> If GPL is as bad as Microsoft says it is, > why do they keep drawing attention to it?
Actually, "no press is bad press" ==> If you are unknown, it is better to have attention-getting bad things said about you, than to be ignored as if you are irrelevant.
So MS is goofing by bringing attention to the GPL. UNLESS, the GPL is already well known - this must be the case, otherwise the biggest spindoctor in the computer industry wouldn't make this mistake.
Yeah, I think the GPL has arrived. Now to explain it to my parents...
>It gets it from all the companies that have higher profits >because they aren't paying the Microsoft tax.
Actually, you're making a lot of sense. If my company has $100, it could either keep it as profits, in which case the government gets (say) $30, or it could spend it on Microsoft stuff. Not all of the money that goes to microsoft is taxable, say only 30% (I recently estimated MS has a 29% margin). So the government gets only 30% x $30 = $9.
In other words, if you are in The Land Of Microsoft, where the government gets revenue ONLY from Microsoft corporation, and no other corporation exists, he's right. In the real world, it's the exact opposite of the truth.
> A Java downcast is dynamically checked and cannot compromise the
> integrity of the virtual machine. It is not "unsafe" in any
> meaningful sense of the word.
exactly. You can only cast to a superclass, period. This is useful and unfortunately necessary in a lot of situations, but not at all unsafe. In general, Java is a relatively bullet-proof environment for a program.
People are not going to run their jvm's in 'unsafe' mode.
You can get around it all and wreak havoc with a native piece of code, but you fine tune that stuff, it's hopefully small. And the majority of your app is safe from the usual bugs.
If you get your app rated "100% Pure Java [registered trademark i think]" there can be no JNI in it and it's bulletproof. I've always been impressed with Java and I wish I was using it now.
Given java, I see no reason for C# to exist at all, other than Microsoft's usual hegemony.
In the US, apparently someone patented the use of a laser pointer for cat amusement.
Realistically, what's the upshot? Nobody can sell laser pointers claiming that they are cat amusement devices, except the patent holder, and anybody he/she licenses out to.
Theoretically they could invade your living room and sue you for royalties on their idea that you are using, but pfft, there's no money in it.
My guess is that these stories (the coat hanger, the wheel) have some details missing. Not sure.
> When will the US government realize that allowing patents
> on common ideas is just wrong?
That's like saying "When will programmers realize that bugs are bad, and we don't want more of them?"
Patents have to be "useful, novel and unobvious", or they should be at least. That's the whole idea. There's plenty of effort spent on disallowing me-too patents and stuff that's already been thought of. And, even if they get through, they can be blown away later.
> That's the problem. Say company A creates something and
> company B is jealous. Well, company B could simply say
> "Oh yeah, we came up with that too, and so did company
> C and D and E. It's common!".
Do you really think that nobody's thought of this before? Patent law goes back hundreds of years, and it's all engineers. Nobody's that dumb. And it's all jealous entrepreneurs who try to rip off other people's ideas, and simultaneous inventions who get their foot in the door a week earlier than the other guys. Nothing new.
If company B wants to say We Came Up With That Too, they have to have some evidence, with dates. A shipping product is REALLY good evidence. An ad in a magazine, given the magazine's date, is really good evidence.
Inventors are supposed to keep notebooks (stitched pages, not looseleaf of course) with their ideas, and signed paragraphs from their friends saying I read it and I understood it, with the date. Receipts are good for establishing dates. Anything that's hard to forge. Sending specifications to the patent office, that's really good evidence.
An official of company B who gets on the stand and vehemently asserts that they got the idea first, is NOT good evidence. Kindof like tobacco industry executives saying tobacco is harmless. Someone without a vested interst is better evidence. Paper is better evidence.
>I don't want to spend the time reading it to find out,
>if that's how they write.
just like os api's, there's a lingo, you get used to it, given time.
> And of course, it is quite possible that between 1994 when they filed the patent and 1997, they discovered which way the RDF team was going and added the appropriate definitions to the patent...
You can't do that, not without a time machine. When you file the application, that date's set. Say, 1994. You cannot add new material later. Just fix details and argue with the patent examiner. (don't ask me the difference, i dunno.)
If you want to add material, say in 1997, you start a new patent application with a new, later date. That patent can be blown away because the RTF's work consitutes prior art, dated before 1997, right?
You can't patent forks and spoons, no matter how many twists and turns. You are not the first one to try, and these are engineers who made these rules.
