I'm a cut-n-paste SQL writer, so I may go off into left field here quickly...
The condition I was referring to, which was called nested looping by the DBAs and development lead at the company I worked at, is the condition where one query causes additional queries to be performed. Think queries against dynamic view tables. In this case, the nightmare query went four or five levels deep. I recently saw an Oracle trace in which a data view was unrolling to 88 levels deep...
Keep in mind this was over a year ago and it probably doesn't represent best coding practice. However, the most popular supported platform performed just fine with it, which tends to make a company loathe to spend much time finding workarounds for the other platforms.
SQL Server gets a lot of flack on/., I'd be interested to know why? I've worked a fair amount with it, Oracle 8 and 9, Postgres, and a little bit of MySQL. I've also done extensive benchmark testing of SQL/Oracle/Postgres handling the same load on the same hardware (shh, don't tell Oracle:).
My experience leads me to beleive the following things: 1) MS-SQL is a high quality database that is ridiculously easy to set up, tune, and maintain. It is also very expensive. 2) Postgres is a high quality database that is ridiculously easy to set up and maintain, but fairly difficult to tune. However, its performance is just as good as SQL server as long as you stay away from nested loops(*). It is also fairly inexpensive (free license, but increased TCO). 3) Oracle is a pig, and it requires a professional, certified swineherd. If you spend an amazing amount of money on licenses, gear, and certified DBAs you will presumably get good performance; I however was never able to get it past 60% of the performance of MS-SQL or Postgres.
(*) Nested loops are like candy to SQL server, and I've heard this is the same for Sybase (understandably). Deep sets of nested loops will kick the other databases I've tested in the teeth. Given an instruction with several nested loops and 16 million rows of data, I got results from SQL server in 5 minutes, results from Oracle 9 in an hour, and results from Postgres in 18 hours. This was a year ago and Postgres has changed, so it might be better now. Does MySQL handle them well?
Working with people on computers? Triple Suckage. Working with computers for people? Triple Suckage. Making a computer help people work? You guessed it.
All computers suck for usability because it's not clear what we want to use them for, because we're people, and people suck. Moreover, a lot of things we say we want to use them for are some sort of work, which sucks.
I find that life sucks less when ample time is reserved for getting away from computers, people, and work.
And yes, I do use all three platforms concurrently (Rick Wakeman, eat your heart out!), and have used all three as a primary desktop at one point or another. OS X sucks just as much as Windows or Linux.
the second product is clearly a knock-off of the first. Patent system is working as it should in this case, move along.
Besides which, it's hardly rocket-science to put together something like this without infringing the patent -- perhaps even with a compelling feature like non-iPod-specificity. That way those of us with better players can use it now, and the iPod owners can use it next year when the little interface at the bottom changes pin-outs.
Linux owns a large chunk of the data center, and if it wasn't for Red Hat's bungling with over-priced licenses and Fedora Community, it would own a lot more in the US.
However, it's already lost the battle for alternative desktop in my opinion, because it's losing geeks-with-jobs. Student geeks still run Linux desktops; people with deadlines run Windows or Mac.
And for those who don't know who jwz is, he wrote Netscape's original Unix port, xscreensaver, and large portions of XEmacs. He also owns a nightclub in San Francisco (DNA Lounge).
I like Sunbird too, even though it won't start on my OS X box anymore...
iCal never made it past the configuration dialog -- I typed in the https URL to my WebDAV share and it said "we doan do no steenken SSL" and I said thanks for playing.
It's not about the chip in the box, it's about selling and then providing an smooth computing experience and a feeling of "I am cool." As long as they maintain their tight control over the hardware stack and ensure that their lifestyle products integrate well and work smoothly, it doesn't matter what they use for a CPU, GPU, USB EHCI controller, hard drive manufacturer, &c, &c....
Another way to think about it is "outsourcing". I bet they're offloading a significant chunk of R&D cost onto Intel with this move.
And Multimedia CD-ROMs before that. Funnily enough, new technology doesn't do much to change who we are -- those who are interested in breaking down borders will do so anyway, while the rest of the world is quite happy to build higher borders.
