Personally, I have my taxes done by a very good tax attorney
Who will type your info into his copy of TurboTax.
If you have somebody who does your taxes the way you like and is willing to stand up to the IRS for you, then that's terrific. My luck has been (ahem) less than good.
By the time I've collected all the data and explained the picky little details of my business to the tax "professional", I might as well have done it myself. "Audit? Yeah, I'll go in with you for $150/hour. No guarnatees, of course."
I agree with you completely about preferential voting, "instant runoff", Australian voting, whatever you want to call it.
HOWEVER, right now the protests against the current crop of fraud-o-trons is tightly focused. This is good; your friendly average election official has the attention span of a hyperactive four-year-old. Tell him two things at once and he won't remember either one.
Experience from many years of various protests -- keep to ONE ISSUE and you'll win. Spread it out (hey, why not eliminate the Electorial College while we're at it?) and you'll lose.
I assure you, the people who manufacture these horrors aren't gonna lose focus.
Epic is doing this so they can say, when they get back the bashed-open CD players, "Look at how Evil all those reviewers are!" Then they can blame the reviewers for the music "escaping". Since the ones probably putting the music on the P2P nets are the guys who mixed it it the first place, it lets Epic shove responsibility off onto someone else.
Or maybe they're just saying that these albums don't deserve anything better.
The ideal solution would be for the reviewers to not review them at all. See what that does to sales.
The real question is, is it easier to teach geeks to train, or to train trainers the technical stuff?
My experience (four years teaching Cisco courses) is that it's possible to teach geeks to train, but it's almost impossible to teach trainiers enough technical information to answer questions.
If you have a non-geek trainer, you might as well just read the book. You're never going to get anything beyond that.
The main benefit that I see to ant- spam laws is that it would put pressure on ISPs to get rid of the spammers. Spammers aren't going to file lawsuits against ISPs to let them continue an illegal operation.
They aren't going to pack up and physically move to Korea, either.
"The Mythical Man-Month", Brooks. Won't really help much, but you'll have the satisfaction of knowing exactly how your pointy haired managers are screwing things up.
"Design Patterns", Gamma, et al. Without this, you simply won't be able to understand current discussions about programs or programming. This book gives you the philosophy and vocabulary to understand what's going on.
"The Art of Computer Programming", Knuth. What can I say? An absolutely mindboggling treasure trove.
"Software Tools", Kernighan & Plauger. A Golden Oldie. The book is ancient, but the "software tool" concept is still solid.
"The Design of the Unix Operating System", Bach, and "The Design and Implememtation of the 4.4 BSD Operating System", McKusick et al. (These are old. I would hope there is something equivalent for Linux and current BSDs). While abstraction is all well and good, at some point you have to open up the black box and figure out what the machine is actually doing in there.
You need the definitive description of the language you're working with. For C, it's "The C Programming Language", Kernighan & Richie. For C++ it's "The C++ Programming Language", Stroustrup, or, if you're a standards junkie like me, INCITS/ISO/IEC 14882-1998, "ANSI Standard C++".
If you're doing anything connected with the Internet, learn about RFCs. Personally, I credit a large part of the success of the Internet to the free availability of its governing standards. (Other standards are freely available, but not available for free. A paper copy of ISO 14882, for example, is US$175.)
There are all sorts of "domain specific" books. What you need depends on what you're doing. I find "Advanced CORBA Programming with C++", henning & Vinoski, to be priceless, but then, I do CORBA programming in C++.
They can't give it away for free. It has value. There are investors who paid hard cash for the development of the code and while they now know that they're not going to get it all back, they'd like to recoup some of their costs at least.
No.
The only "value" anything has is what somebody is willing to pay for it. If nobody is willing to buy something, it has a value of zero.
"But we paid a lot to develop this!". Tough. Just because you want something to be worth a lot of money doesn't mean it is. Want to buy some WorldCom stock? The proper business term is "fully depreciated".
What is happening is that NaN is carrying something on the books as an "asset" that has no value. Gotta keep the stock price up....
I've had the same problem. It's totally bogus, of course:
You're gonna sue Microsoft. Or IBM. Or CA. Right. Tell me another one.
Go back and read your EULA. Basically, they warrant that the media is Genuine Media. If you happen to put the disk in a reader, and there happens to be some software on it, and if you happen to install it, you're on your own.
Point this out, and the argument changes to another bogus anti- open source rant.