> However, serious inventors and scientists don't make up
> words like "Endo-dynamic information node."
The claims of a patent use words defined in the spec for the patent. The spec goes on for pages and has pictures and must be understandable by you, otherwise it gets thrown out. Must describe to someone "skilled in the art" how to implement it. Otherwise there's no reason for patents.
So everybody reads the spec so everybody's on the "same page". Then everybody reads the claims to see what they think is a new idea that they're claiming.
> The "inventor" is utterly unaware of existing work in the area,...
> He also doesn't know about prior art.
Same thing.
If this is true, the patent is easy to blow away, as long as someone brings the lawsuit to do so. This is why nobody can patent forks and spoons, cuz of the prior art in any dining room. All you have to do is show that somewhere, anywhere in the world, did it or thought it up (and put some effort into it), on a date before the earliest date bragged by the patent owner.
The patent office SHOULD prevent patents from being issued like this, but they are overworked and underpaid, and they just want to score some years and go into private practice. Therefore, existing patents can be broken if the truth comes out later. It's only fair.
>...he is either a crack-pot or his marketing side is a lot stronger
> than his technical side, so I'd wonder about whether he really
> can invent anything.
Unfortunately there's no intelligence or age requirements on patent inventors, any pinhead can do it, if you can get the paperwork past the patent examiner (the overworked one).
> The non-standard terminology is deliberately used so that no one doing a patent search, is likely to find it.
That's what PHBosses say about programmers, and everybody says about lawyers, politicians, etc. But it's lingo. If you do patent searches, you get used to it.
> This lets the patent holder wait until someone has committed their business to using this technology, instead of working around it by changing the implementation, and then spring the patent on them. If that's not fraud, it ought to be.
If it's really obvious, Use the Prior Art, Luke. This stuff can be blown out of the water.
Usually patents are on the invention, the concept. Trivial changes won't get you out of it (as with copyrights), but better ideas might. This one sounds really obvious to me - there's an obvious requuirement too. Ask a patent lawyer.
you... you mean... I won't get laid cuz I wrote a cool OSS program? They lied to me!!
I'll just have to go work for this other outfit I heard of. They promise that after my "mission", I'll have tons of really sexy babes in this place that they send me to. They had a name for it, can't remember, "after here" or something. All I have to learn to do is to fly a plane, and Hamid said it'll be really cool. He wouldn't tell me anything else, though.
> Correct me if I am wrong, but doesn't the Walnut Creek
>FreeBSD CDROM subscription money go in part to the team?
Maybe. Unless, cough, they're having a short year and they gotta pay the rent.
This is one of the problems with all this OSS contribution stuff - when push comes to shove, the voluntary contributions are the last to get paid. And the nasty landlords (who can evict) and the DSL provider (who can cut you off) and the commercial software publishers (whose bugfixes you need) are the first to get paid.
>...To make up the difference, the government will give you a rebate of $1,500....
Couple of details here:
- that applies to US tax law, as well as what I'm saying here. Your tax laws in your country or planet may vary. Will vary.
- Charities go onto Schedule A. There's a big, um, deductible thingee in the way, like the first idunnohowmuch doesn't count. If you own a house and pay mortgage, the interest also goes on Schedule A, so you're probably way past the deductible thingee. If you paid a zillion dollars to doctors for your cancer treatments last year, you might also be past. If you're an average person renting an apartment, your charities might score you Zero on your tax return.
- many US states have state income tax, too, which will add a few (ten?) percent to your tax rate, for the above calculation. On your state's version of Schedule A. With its own inscrutible rules, that might blah blah blah....
- If you have a business, you might be able to call it a biz expense. If not, it's a longshot.
- If you forget until after Jan 1, try postdating the check. Works for me. Give or take, it all comes out in the wash anyway, you deduct this year or next year, whatever, uncle sam gets their bux. UNLESS your tax situation changes A LOT from 2001 to 2002, in which case blah blah blah...
- Consult your accountant. If you don't have one, get one. He or she already knows the stuff you will spend hours trying to learn by yourself, even if you have turbo tax. And, far better than I do or ecampbel or any other tax amateur on slashdot. Like, they never call them "deductible thingees", and they never say "blah blah blah".
>I think a lot of people would be shocked by how corrupt a lot
>of high-profile organizations are, and how small the
>percentage of donations go to the intended receivers.