Take the MAC addresses of your problem children and give them reserved addresses in a blackhole net that has no Internet access. Sure, the smart ones can snag their own static IPs, but the smart ones aren't your problem (and are probably already doing that).
Telephony For Computer Professionals, by Janet Laino... Amazon link here.
Excellent resource for this very purpose, and the fact that it predates VoIP means that you can compartmentalize the one technology from the other (first understand what you mean to emulate, then understand the emulation).
He was an SE from one of the larger security vendors who claimed to be an expert with their packet sniffer, but couldn't explain ARP, DHCP, or a TCP handshake. Now in theory he didn't need those skills for doing the job we had open, but in practice understanding networking at least well enough to call bullshit when you're being lied to was a crucial requirement. Thanks for playing.
yeah -- it's via word of mouth for me though, since it doesn't have an auto-find-servers function like q3a and I'm not going to buy gamespy. My friends tear me up quite well without help.
With GL graphics the game looks quite nice IMHO -- explosions and such aren't as nice as with newer games, and the characters can look clunky, but the maps are well designed and a pleasure to play in.
yup -- I play games so I can stop thinking and spend a little time letting the lizard brain deal. So they need to be fast, with simple controls and all the complexity embedded in matters of gameplay. Quake2 and BZFlag are the ones that fit that bill, and both are cross-platform and lightweight enough to play well on a laptop.
_Pattern Recognition_ wasn't bad, but it wasn't that good either -- sad fact is that William Gibson got lucky by writing _Neuromancer_ at the right time to generate his fifteen minutes. Still, I'd rather read an okay book like _Pattern Recognition_ than some other books.
David Brin is another so-so writer, in my opinion -- not usually bad, but never really making it to great. Hey, we can't all be the best thing since sliced bread, and there's a place for "pretty good". He's usually good for ideas, but with a tin ear for dialogue and character -- however, if you love Arthur C. Clarke and Isaac Asimov, you probably won't be bothered by Brin's shortcomings.
it's Mieville, not Melville, and yes, he's over-rated. I like Perdido St. Station, but it was not even within shouting distance of _Jonathan Strange & Mr. Norrell_.
Ditto for Charles Stross -- _Atrocity Archives_ is an enjoyable read, but _Singularity Sky_ is a third-rate potboiler. It's as if Tom Clancy decided to write sci-fi, yech.
the dirty little secret of dependency management systems is that they have to be managed. As a desktop user or server administrator, you have very little control over that management process. In other words, you're putting your trust into the distribution's package management team. They need to package sanely, define dependencies sanely, and handle changes sanely. This is a lot to ask for a distribution that follows the current state of Linux; it becomes brutally difficult when the same people are trying to track current state plus everything that's happened in the last three years.
My own distribution of choice, Mandrake, illustrates this problem well -- stick with the current release, and your packaging problems will be rare. The biggest challenge I face in maintaining my current Mandrake systems is keeping my mirror list up to date. However, if you let a system fall back a release or two from current, even while still in the "supported" window, you've got a problem. Package naming and numbering conventions change, people quit being so careful when building update packages, and the "screw backward compatibility" decisions of hundreds of FOSS developers start to add up. Eventually, there's nothing for it but to upgrade to current or wipe the system drive and reinstall.
The tool used to interface with all of this is irrelevant. urpmi, apt-get, emerge, make (port), cast, yum, red-carpet, whatever. Might as well compare the quality of hammers when the question is whether the boards to be nailed are actually lined up.
Well, there's also the fact that a lot of Unix/Linux people haven't actually touched Windows in a meaningful way since Win98 was the new hotness. If you've never managed a Windows environment, you're not really qualified to talk about how and why it sucks IMHO. It does suck, but often not for the reasons that are given by people with no experience.
OK, you're inducing the facial tics again... Been there, done that, it wasn't fun.
The only sane way to be making this sort of decision is to benchmark, because everything relies on how the existing SQL is written. If you don't have time and money to benchmark, then follow this simple metric:
a) who's the lead database programmer? b) which database does s/he write for first?
Pick that one, because you can bet that the programmer has been picking up all sorts of database-specific tips and tricks for the last few years. The SQL they've generated may actually work on another database, but its performance is probably going to stink. This will become clear as soon as you get around to benchmarking.