From what I've seen in various places, he's trying to describe everything in terms of cellular automata.
But a cellular automaton is just the discrete version of a differential equation. Physics has been described in terms of differential equations for 150 years.
So is he getting any new insights? Simplifying the calculations? Or is he just rearranging the the mathematical furniture?
... why we should care? What have future civilizations ever done for us, after all?
A simple trefoil will explain to anybody what's there, as long as there are any vestiges of our current civilization left anywhere. It'll take a *long* time for the trefoil to be forgotten.
We don't worry about anything else that might affect our long-distant ancestors. Why this?
This hoohah about radioactive waste is nothing more than another bogus "reason" to scream "No Nukes!" without having to think. (Keeping nuclear waste where it is now is *ever* so much safer than putting it in the middle of the Nevada desert, after all!) Public attitudes toward nuclear power have been formed more by monster movies than any rational thought process.
IMHO, nuclear power is a Bad Idea, but for boring old economic reasons. No need to get hysterical.
Two reasons. Minor reason: It Is Assumed that a "better algorithm" will use "more resources", whether it does or not. (Defintions of the terms in quotes are optional.)
Major reason: Defense contractors cannot make money by building a system strictly as specified. Defense contractors make boxcar loads of money on change orders. The normal sequence here would be:
Build the system strictly to spec.
Write a discrepency report identifying a potential clock problem and outlining a fix.
Get a change order and a boatload of money from the government.
Do it the way it should have been done in the first place.
Any reasonably sized system will have thousands of these change orders.
The Patriot time-drift was caused by the system being operated outside of its dsign parameters. It was designed to operate during a Soviet invasion of Western Europe, and expected to have to relocate every 8 hours or so. The spec, therefore, assumed that the software would reboot every 8-12 hours. From my experience with the military, if a programmer had put in a clock algorithm that would track indefinitely, he or she would have been ordered to take it out. (Been there. Done that. Broke the coffee mug.)
The Yorktown crash was the result of mixing mission-critical and non-mission-critical programs on the same box. Big no-no.
So we have a specification problem and a system design problem. Neither is a pure "programming problem".
Software crashes are like airplane crashes -- blame the lowest guy on the totem pole. In air crashes, it's the pilot. In software, it's a coder.
Another point -- require them to keep copies of any hardware needed to play the media, and require them to provide hardware to law enforcement, government archives, or the copyright office at any time. This would last for the lifetime of the copyright. As a hedge against them going out of business, the hardware (and all specifications needed to rebuild it if necessary) should be escrowed somehow.
This is a serious problem already -- according to a friend at the US National Archives, they're already having problems getting players for seven inch open-reel audio tapes.
IIRC, this is exactly what happened when Keith Henson was sued by the Church of Scientology. He spread around a bunch of their "secret scriptures" advocating illegal action and they sued him for breach of trade secrets.
With no worries about antitrust prosecution, we're going to see a lot more of this stuff. We've already seen them state flat out that "donated" computers must have a legal Microsoft OS and attack the GPL directly; more FUD will surely follow.
The only question is how far their "customers" can be pushed. My guess is pretty far. Never underestimate the pointy-hair factor. Most places, "learning something new" is interpereted as "complete retraining". PHBs regard doing anything new the way a nun would regard going to work in a brothel.
About the only thing we can do is to make sure Open Source solutions don't get wired out due to:
1. Laws or standards that mandate the use of patented/licensed technology. (*Must* use GIF, *must* pony up US$5000 to Unisys.)
2. Laws that specify "maufacturer's liability" (release an Open Source program; get sued if somebody doesn't like it.)
3. Laws mandating DRM hardware/software.
I'm sure we're going to see a flood of these from the Microsoft keiretsu.
We run air over calcium oxide, which absorbs CO2 and turns into calcium carbonate. We then heat the calcium carbonate to turn it back into calcium oxide and CO2. Now we react the CO2 with something to take it out of the actmsphere. Like, say, calcium oxide.
How do we make calcium oxide in the first place? Heat calcium carbonate to drive off the CO2.
Looks to me like the greenhouse gas version of perpetual motion.
is Microsoft Outlook with a bunch of plugins. I rather suspect that they've never used anything except Outlook and Outlook Express for serious work. Their "wish list" is almost all user interface doodles and twiddles.