I was once told, by an unreliable source, that the legal limit was 4% - if less than 4% of contributions actually went to the intended organization, rather than the mailing house, administrators, contribution mailing ad copy writers, envelope lickers, etc, then it didn't qualify as a "charity". This was decided because Boys Town Nebraska was at 4% and they wanted to let them in. (If this doesn't make sense, blame my unreliable source.)
So, yeah, send your bux to XFree86.org or wherever, directly. And vote with your dollars (or euros or whatever).
>If everyone does something in the community, everyone is
>getting something for free.. just as long as everyone does
>something good.
That's the problem - it rarely adds up to a positive number. Human nature.
- some people produce stuff that's bad
- some people forget to produce
- some people are just lazy
- some people will do it for certain... tomorrow
- some people are just selfish and think open sourcers are stupid to give it away
- most people rationalize that they are contributing enough, whether or not they are
I read somewhere that in a "significant other" relationship (eg wife&husband), if each person feels like they are contributing about 60% of the total work+money+whatever, that's "about right", each person is really contributing 50%. That's when the two people LOVE each other. I have never loved a programmer like that, nor a company.
So GNU goes begging while Bill Gates eats caviar in his $100k antique chair. I'm not happy about it, I just see it happening again and again.
> I left out the really fun part about actually installing OSX on a G3. Spent hours dinking around with it. I finally called up Apple tech support about this. According to Apple I had to create an 8gig partition in order for it to work. No more, no less. Nothing in any of the documentation could I find information about this, and I would have thought others would have different partition sizes. Didn't explore it further after that.
Well, that's never happened to me. Honestly, I haven't had much experience installing OS X. Mostly cuz it works the first time, and it's done. What is there to spend hours on? I had one beta version fail to install, and that's it, ever. Worked the second time, that time. What's this 8gb thing? I've never heard of something so absurd. I don't think any of my installations have been in 8gb partitions. how do you get the installer to fail? I don't get it.
> When evaluating OSX for some Mac users I support I ran into serious difficulties in how to make basic changes to the GUI. Dumb things, like the background graphic, system colors, and other stuff along them lines.
Go to The Control Panel. There's only one of them, unlike on Linux, where there's countless control panels, depending on your distro and the whims of the programmers and the block diagram of how it's put together.
> On the other hand, my first experience with KDE (back on 1.12 as I recall) I managed to locate all kinds of tweaks to the UI with mostly all the control center objects being where I expected to find them.
You are a user who knows how it's put together. "ui tweaks" are part of the window manager, session manager, blah blah blah. If you know X windows and the gui stack and how it's put together, you can find all that stuff. I just want to use my computer.
My first experience on Gnome, I was lost, and the only help message told me how to drag a title bar to move a window, duuu. Where's the control panel that adjusts the screen depth? Took me 6 months to find out on Linux. On MacOS and Windows it takes a minute.
only if you exclude hassle cost. You plug stuff in and it works. If your time is worth money, macs are cheaper.
>...you can get hardware that will run it for next-to-nothing...
but if your time is worth money, that's really going to be expensive.
> The article's notation of a proliferation of ibooks at this or that conference is close to meaningless. Those conferences are damn expensive!
People whose time is worth money, buy Macs and go to conferences, because they are well paid, and their companies are well funded. If your computers doubled in price, but your paycheck also doubled at the same time, wouldn't that be a deal?
gimme a pc catalog. OK. The dell you got... ok change that to a compaq. here's one for $1599, 1ghz. 14.1" screen. dvd/cdrw 128mb/20gb. 10/100 ethernet. Free printer. Mid-priced in its class.
Titanium for $2994, has 15.2" megawide screen. 512mb ram/30gb. Firewire port. DVD-ROM, oops. 10/100/1000 ethernet. Free Printer. Sleep works correctly. This is the top of the line mac.
Not the same beast.
OK here's a mac ibook, $1694. 600mhz. 128mb./20gb. 12.1" screen, 10/100 ethernet. dvd-rom/cdrw. firewire. free printer. Sleep works correctly. That's more like the compaq, and only $100 more. We're talking 7% more expensive for a Mac.
I'm having a hard time finding that 4.2ghz laptop. Can't even find a 1.7ghz pentium laptop. hmmm... Keep in mind that 1 powerpc mhz is about 1.5 or 2mHz for pentium. And your cutting edge speeds don't show up in laptops.
>And 5 minutes later your box is trojaned because you left the
>shares world-writeable. That is the problem with Windows:
>By oversimplifying things that are complex by nature,
>it sets up its users for disaster.