If you're at a point where rewriting all that code becomes a realistic option, I'm sorry. Anyway, you're still going to be a lot better off by following the same metric, unless you want to add "waiting for the developers to learn how to optimally use the new database" to your soft costs. And since these are developers, you'll also need to add "waiting for the developers to realize that they need to learn how to optimally use the new database", which can take a full release cycle.
There's a lesson here, and it's this: "being right is not the same thing as being successful."
Most successful customer-facing people have learned to combine these approaches; act like #2 in front of the customer, but be careful to use some qualifying terms. "No problem, we should have this done in two days". Keep the rest of the sentence "if pigs start flying" under your hat, they don't really need to know. If they're smart enough to pry, then tell them the truth, but 99% just want to hear you say what they want to hear. They're not going to be listening to you anyway and will be perfectly happy with "blah blah two days blah blah". Do not discuss pricing. Do not discuss pricing. If you feel you must answer a pricing question, give them an insane range and when they complain, tell them you don't know anything about pricing and to ask the person who ought to be handling pricing. There's nothing for you to gain from mentioning what you think it will cost before you know what it will really cost.
When you're back to the shop or working with internal process, then you're #1 all the way -- do the process, do it right. You're not going to get anything but enemies by making your coworker's jobs harder.
Absolutely -- I love my 120 passionately. Better sound quality than an iPod, doesn't need iTunes, doesn't need DRM, and continues to work just great after a year of daily use.
I'm a little confused by this whole article though, as I thought Rockbox was for the Archos, not iRiver. Can't say that I'm likely to replace firmware anyway, as the iRiver firmware is working just fine. The most I'd look for in a different firmware is better fonts.
Slashdot Mirror
I'm a cut-n-paste SQL writer, so I may go off into left field here quickly...
The condition I was referring to, which was called nested looping by the DBAs and development lead at the company I worked at, is the condition where one query causes additional queries to be performed. Think queries against dynamic view tables. In this case, the nightmare query went four or five levels deep. I recently saw an Oracle trace in which a data view was unrolling to 88 levels deep...
Keep in mind this was over a year ago and it probably doesn't represent best coding practice. However, the most popular supported platform performed just fine with it, which tends to make a company loathe to spend much time finding workarounds for the other platforms.
SQL Server gets a lot of flack on /., I'd be interested to know why? I've worked a fair amount with it, Oracle 8 and 9, Postgres, and a little bit of MySQL. I've also done extensive benchmark testing of SQL/Oracle/Postgres handling the same load on the same hardware (shh, don't tell Oracle :).
My experience leads me to beleive the following things:
1) MS-SQL is a high quality database that is ridiculously easy to set up, tune, and maintain. It is also very expensive.
2) Postgres is a high quality database that is ridiculously easy to set up and maintain, but fairly difficult to tune. However, its performance is just as good as SQL server as long as you stay away from nested loops(*). It is also fairly inexpensive (free license, but increased TCO).
3) Oracle is a pig, and it requires a professional, certified swineherd. If you spend an amazing amount of money on licenses, gear, and certified DBAs you will presumably get good performance; I however was never able to get it past 60% of the performance of MS-SQL or Postgres.
(*) Nested loops are like candy to SQL server, and I've heard this is the same for Sybase (understandably). Deep sets of nested loops will kick the other databases I've tested in the teeth. Given an instruction with several nested loops and 16 million rows of data, I got results from SQL server in 5 minutes, results from Oracle 9 in an hour, and results from Postgres in 18 hours. This was a year ago and Postgres has changed, so it might be better now. Does MySQL handle them well?
Computers... suck.
People... suck.
Work... sucks.
Working with people on computers? Triple Suckage.
Working with computers for people? Triple Suckage.
Making a computer help people work? You guessed it.
All computers suck for usability because it's not clear what we want to use them for, because we're people, and people suck. Moreover, a lot of things we say we want to use them for are some sort of work, which sucks.
I find that life sucks less when ample time is reserved for getting away from computers, people, and work.