Problems with Outlook and OE are obvious to/.ers:
Very poorly designed overall
Wildly insecure
Wired to other proprietary products like Exchange Server
I've considered writing my own mail/news client, but it seems like an awful lot of work for relatively little return -- current programs are Almost Good Enough. Anyway, what I'd look for:
It should work correctly. Duh! But it's amazing what doesn't work in current clients. At the very least, it shouldn't crash and should never, never, never corrupt stored mail or lose mail from the server.
Modular system architecture that makes maximum use of plugins. I can't think up all the features an e-mail client could use and I certainly can't implement all of them.
Configurable everything. This implies a good configuration structure and an editor that lets a user actually set things up the way s/he wants.
Data exchange with other programs, using standard formats wherever possible. Contact lists are first priority, with calendars a close second.
Filtering. Current filtering systems are simply not very powerful. At the very least, should have full regular expressions on all header fields and the body of the message (bodies, for multipart messages), plus the ability to flag, prioritize, download, autoreply, sort, delete, file, etc, etc, the sorted messages. A nice interface that makes this all usable will be a real challenge.
Fine grained control over what applications get invoked with attachments. Must be independant of main system. When I double click on a.doc file in the main system, I want to open Word. When I get an application/msword attachment, I do not want to open Word.
Unlimited archiving capabilities. Data storage should be a database, or something with a lot of database characteristics. Should be able to get full messages (with all headers!), however. Have to be careful here -- those archives can get very large very quickly.
If the architecture is done properly, it should be possible to add all the user interface twiddles that anybody could want (icons for the sender showing their Webcam?), while keeping a solid base system that will handle the mail properly.
It's both control of the passengers (You *will* drop your trousers and paint your arse green!) and control of the drelbs who run the security checkpoints (follow *every* rule *exactly* or you're fired!) Security- related professions are magnets for rule-bound control freaks.
Most of the stuff is ridiculous. "Turn the laptop on and off". Tweezers. Fingernail clippers. Very little about security and a whole lot about "I'm in charge and you're not!"
I kicked this around as a busness idea a while back.
The software is completely straightforward, except for string-matching algorithms. (This just means that we probably couldn't use standard regexp stuff, as the block/pass lists will be quite long.)
First thing it would do on connection is read the filter list(s) from whatever lists the user subscribes to. A user subscribes to lists that match his/her prejudices about what kids should be allowed to see.
The software should certainly be open source, but the filter lists need full time maintainers. Salaries. Offices. Organization. Marketing. This means money. 'Way beyond my organizational abilities.
Fine. So Micro$oft is jerking the hardware makers around. I'm sure this really surprises everybody.
The obvious solution is to use a non-Intel motherboard. Unfortunately, a bit of Googling turned up, basically, nothing.
IBM supposedly had a "reference design" for a PowerPC mobo, but the only implementation I saw for sale was about US$3500 (yikes!). Only 3 PCI slots and no USB.
The reason people use Wintel mobos is that they're cheap and powerful. You can get a pretty nice Intel mobo for US$500 (including processor). What can you get for that in a PowerPC/MIPS/SPARC/HPPA mobo?
Slashdot Mirror
How so? The companies will still be there -- somebody has to translate the XML into something that people can use. They'd just fire some programmers.
Personally, I have my taxes done by a very good tax attorney
Who will type your info into his copy of TurboTax.
If you have somebody who does your taxes the way you like and is willing to stand up to the IRS for you, then that's terrific. My luck has been (ahem) less than good.
By the time I've collected all the data and explained the picky little details of my business to the tax "professional", I might as well have done it myself. "Audit? Yeah, I'll go in with you for $150/hour. No guarnatees, of course."
I agree with you completely about preferential voting, "instant runoff", Australian voting, whatever you want to call it.
HOWEVER, right now the protests against the current crop of fraud-o-trons is tightly focused. This is good; your friendly average election official has the attention span of a hyperactive four-year-old. Tell him two things at once and he won't remember either one.
Experience from many years of various protests -- keep to ONE ISSUE and you'll win. Spread it out (hey, why not eliminate the Electorial College while we're at it?) and you'll lose.
I assure you, the people who manufacture these horrors aren't gonna lose focus.
I think Pearl Jam also has a solid core of fans who will buy the album, reviews or no.
So why mess around with reviewers at all?
Epic is doing this so they can say, when they get back the bashed-open CD players, "Look at how Evil all those reviewers are!" Then they can blame the reviewers for the music "escaping". Since the ones probably putting the music on the P2P nets are the guys who mixed it it the first place, it lets Epic shove responsibility off onto someone else.