The problem is not the oversimplifying. Or, at least, it doesn't have to be.
In this specific case, I think the devil is Microsoft's inability to think in terms of security. [don't get me started... a topic for another megabyte of commentary]
On MacOS, the same barn door is defaulted to CLOSED. (OK, part of that is the fact that the macos TCP/IP stack does not inherently route.) I think the best default is something that's easy to open up to neighboring machines but hard to open up to the internet at large.
Defaults should be set for the MOST NAIVE user to have SOME KIND OF SUCCESS with the MINIMUM DANGER.
MOST NAIVE - the more advanced users know where the config settings are, and can change them to suit. Or they can find them if they don't know where at the start. The naive users, and THOSE EXPERTS who DON'T HAVE THE TIME to mess with it, are stuck with the defaults. So often, in the Unix world, the defaults are set for the convenience of the experts who wrote the software, when the defaults should be for the beginner. Therefore, for instance, stty should set the default backspace-delete character to BOTH del and bs, cuz beginners never know how to change the setting it's at, and often don't even realize what's going on, what the problem is.
SOME KIND OF SUCCESS - without success that is obvious, an inexperienced user will get discouraged, or will stumble onto the right thing but wander away when it's not confirmed. Microsoft gets this right with their network config... well, not "right" but not bad. Relatively easy success. Messages that confirm "you are now serving/customerdata to the network" are good. Systems with two simultaneous malfunctions, with similar symptoms, are a disaster, because even if the beginner stumbles onto one fix, the other malfunction will fool them into thinking they should try something else. (better error handling helps.)
MINIMUM DANGER - the newbie also does not know how to recover from mishaps. This also partly goes for the expert who doesn't want to hassle. On the "rm" comand, for instance, the -i option (ask before deleting) should be ON by default. Those experts who know how, will turn it off after they get sick of answering the same damn question over and over. This is what microsoft messes up all the time with security - the barn door is once again left open, and it's the user's fault if they don't know enough to close it.
>...hardware gets adopted by Linux distros pretty well....
>...a better and better job recognizing and installing hardware
>without me having to go in and recompile or anything.
I once tried to install a Debian distro. Ethernet wasn't working, dunno why. After messing with that, and other hassles, for a day and a half, I was confronted with downloading source for a replacement driver, building it up, redoing the kernel,... I said to myself, if it works this badly in the dealer's lot, what happens when I get it out onto the highway? Far cheaper was to pay $30 and get a Caldera distro that just worked. Cheaper in terms of my time.
> what you did was a bit unnecessary. Since you're running
> RH 6.2, you need to grab the packages that are built for
> 6.2, wu-ftpd-2.6.1-0.6x.21.i386.
in other words, "blah blah blah, blah blah blah, blah blah". Exactly where did you learn the fact that If "you're running RH 6.2, you need to grab the packages that are built for 6.2" ??!?! Does this rule extend to version 7.2? To version 5.2? How do you know that it does? Where did you learn it? My mom doesn't know it. It is not instinctive.
"Oh of course RPM didn't exist before redhat version blah... therefore it's a moot point". Notice how a significant part of the facts needed to understand how to use stuff, is history: this came before that, that's why it's the way it is, then this other guy, really arrogant, he did something else that way... so if you start using Linux in 2001 you will be at a disadvantage cuz of all the things you didn't live through.
These are the details that make the difference. These are the facts that new users don't know, or aren't sure of.
> "setup.exe" is no easier or harder than double-clicking on an RPM
Last time I double clicked on an RPM, well, first, it started doing something on the first click, not the second, what browser was I using?
Then some program came up, there was a window with lots of doodads around. We looked at each other and said "what next" (me and that program). Maybe there was a button that I thought meant "install". I cliked it, something happened. Did it install? dunno. Try "installing" it again, see if it does something different, like says it's already installed. Clicked again. Something happened again, no difference. One way or another, my software was not installed. I think. No way to tell.
I generally install RPMs on the command line. (I'm a mac guy, mind you.) You read the man page for rpm and it goes on and on about the theory of the RPM system and all of the commands to build, update and make sure these options are all mutually exclusive except you can combine the -X option with -T if you haven't specified a default registration directory in the system-wide config file blah blah blah blah.
