20 site?", it's fairly easy to see how it can be done with pages that are dynamic *to the second*, without needing anywhere near ebay's level of infrastructure.
Perhaps others have better ideas...
We're working on it....
For example:
-Free items list for free in our "free stuff" area. No limit on listings, items, etc. It's simple, and cheap.
-We're really trying to pay attention to UI. I don't think very many people would even feel like defending eBay's UI, but we have meetings about color schemes, table vs. CSS tradeoffs, etc.
-Did I mention the first photo is free? ;D
-No reserve fees.
-No minimum/starting bid fees.
-No "Buy It Now" fees.
-Sellers can choose to offer any payment method they want.
-Everyone in the company is customer service first, coder/designer/whatever second. Contrast to "we don't care, we already have your money" sites, of which there are many.
-RSS feeds for "stores" are pre-built for store users.
(etc. etc. etc.)...
Basically, we're trying to make something like...iBay(?), or what Apple would do if they did an auction/classified site. Offer simpler features, with better UI, resulting in a better end-user experience.
Correction: There is erosion with similar *characteristics* to water erosion on the sphinx. This doesn't mean the sphinx predates the surrounding desert climate.
:-) )
1). It could be water erosion, brought about by channeling/bringing water to the sphinx in a ritualistic fashion, even during desert times.
2). For those of you who have never lived in a desert, well, many deserts flood, and have severe damage, but only every 50/100/500 years. They are devastating enough to rip whole stone/concrete/etc. structures apart.
3). It could be erosion *similar to*, but not *indicitive of* water erosion, or even a combination of the two.
(Oh, wait, look, there are canals on mars! There *must* be people there!
Real science is not about drawing simple conclusions based on what we know, but about finding many possible conclusions based on what we currently know, and being ready to dismiss as "bad" any conclusions that are based on absurd prepositions, if we can find simpler prepositions.
Since most of the "old sphinx" tripe comes from such reputable folks as alien researchers, astrologists, and similarly accedited (which is to say: not at all) types, I find it generally more amusing (and of intellectual value) to point a finger at the general clown show and laugh. When a real scientist steps in (and they have), they're worth a listen, but they don't make similar absurd and bizarre "statements of fact"... because doing such a thing, quite frankly, isn't science. It's mere showmanship.
Yes, it seems like the sphinx is only 3,500 years old. Yes, there is odd weathering, that looks like water weathering. That doesn't mean that the sphinx is much older, or teleported to us from another planet, it means that we have a currently dated as 3,500 year old object, which requires new research, to figure out why it shows weathering signs of being dated older. To state it is much older *because* of unusual weathering... well, you might as well start telling your next-door neighbor about the anal probe the greys gave you.
-Ronabop
It's not really a broken link. A complete, working, version of the link is due out any day now, they're just in the final stages of debugging it.
SAN FRANCISCO Title: Bay Area Electronic Crimes Task Force Address: 345 Spear Street San Francisco, CA 94105 Phone: 415/744-9026 Fax: 415/744-9051 Contact: SA Erik Enos Office #: 415/744-9026 Alternate: SA Susan Broad Office #: 415/744-9026 Email: eenos@usss.treas.gov From: http://www.ectaskforce.org/regional_locations.htm
I suggest you investigate the earlier versions of the Oompa Loompas, before they were made a more politically correct green.
Seriously.
http://www.roalddahlfans.com/books/charoompa.php
Just imagine what life would be like for geeks if we could only choose the best "pre-fab" systems. No tuning. No alternate systems for some parts. No replacement of software or hardware. Just the top 20 "out of the box" systems.
If I was in charge, I'd run *all* entrants, award top prize to the best overall, but also keep the best traction group, best collision avoidance group, best navigation group (etc.) in my rolodex, and divide the problems out.
However, they are allowed to discriminate against pre-existing conditions... even if you didn't know about them.
So, as the genome is mapped, and causitive genes are found, they can drop you for a disease you may not get for 50 years. (If you live that long).
Privacy... yeah, right. Privacy and protection for those with a "non-deterministic" genome, maybe. Those who have causitive genes for later-life diseases are SOL.
Hitler was *not* voted in as dictator.
Since german political systems may be a bit hazy to americans, it worked like this:
His party got enough people elected in the 'congress' (reichstag). They then effectively voted to disband themselves, after appointing Hitler.
Nothing like an american election to install a dictator. The was no 'one person, one vote' to create a central leader.
-Bop
I wonder if I'm going to wake up tomorrow and find out that I'm an armed robber instead of a healthcare professional.
What, there's a difference? Are you not in the US?
Wanna diagram it?
I need a life with less ascii. And less spam words in my vocabulary.
So, I have this really ugly couch stuffed with 7 tons of down feathers, so it's now comfortable for everyone. To put it anywhere, I need a crane, and a team of 3 movers.
I don't mind though, because everyone *I* know who uses it for their guestbooks and web-to-mail forms and other wep-app-101 tutorial tasks is happy.