And yes, I do use all three platforms concurrently (Rick Wakeman, eat your heart out!), and have used all three as a primary desktop at one point or another. OS X sucks just as much as Windows or Linux.
the second product is clearly a knock-off of the first. Patent system is working as it should in this case, move along.
Besides which, it's hardly rocket-science to put together something like this without infringing the patent -- perhaps even with a compelling feature like non-iPod-specificity. That way those of us with better players can use it now, and the iPod owners can use it next year when the little interface at the bottom changes pin-outs.
Linux owns a large chunk of the data center, and if it wasn't for Red Hat's bungling with over-priced licenses and Fedora Community, it would own a lot more in the US.
However, it's already lost the battle for alternative desktop in my opinion, because it's losing geeks-with-jobs. Student geeks still run Linux desktops; people with deadlines run Windows or Mac.
And for those who don't know who jwz is, he wrote Netscape's original Unix port, xscreensaver, and large portions of XEmacs. He also owns a nightclub in San Francisco (DNA Lounge).
I like Sunbird too, even though it won't start on my OS X box anymore...
iCal never made it past the configuration dialog -- I typed in the https URL to my WebDAV share and it said "we doan do no steenken SSL" and I said thanks for playing.
It's not about the chip in the box, it's about selling and then providing an smooth computing experience and a feeling of "I am cool." As long as they maintain their tight control over the hardware stack and ensure that their lifestyle products integrate well and work smoothly, it doesn't matter what they use for a CPU, GPU, USB EHCI controller, hard drive manufacturer, &c, &c....
Another way to think about it is "outsourcing". I bet they're offloading a significant chunk of R&D cost onto Intel with this move.
And Multimedia CD-ROMs before that. Funnily enough, new technology doesn't do much to change who we are -- those who are interested in breaking down borders will do so anyway, while the rest of the world is quite happy to build higher borders.
environmental chilling... fire the PC into space, then use a very long KVM cable-set. What's even radder is that the KVM doubles as a space elevator!
"an article like this does a disservice because it's misleading the /. crew. ...in its current state it changes nothing for the average end-user."
/. crew is a bunch of average end-users with a lot of attitude... Hopefully, that isn't true yet.
So the
Take the MAC addresses of your problem children and give them reserved addresses in a blackhole net that has no Internet access. Sure, the smart ones can snag their own static IPs, but the smart ones aren't your problem (and are probably already doing that).
people suck. work sucks. computers suck.
what do you think happens when people work on/with/for computers?
The suckitude is cubed!!!
Telephony For Computer Professionals, by Janet Laino... Amazon link here.
Excellent resource for this very purpose, and the fact that it predates VoIP means that you can compartmentalize the one technology from the other (first understand what you mean to emulate, then understand the emulation).
I think I interviewed that guy...
He was an SE from one of the larger security vendors who claimed to be an expert with their packet sniffer, but couldn't explain ARP, DHCP, or a TCP handshake. Now in theory he didn't need those skills for doing the job we had open, but in practice understanding networking at least well enough to call bullshit when you're being lied to was a crucial requirement. Thanks for playing.
yeah -- it's via word of mouth for me though, since it doesn't have an auto-find-servers function like q3a and I'm not going to buy gamespy. My friends tear me up quite well without help.
With GL graphics the game looks quite nice IMHO -- explosions and such aren't as nice as with newer games, and the characters can look clunky, but the maps are well designed and a pleasure to play in.
yup -- I play games so I can stop thinking and spend a little time letting the lizard brain deal. So they need to be fast, with simple controls and all the complexity embedded in matters of gameplay. Quake2 and BZFlag are the ones that fit that bill, and both are cross-platform and lightweight enough to play well on a laptop.
_Pattern Recognition_ wasn't bad, but it wasn't that good either -- sad fact is that William Gibson got lucky by writing _Neuromancer_ at the right time to generate his fifteen minutes. Still, I'd rather read an okay book like _Pattern Recognition_ than some other books.
David Brin is another so-so writer, in my opinion -- not usually bad, but never really making it to great. Hey, we can't all be the best thing since sliced bread, and there's a place for "pretty good". He's usually good for ideas, but with a tin ear for dialogue and character -- however, if you love Arthur C. Clarke and Isaac Asimov, you probably won't be bothered by Brin's shortcomings.
it's Mieville, not Melville, and yes, he's over-rated. I like Perdido St. Station, but it was not even within shouting distance of _Jonathan Strange & Mr. Norrell_.