Scanning down the man page, you find out that the command to actually INSTALL one of these rpm files (think: 99% of the use rpm will have is a simple install) you just say:
install -i filename.rpm
So I memorized that. That's why I use the command line in Linux: stuff works. It's the user interface that the programmers themselves use.
why didn't they put that at the beginning of the man page? why didn't they make that the default? You unzip a zip file like this:
unzip filename.zip
PS: zip was made by pc people. yes i'm trolling, sorry i'll stop.
You untar a tar file like this:
tar -xvf filename.tar
what's with the xvf? Oh by the way you don't need the minus if it's the first option and you can pipe the input from stdout by including the blah blah blah options and bada bada bada bada....
You ungzip a gzipped file and it actually EATS and REPLACES THE PREEXISTING FILE. Holy disappearing data batman! what if something goes wrong?
> People do make it out to be harder than it really is.
No.
It's that we linux users have lost track of all of the facts and details we've learned over the years and months that we use all the time to make it work. So it SEEMS easy to us. It's just like knowing or not knowing a foreign language. It seems so easy for the ones who know, but for learners, all it takes is one detail that you don't know and you're paralyzed.
> It seems that the difficult part is being able to empathize
> with whomever you are attempting to help. In my experience,
> the ability to truly visualize things from another person's
> mindset is rather rare.
YEAH! This ability, to put yourself into the mind of the other person, is important for:
- doing tech support
- writing documentation
- writing user interfaces
- teaching
- social interactions in general
If a programmer is bad at it, they are going to make bad user interfaces. Knowing how to fling the GUI API does not make good user interfaces.
It is mental work, it does require thinking, but unfortunately this thinking is not perceived among technical people as being very merit-worthy. The programmers with the fame are the ones who hack the kernel or fix the driver, not the ones who adapt the kernel to a human being. The academics with fame are the ones who figure out relativity or DNA, not the ones who teach it to new students, even though the mental effort is similar in magnitude. It just doesn't seem "macho" enough. (Handful of exceptions: Einstein, Feynmann, RM Stallman and Don Knuth were/are excellent at teaching/writing, they do both sides well.)
See article in the recent Wired Magazine about asperger's syndrome. Quick sumary: Asperger's syndrome is like mild Autism. These people have trouble visualizing what's going on in other people's brains, among other symptoms. And it's rampant in Silicon Valley.
Yeah, I remember the first time I got Linux running on my laptop, I was desperate to change the display resolution. My screen was 1024x768, but the installer let me set it higher, what's up with that? So I did. Well, I got a "stepping out" style screen which I hated, gotta drag to the edge to see the pixels beyond the edge. Battling seasickness, I struggled to find the "monitors control panel" or the "display control panel" in vain. IN a moment of reflection, my mouse rested on a titlebar, and a tooltip came up, "To move this window around, drag it around in the title bar" or something. DUUU!!!
observation: the tooltip help items are generally worthless on Linux. Any time your tooltip say something the user already knows, that's worthless. Another example: see item that says "foobar". The tooltip says "foobar". Exact same spelling. Zero new information. Doesn't say WHAT foobar is or what it does or why the heck you would use it. People who write software like this deserve a pie in the face.
I didn't know at the time that there isn't one "control panel" directory, or any central place, there's like a half dozen different "control panel" programs and configuration programs, they all work different. They are not intuitively organized, they are organized according to the subsystems which you have yet to learn, and the programmers who worked on each. EG hardware like mice are not configured in a section called "hardware" or a control panel named "GUI settings", it's in XFree86Config (of course!) You can search the entire file system for the word "mouse" and find tons of stuff, none of which tell you to go to this config program.
I ended up reinstalling the whole thing. The only control panel I ever saw that controlled the screen resolution was in the installer, so it was easier to redo that than to keep spinning my wheels.
I reinstalled it about two dozen times. I was also trying to multiboot NT and Win98, so that added to the complexity, but I was determined to learn it. Being able to blow away the hard disk allows lots of educational experiments.
By contrast, MacOS and Windows pretty much install the first time. Their control panels are all in one place so after 15 seconds you've found what you need. That's why bill gates is laughing all the way to the bank.
I think that Linux can be made "easy for newbies" in a sortof abstract sense. If you take someone and keep them in the watered-down gui tools that are there, you can think that they have everything they need. If you try to live like that, yourself, however, you realize the folly of that.