Funny thing, though, people who care about speed and performance and portability keep complaining about it.
What does is matter if it's big, heavy, ugly, and unportable... it's comfy!
What can I tell these poor souls who don't enjoy my couch, and want high-speed lightweight benches, or standing-room-only environments?
-Bop
K = Keycolor.
Two years re-reading all the GATF tech manuals for you, and, if it will help, making keylines.
I traded rubylith for this?
These are mostly maintainence jobs, which suck.
Is there a new wave of technology sweeping the industry, where cool new killer-apps are popping up like mad, because of new language idioms? I think not.
Can Java solve any single problem that wasn't solvable before? Of course not. It doesn't put new magic into the box, it's just another kind of magic.
This survey seems to be a measure of the buzzword bingo managers who are *still* not realizing the simple fact that: fads don't fix problems. I thought most of them dot-bombed in the 90's, but I guess we have some stragglers.
It doesn't matter what the fad language is, I can still do twice the work, in half the time, by cobbling together bash, perl, php, C, and whatever else is optimal to solve a problem. Any language which is slower to write in, or requires more design time, or stupid ROSE diagrams, ER committees, or 24 hour code reviews for 100 lines, is a *massive waste* of time.
Customers don't pay for design, theory, or somebody's philosophy, they pay for solutions. Now that hardware is cheaper than coders, design matters less than ever, as long as it can scale across boxes.
(Amusing note: I had a contract job where I distilled 300 lines of perl to one line of bash... the spec was re-written to not be language specific, and boom, I had replaced the perl bloat, where they re-implemented bash functions in perl).
Caveat: I get paid for solutions, not buzzword compliance. I don't do gigs where I waste hours on "modeling charts for the extremely daft".
-Bop
Or does python just not care about, oh, things like syntax?
-Depthchargingthekarmabop
Why should this even matter? (rhetorical)
What is so fragile about our 'net that one single organization can stop it? (conversational)
Why place eggs in one (censorware) basket (or even ten, or twenty), instead of hundreds (watermarked data containers, laundry systems, radio packets, pigeon-net, new encryption, new data hiding)?
Heck, if we can't easily stop it with obvious foes (China), what about more insipid ones? (DMCA)
I'm going to stop whining now, and keep coding.
I'm working on proto-coding various shared-secret (yeah, I know) based byte-use methods, so gathering an assembled message may be in 50 website gifs, an mp3 file, an ISO image, etc. etc.
Example: Byte 1297 of this porn image, Byte 16 of that slashdot header, Byte 1027,7189,1081 of a specific NT distro....so it's functionally impossible to censor all possible sources until a "reader key" is compromised (so, single-use keys are best). It's a WW2(1? prior?) transmission technology (words 197, 213,618,541 in a local newspaper), only the number of available sources to hide data has increased exponentially....
Even in Red Lobster's website... (ponder that for a second...)
Of course, this is probably *already* being done, so any links to existing sources would be helpful. :-) The hardest nut to crack (so far) seems to be gathering an accurate profile of the non-random destination user (to not trigger "abnormal behavior"... "hey, send me your browsing history, ji-chang, I like your p0rn"!), quickly find words/bytes/bits (or even chawmps, playtes, varying sizes to increase obsfuscation) in their "typical" stream to use, and, of course (as with all shared secrets) to transfer the key itself (pigeon-net! spaces/tabs in an HTML page (each space=0, tab=1)! and so on).
-WaitingForKnocksAtTheDoorBop
1. No. The *nix geeks regularly evangelize on every support line, in every tip site, and every magazine. They know it's there, even Apple pounds it into them.
2. Yes. The fear is there.
Apropos of this, I suggest one of the most important things these students can learn is that it's okay to do things that break, that wedge, that "go boom" in funny ways, because they can actually fix it themselves with OSS, or at least, "add eyeballs" to the problem!
Don't forget that some of Mac OS X's PPP issues were resolved by a teenager...
-Bop
Hm. Your family foods are being defined as a quantifiable "service". With children and elderly defined by cost.
My family is not a cost (or revenue) center. Good food is much more important than cheap food, even if it's web delivered.
If you can't pack dinner for a family of five, get in shape! It's only 8 pounds or so!
Is lugging 8 pounds 4 blocks more important than your family?
Good food matters more to me than the convenience.
-Bop
Two week old food. Yum! -Bop
You might try it.
In addition, have you considered that the "big-box" grocery may have lower quality foods?
Get meat at the butcher, fruit at the fruit stand, all on the way home....
It takes me about 15 minutes a day to get good food.
On foot.
You might realize, one day, that there are hundreds of small businesses who aren't big box.
Give them a try.
Or use what most people use, buy where most people buy, and then wonder why you're reading /. :)
-Ronabop
Running into people, crashing cars, etc, is much more deadly, but it's not the same thing, is it?