Or maybe they're just saying that these albums don't deserve anything better.
The ideal solution would be for the reviewers to not review them at all. See what that does to sales.
The real question is, is it easier to teach geeks to train, or to train trainers the technical stuff?
My experience (four years teaching Cisco courses) is that it's possible to teach geeks to train, but it's almost impossible to teach trainiers enough technical information to answer questions.
If you have a non-geek trainer, you might as well just read the book. You're never going to get anything beyond that.
The main benefit that I see to ant- spam laws is that it would put pressure on ISPs to get rid of the spammers. Spammers aren't going to file lawsuits against ISPs to let them continue an illegal operation.
They aren't going to pack up and physically move to Korea, either.
"The Mythical Man-Month", Brooks. Won't really help much, but you'll have the satisfaction of knowing exactly how your pointy haired managers are screwing things up.
"Design Patterns", Gamma, et al. Without this, you simply won't be able to understand current discussions about programs or programming. This book gives you the philosophy and vocabulary to understand what's going on.
"The Art of Computer Programming", Knuth. What can I say? An absolutely mindboggling treasure trove.
"Software Tools", Kernighan & Plauger. A Golden Oldie. The book is ancient, but the "software tool" concept is still solid.
"The Design of the Unix Operating System", Bach, and "The Design and Implememtation of the 4.4 BSD Operating System", McKusick et al. (These are old. I would hope there is something equivalent for Linux and current BSDs). While abstraction is all well and good, at some point you have to open up the black box and figure out what the machine is actually doing in there.
You need the definitive description of the language you're working with. For C, it's "The C Programming Language", Kernighan & Richie. For C++ it's "The C++ Programming Language", Stroustrup, or, if you're a standards junkie like me, INCITS/ISO/IEC 14882-1998, "ANSI Standard C++".
If you're doing anything connected with the Internet, learn about RFCs. Personally, I credit a large part of the success of the Internet to the free availability of its governing standards. (Other standards are freely available, but not available for free. A paper copy of ISO 14882, for example, is US$175.)
There are all sorts of "domain specific" books. What you need depends on what you're doing. I find "Advanced CORBA Programming with C++", henning & Vinoski, to be priceless, but then, I do CORBA programming in C++.
No.
The only "value" anything has is what somebody is willing to pay for it. If nobody is willing to buy something, it has a value of zero.
"But we paid a lot to develop this!". Tough. Just because you want something to be worth a lot of money doesn't mean it is. Want to buy some WorldCom stock? The proper business term is "fully depreciated".
What is happening is that NaN is carrying something on the books as an "asset" that has no value. Gotta keep the stock price up ....
I've had the same problem. It's totally bogus, of course:
Point this out, and the argument changes to another bogus anti- open source rant.
From what I've seen in various places, he's trying to describe everything in terms of cellular automata.
But a cellular automaton is just the discrete version of a differential equation. Physics has been described in terms of differential equations for 150 years.
So is he getting any new insights? Simplifying the calculations? Or is he just rearranging the the mathematical furniture?
Bytes on a Pentium, of course!
... why we should care? What have future civilizations ever done for us, after all?
A simple trefoil will explain to anybody what's there, as long as there are any vestiges of our current civilization left anywhere. It'll take a *long* time for the trefoil to be forgotten.
We don't worry about anything else that might affect our long-distant ancestors. Why this?
This hoohah about radioactive waste is nothing more than another bogus "reason" to scream "No Nukes!" without having to think. (Keeping nuclear waste where it is now is *ever* so much safer than putting it in the middle of the Nevada desert, after all!) Public attitudes toward nuclear power have been formed more by monster movies than any rational thought process.
IMHO, nuclear power is a Bad Idea, but for boring old economic reasons. No need to get hysterical.
Two reasons. Minor reason: It Is Assumed that a "better algorithm" will use "more resources", whether it does or not. (Defintions of the terms in quotes are optional.)
Major reason: Defense contractors cannot make money by building a system strictly as specified. Defense contractors make boxcar loads of money on change orders. The normal sequence here would be:
Any reasonably sized system will have thousands of these change orders.
So we have a specification problem and a system design problem. Neither is a pure "programming problem".
Software crashes are like airplane crashes -- blame the lowest guy on the totem pole. In air crashes, it's the pilot. In software, it's a coder.