The problem is that stuff goes wrong. If you don't have the skills to fix it, your computer just turned into a big paperweight. And really the only way to be able to fix stuff in Linux is:
1) learn the command line (big time committment)
2) read the docs (big time committment)
MacOS and to a lesser extent Windows are designed by people who actually live in the gui tools that they build for themselves, and they make them work, on a power user level. I sit there in macos, grab a hundred source files, throw them at the editor and a hundred windows open, boom boom boom. Then I do global search and replaces on them, in unison. (and macos was doing this in the late 80's.) Try that with vi.
When things go wrong, I fix it with gui tools - Resourcerer, the finder, and just regular control panels and other facilities that (surprise!) have error messages that actually say what they mean. Stuff just works better in general, so you don't have to go around fixing stuff as often.
Example: install Linux. Ethernet isn't up. The ethernet card IS on the list of approved cards... after you go websurfing for hours learning stuff you never knew before about which ethernet boards use which chips and emulate which other chips and how well they emulate them. OK now spend more hours struggling to find the reason why it isn't working. Try downloading another driver. Try rebuilding the kernel. Read the random error messages that squirt out of each of these processes, in each case, make a guess as to whether the error message is (a) important (b) unimportant (c) correct (d) incorrect. When you get to the end and it still doesn't work, reevaluate your decision on each of those error messages to see if maybe you should have paid attention anyway.
Example: get digital camera for older Mac. "The usb link with this camera will only work on a usb that comes native with your mac, not on an add-on usb adapter". I have an add-on usb adapter. I try it anyway. It works. It works the first time, in fact. No error messages, no questions as to whether I'm doing this right or that right, obviously it's right, or right enough.
Slashdot Mirror
> If GPL is as bad as Microsoft says it is,
> why do they keep drawing attention to it?
Actually, "no press is bad press" ==> If you are unknown, it is better to have attention-getting bad things said about you, than to be ignored as if you are irrelevant.
So MS is goofing by bringing attention to the GPL. UNLESS, the GPL is already well known - this must be the case, otherwise the biggest spindoctor in the computer industry wouldn't make this mistake.
Yeah, I think the GPL has arrived. Now to explain it to my parents...
>It gets it from all the companies that have higher profits
>because they aren't paying the Microsoft tax.
Actually, you're making a lot of sense. If my company has $100, it could either keep it as profits, in which case the government gets (say) $30, or it could spend it on Microsoft stuff. Not all of the money that goes to microsoft is taxable, say only 30% (I recently estimated MS has a 29% margin). So the government gets only 30% x $30 = $9.
In other words, if you are in The Land Of Microsoft, where the government gets revenue ONLY from Microsoft corporation, and no other corporation exists, he's right. In the real world, it's the exact opposite of the truth.
So who's getting fooled by this hogwash?
> A Java downcast is dynamically checked and cannot compromise the
> integrity of the virtual machine. It is not "unsafe" in any
> meaningful sense of the word.
exactly. You can only cast to a superclass, period. This is useful and unfortunately necessary in a lot of situations, but not at all unsafe. In general, Java is a relatively bullet-proof environment for a program.
People are not going to run their jvm's in 'unsafe' mode.
You can get around it all and wreak havoc with a native piece of code, but you fine tune that stuff, it's hopefully small. And the majority of your app is safe from the usual bugs.
If you get your app rated "100% Pure Java [registered trademark i think]" there can be no JNI in it and it's bulletproof. I've always been impressed with Java and I wish I was using it now.
Given java, I see no reason for C# to exist at all, other than Microsoft's usual hegemony.
In the US, apparently someone patented the use of a laser pointer for cat amusement.
Realistically, what's the upshot? Nobody can sell laser pointers claiming that they are cat amusement devices, except the patent holder, and anybody he/she licenses out to.
Theoretically they could invade your living room and sue you for royalties on their idea that you are using, but pfft, there's no money in it.
My guess is that these stories (the coat hanger, the wheel) have some details missing. Not sure.
> When will the US government realize that allowing patents
> on common ideas is just wrong?
That's like saying "When will programmers realize that bugs are bad, and we don't want more of them?"
Patents have to be "useful, novel and unobvious", or they should be at least. That's the whole idea. There's plenty of effort spent on disallowing me-too patents and stuff that's already been thought of. And, even if they get through, they can be blown away later.
> That's the problem. Say company A creates something and
> company B is jealous. Well, company B could simply say
> "Oh yeah, we came up with that too, and so did company
> C and D and E. It's common!".