Until the US can intelligently think about such silly 'slippery slope' arguments, every thing that could have fatal consequences, based on correlative, not causative, statistics, is up for grabs.
Let's apply that same enlightened thinking of 'drunk driving is deadly" to other items, shall we?
Linux/BSD users are (or will be) crackers, and should be in jail.
People who express anti-government opinions now are future terrorists, and should be in jail.
If you have a single illegal mp3, you are probably a future multi-billion-dollar-pirate, and should be in jail.
By all means, punish those who commit crimes against others, but to arrest someone *only* for drunk driving is to arrest them for future-crime, for the "potential" and "likelihood" of harming others.
I live in AZ, where 'drunk driving' is one (1) beer before driving. And I use good computers, which tend to kill less people than cars, and I own a handgun.
Three strikes against me on future-crime, I guess.
-Bop
They don't have the money to waste on absurdities like 1 gig drives, 486 CPU's, video cards with more than 128K of RAM, or all of the much more excessive crap the first world uses to do absurdly simple, basic, tasks, like send messages or analyze information.
That doesn't mean, however, that effective tools, that can benefit impoverished and starving peoples, don't need to be written and then run on simple (cheap) computers in the few places where power and electricity do exist, or that alternate infrastructure technologies (solar power, cheap radio transmission, etc) can't be deployed. They don't need "lickable", they don't need 3D, they'll do just fine with monochrome characters on 2 VT100's hooked into a 386 if it saves and improves lives.
Some Examples: :-)
1. Route planning software for food/medicine deliveries in countries with unreliable transportation systems.
2. Accounting software to reduce the amount of graft and corruption involved in monetary aid. (If bribes are required, as least they can be tracked, budgeted for in the future, and properly expensed.)
3. Weather prediction software to help in crop planning.
4. Water management software to help prevent over-usage, which creates (artificial) droughts.
5. Simple communications hardware/software to relay information over low-tech, low-bandwidth, links. You don't need an email client->TCP/IP->DSL to send "Tornado coming, evacuate", you can do just fine with anything that does 300 (or even less) baud, like shortwave radio to a TTY, or, with human operators, 'net weather service->voice->shortwave radios work, too.
6. Inventory management software for food supplies, medical supplies.
7. Simple message aggregation systems (remember why UUCP batched email was a good thing?) to speed up communications... (Village 197 needs penicillin brought in by jeep. Village 123 road out, needs 'dozer or shovels). Much faster than human messenger systems, and without the absurd overhead of the current crap we use.
8. Water testing/sanitation issues/road building/disease statistics/famine information/etc. aggregation systems, so the minimal resources available can be focused on the neediest areas.
9. Software to enable translation (and working translators) into *many* more languages than the computer world currently supports.
I'm surprised by the cynicism in many of the posts, which smack of first world arrogance. "Oh yeah, we'll send them our current technology, and they can't use it, because they lack our current infrastructure..." They need the computing ideas and technology of the 40's to 80's, not 2002's. Right now, many places are trying to survive on the same level of technology (or worse) than the US had in the late 1700's.
Come to think of it, the CPU cycles we *waste* on lickable interfaces, skinnable GUIs, looking for aliens, ripping crappy joke songs, or adding transparency to everything, could be better applied to *many* issues greater than the ones the first world faces.... with the CPU I use on a single protien fold sim, I could probably calculate *all* possible routes for a single food shipment to a village in Ghana.
Apple had a campaign centered around a current desktop being equivalent to what used to be a "supercomputer". If every third world country could now afford a 70's era supercomputer, they could be doing many of the basic tasks we all take for granted now. For only 10 grand, *any* nation in the world can afford the equivalent of two Cray-1's, and accomplish *all* of the computing tasks outlined above, with only two power cords, and one phone line. It's not about the infrastructure to send colored IM text to the next village, it's about extremely simple needs to help solve issues too complex to be done on paper.
It doesn't take a computer on every villager's "desktop", it takes computing power in the very few places that can benefit from it. Why would it need to be deployed in the same way? We already learned these lessons in the 60's, the first computing power to help the public was shared by very few people. It didn't look "pretty", it wasn't "fun", but that wasn't the point.
Addressing another point, those same Apple machines cannot be shipped to some countries, because they're considered powerful enough to be weapons. Again, we *also* learned this lesson in the 60's: The military wants first crack at all of the available CPU cycles, and will use them. Despots will use them t enrich their power... that's a side-effect, side problem, but as long as the CPU power is kept low enough, it'd still be legal.
-Bop
non-MS?
Nope. They don't give a damn who the vendor is.
Not even the "non-MS desktop" ideology.
What they want is an easy to use computer. MS made apple's life hard by making their OS/apps "harder" to use with Macs, and vice versa.
Users don't care about about evil companies, when MS was found guilty, users didn't flee...
-Bop
Those who lived on the land, built the roads, or created the knowledge itself had no votes.
Wonder why societies changed that?
-Bop