Ditto for Charles Stross -- _Atrocity Archives_ is an enjoyable read, but _Singularity Sky_ is a third-rate potboiler. It's as if Tom Clancy decided to write sci-fi, yech.
the dirty little secret of dependency management systems is that they have to be managed. As a desktop user or server administrator, you have very little control over that management process. In other words, you're putting your trust into the distribution's package management team. They need to package sanely, define dependencies sanely, and handle changes sanely. This is a lot to ask for a distribution that follows the current state of Linux; it becomes brutally difficult when the same people are trying to track current state plus everything that's happened in the last three years.
My own distribution of choice, Mandrake, illustrates this problem well -- stick with the current release, and your packaging problems will be rare. The biggest challenge I face in maintaining my current Mandrake systems is keeping my mirror list up to date. However, if you let a system fall back a release or two from current, even while still in the "supported" window, you've got a problem. Package naming and numbering conventions change, people quit being so careful when building update packages, and the "screw backward compatibility" decisions of hundreds of FOSS developers start to add up. Eventually, there's nothing for it but to upgrade to current or wipe the system drive and reinstall.
The tool used to interface with all of this is irrelevant. urpmi, apt-get, emerge, make (port), cast, yum, red-carpet, whatever. Might as well compare the quality of hammers when the question is whether the boards to be nailed are actually lined up.
Well, there's also the fact that a lot of Unix/Linux people haven't actually touched Windows in a meaningful way since Win98 was the new hotness. If you've never managed a Windows environment, you're not really qualified to talk about how and why it sucks IMHO. It does suck, but often not for the reasons that are given by people with no experience.
OK, you're inducing the facial tics again... Been there, done that, it wasn't fun.
.
The only sane way to be making this sort of decision is to benchmark, because everything relies on how the existing SQL is written. If you don't have time and money to benchmark, then follow this simple metric:
a) who's the lead database programmer?
b) which database does s/he write for first?
Pick that one, because you can bet that the programmer has been picking up all sorts of database-specific tips and tricks for the last few years. The SQL they've generated may actually work on another database, but its performance is probably going to stink. This will become clear as soon as you get around to benchmarking.
If you're at a point where rewriting all that code becomes a realistic option, I'm sorry. Anyway, you're still going to be a lot better off by following the same metric, unless you want to add "waiting for the developers to learn how to optimally use the new database" to your soft costs. And since these are developers, you'll also need to add "waiting for the developers to realize that they need to learn how to optimally use the new database", which can take a full release cycle
Score funny? What for? Looks pretty insightful to me.
There's a lesson here, and it's this: "being right is not the same thing as being successful."
Most successful customer-facing people have learned to combine these approaches; act like #2 in front of the customer, but be careful to use some qualifying terms. "No problem, we should have this done in two days". Keep the rest of the sentence "if pigs start flying" under your hat, they don't really need to know. If they're smart enough to pry, then tell them the truth, but 99% just want to hear you say what they want to hear. They're not going to be listening to you anyway and will be perfectly happy with "blah blah two days blah blah". Do not discuss pricing. Do not discuss pricing. If you feel you must answer a pricing question, give them an insane range and when they complain, tell them you don't know anything about pricing and to ask the person who ought to be handling pricing. There's nothing for you to gain from mentioning what you think it will cost before you know what it will really cost.
When you're back to the shop or working with internal process, then you're #1 all the way -- do the process, do it right. You're not going to get anything but enemies by making your coworker's jobs harder.
Absolutely -- I love my 120 passionately. Better sound quality than an iPod, doesn't need iTunes, doesn't need DRM, and continues to work just great after a year of daily use.
I'm a little confused by this whole article though, as I thought Rockbox was for the Archos, not iRiver. Can't say that I'm likely to replace firmware anyway, as the iRiver firmware is working just fine. The most I'd look for in a different firmware is better fonts.
3 minutes to server death. I'm sure it was an interesting presentation anyway :)