Do you really think that nobody's thought of this before? Patent law goes back hundreds of years, and it's all engineers. Nobody's that dumb. And it's all jealous entrepreneurs who try to rip off other people's ideas, and simultaneous inventions who get their foot in the door a week earlier than the other guys. Nothing new.
If company B wants to say We Came Up With That Too, they have to have some evidence, with dates. A shipping product is REALLY good evidence. An ad in a magazine, given the magazine's date, is really good evidence.
Inventors are supposed to keep notebooks (stitched pages, not looseleaf of course) with their ideas, and signed paragraphs from their friends saying I read it and I understood it, with the date. Receipts are good for establishing dates. Anything that's hard to forge. Sending specifications to the patent office, that's really good evidence.
An official of company B who gets on the stand and vehemently asserts that they got the idea first, is NOT good evidence. Kindof like tobacco industry executives saying tobacco is harmless. Someone without a vested interst is better evidence. Paper is better evidence.
See Nolo Press: Patent It Yourself.
>I don't want to spend the time reading it to find out,
...
...he is either a crack-pot or his marketing side is a lot stronger
>if that's how they write.
just like os api's, there's a lingo, you get used to it, given time.
> And of course, it is quite possible that between 1994 when they filed the patent and 1997, they discovered which way the RDF team was going and added the appropriate definitions to the patent...
You can't do that, not without a time machine. When you file the application, that date's set. Say, 1994. You cannot add new material later. Just fix details and argue with the patent examiner. (don't ask me the difference, i dunno.)
If you want to add material, say in 1997, you start a new patent application with a new, later date. That patent can be blown away because the RTF's work consitutes prior art, dated before 1997, right?
You can't patent forks and spoons, no matter how many twists and turns. You are not the first one to try, and these are engineers who made these rules.
> However, serious inventors and scientists don't make up
> words like "Endo-dynamic information node."
The claims of a patent use words defined in the spec for the patent. The spec goes on for pages and has pictures and must be understandable by you, otherwise it gets thrown out. Must describe to someone "skilled in the art" how to implement it. Otherwise there's no reason for patents.
So everybody reads the spec so everybody's on the "same page". Then everybody reads the claims to see what they think is a new idea that they're claiming.
> The "inventor" is utterly unaware of existing work in the area,
> He also doesn't know about prior art.
Same thing.
If this is true, the patent is easy to blow away, as long as someone brings the lawsuit to do so. This is why nobody can patent forks and spoons, cuz of the prior art in any dining room. All you have to do is show that somewhere, anywhere in the world, did it or thought it up (and put some effort into it), on a date before the earliest date bragged by the patent owner.
The patent office SHOULD prevent patents from being issued like this, but they are overworked and underpaid, and they just want to score some years and go into private practice. Therefore, existing patents can be broken if the truth comes out later. It's only fair.
>
> than his technical side, so I'd wonder about whether he really
> can invent anything.
Unfortunately there's no intelligence or age requirements on patent inventors, any pinhead can do it, if you can get the paperwork past the patent examiner (the overworked one).
> The non-standard terminology is deliberately used so that no one doing a patent search, is likely to find it.
That's what PHBosses say about programmers, and everybody says about lawyers, politicians, etc. But it's lingo. If you do patent searches, you get used to it.
> This lets the patent holder wait until someone has committed their business to using this technology, instead of working around it by changing the implementation, and then spring the patent on them. If that's not fraud, it ought to be.
If it's really obvious, Use the Prior Art, Luke. This stuff can be blown out of the water.
Usually patents are on the invention, the concept. Trivial changes won't get you out of it (as with copyrights), but better ideas might. This one sounds really obvious to me - there's an obvious requuirement too. Ask a patent lawyer.
you ... you mean... I won't get laid cuz I wrote a cool OSS program? They lied to me!!
I'll just have to go work for this other outfit I heard of. They promise that after my "mission", I'll have tons of really sexy babes in this place that they send me to. They had a name for it, can't remember, "after here" or something. All I have to learn to do is to fly a plane, and Hamid said it'll be really cool. He wouldn't tell me anything else, though.
Damn, that website's not up anymore.
> Correct me if I am wrong, but doesn't the Walnut Creek
>FreeBSD CDROM subscription money go in part to the team?
Maybe. Unless, cough, they're having a short year and they gotta pay the rent.
This is one of the problems with all this OSS contribution stuff - when push comes to shove, the voluntary contributions are the last to get paid. And the nasty landlords (who can evict) and the DSL provider (who can cut you off) and the commercial software publishers (whose bugfixes you need) are the first to get paid.