Another point -- require them to keep copies of any hardware needed to play the media, and require them to provide hardware to law enforcement, government archives, or the copyright office at any time. This would last for the lifetime of the copyright. As a hedge against them going out of business, the hardware (and all specifications needed to rebuild it if necessary) should be escrowed somehow.
This is a serious problem already -- according to a friend at the US National Archives, they're already having problems getting players for seven inch open-reel audio tapes.
IIRC, this is exactly what happened when Keith Henson was sued by the Church of Scientology. He spread around a bunch of their "secret scriptures" advocating illegal action and they sued him for breach of trade secrets.
They won.
With no worries about antitrust prosecution, we're going to see a lot more of this stuff. We've already seen them state flat out that "donated" computers must have a legal Microsoft OS and attack the GPL directly; more FUD will surely follow.
The only question is how far their "customers" can be pushed. My guess is pretty far. Never underestimate the pointy-hair factor. Most places, "learning something new" is interpereted as "complete retraining". PHBs regard doing anything new the way a nun would regard going to work in a brothel.
About the only thing we can do is to make sure Open Source solutions don't get wired out due to:
1. Laws or standards that mandate the use of patented/licensed technology. (*Must* use GIF, *must* pony up US$5000 to Unisys.)
2. Laws that specify "maufacturer's liability" (release an Open Source program; get sued if somebody doesn't like it.)
3. Laws mandating DRM hardware/software.
I'm sure we're going to see a flood of these from the Microsoft keiretsu.
We run air over calcium oxide, which absorbs CO2 and turns into calcium carbonate. We then heat the calcium carbonate to turn it back into calcium oxide and CO2. Now we react the CO2 with something to take it out of the actmsphere. Like, say, calcium oxide.
How do we make calcium oxide in the first place? Heat calcium carbonate to drive off the CO2.
Looks to me like the greenhouse gas version of perpetual motion.
is Microsoft Outlook with a bunch of plugins. I rather suspect that they've never used anything except Outlook and Outlook Express for serious work. Their "wish list" is almost all user interface doodles and twiddles.
Problems with Outlook and OE are obvious to /.ers:
I've considered writing my own mail/news client, but it seems like an awful lot of work for relatively little return -- current programs are Almost Good Enough. Anyway, what I'd look for:
If the architecture is done properly, it should be possible to add all the user interface twiddles that anybody could want (icons for the sender showing their Webcam?), while keeping a solid base system that will handle the mail properly.
It's about control.
It's both control of the passengers (You *will* drop your trousers and paint your arse green!) and control of the drelbs who run the security checkpoints (follow *every* rule *exactly* or you're fired!) Security- related professions are magnets for rule-bound control freaks.
Most of the stuff is ridiculous. "Turn the laptop on and off". Tweezers. Fingernail clippers. Very little about security and a whole lot about "I'm in charge and you're not!"
Control freaks at play.
a power higher than yourself
You mean the Government? God is the Government? The Government is God? You need to watch your phrasing; you're not talking to True Believers here.
I live by moral codes derived from social norms, scientific principals, and Government laws. You live by moral codes that "God" whispers in your ear.
You may call it "morals"; I call it schizophrenia.
I kicked this around as a busness idea a while back.
The software is completely straightforward, except for string-matching algorithms. (This just means that we probably couldn't use standard regexp stuff, as the block/pass lists will be quite long.)
First thing it would do on connection is read the filter list(s) from whatever lists the user subscribes to. A user subscribes to lists that match his/her prejudices about what kids should be allowed to see.
The software should certainly be open source, but the filter lists need full time maintainers. Salaries. Offices. Organization. Marketing. This means money. 'Way beyond my organizational abilities.
I beleive in fairies! I can fly!
Pardon me while I go up to the roof with a parasol.
Fine. So Micro$oft is jerking the hardware makers around. I'm sure this really surprises everybody.
The obvious solution is to use a non-Intel motherboard. Unfortunately, a bit of Googling turned up, basically, nothing.
IBM supposedly had a "reference design" for a PowerPC mobo, but the only implementation I saw for sale was about US$3500 (yikes!). Only 3 PCI slots and no USB.
The reason people use Wintel mobos is that they're cheap and powerful. You can get a pretty nice Intel mobo for US$500 (including processor). What can you get for that in a PowerPC/MIPS/SPARC/HPPA mobo?
Yo! Hardware guys! Market niche here!