Capitalism: you can run, but you can't hide.
> ...To make up the difference, the government will give you a rebate of $1,500....
Couple of details here:
- that applies to US tax law, as well as what I'm saying here. Your tax laws in your country or planet may vary. Will vary.
- Charities go onto Schedule A. There's a big, um, deductible thingee in the way, like the first idunnohowmuch doesn't count. If you own a house and pay mortgage, the interest also goes on Schedule A, so you're probably way past the deductible thingee. If you paid a zillion dollars to doctors for your cancer treatments last year, you might also be past. If you're an average person renting an apartment, your charities might score you Zero on your tax return.
- many US states have state income tax, too, which will add a few (ten?) percent to your tax rate, for the above calculation. On your state's version of Schedule A. With its own inscrutible rules, that might blah blah blah....
- If you have a business, you might be able to call it a biz expense. If not, it's a longshot.
- If you forget until after Jan 1, try postdating the check. Works for me. Give or take, it all comes out in the wash anyway, you deduct this year or next year, whatever, uncle sam gets their bux. UNLESS your tax situation changes A LOT from 2001 to 2002, in which case blah blah blah...
- Consult your accountant. If you don't have one, get one. He or she already knows the stuff you will spend hours trying to learn by yourself, even if you have turbo tax. And, far better than I do or ecampbel or any other tax amateur on slashdot. Like, they never call them "deductible thingees", and they never say "blah blah blah".
>I think a lot of people would be shocked by how corrupt a lot
>of high-profile organizations are, and how small the
>percentage of donations go to the intended receivers.
I was once told, by an unreliable source, that the legal limit was 4% - if less than 4% of contributions actually went to the intended organization, rather than the mailing house, administrators, contribution mailing ad copy writers, envelope lickers, etc, then it didn't qualify as a "charity". This was decided because Boys Town Nebraska was at 4% and they wanted to let them in. (If this doesn't make sense, blame my unreliable source.)
So, yeah, send your bux to XFree86.org or wherever, directly. And vote with your dollars (or euros or whatever).
>If everyone does something in the community, everyone is
... tomorrow
>getting something for free.. just as long as everyone does
>something good.
That's the problem - it rarely adds up to a positive number. Human nature.
- some people produce stuff that's bad
- some people forget to produce
- some people are just lazy
- some people will do it for certain
- some people are just selfish and think open sourcers are stupid to give it away
- most people rationalize that they are contributing enough, whether or not they are
I read somewhere that in a "significant other" relationship (eg wife&husband), if each person feels like they are contributing about 60% of the total work+money+whatever, that's "about right", each person is really contributing 50%. That's when the two people LOVE each other. I have never loved a programmer like that, nor a company.
So GNU goes begging while Bill Gates eats caviar in his $100k antique chair. I'm not happy about it, I just see it happening again and again.
> I left out the really fun part about actually installing OSX on a G3. Spent hours dinking around with it. I finally called up Apple tech support about this. According to Apple I had to create an 8gig partition in order for it to work. No more, no less. Nothing in any of the documentation could I find information about this, and I would have thought others would have different partition sizes. Didn't explore it further after that.
Well, that's never happened to me. Honestly, I haven't had much experience installing OS X. Mostly cuz it works the first time, and it's done. What is there to spend hours on? I had one beta version fail to install, and that's it, ever. Worked the second time, that time. What's this 8gb thing? I've never heard of something so absurd. I don't think any of my installations have been in 8gb partitions. how do you get the installer to fail? I don't get it.
> When evaluating OSX for some Mac users I support I ran into serious difficulties in how to make basic changes to the GUI. Dumb things, like the background graphic, system colors, and other stuff along them lines.
Go to The Control Panel. There's only one of them, unlike on Linux, where there's countless control panels, depending on your distro and the whims of the programmers and the block diagram of how it's put together.
> On the other hand, my first experience with KDE (back on 1.12 as I recall) I managed to locate all kinds of tweaks to the UI with mostly all the control center objects being where I expected to find them.
You are a user who knows how it's put together. "ui tweaks" are part of the window manager, session manager, blah blah blah. If you know X windows and the gui stack and how it's put together, you can find all that stuff. I just want to use my computer.
My first experience on Gnome, I was lost, and the only help message told me how to drag a title bar to move a window, duuu. Where's the control panel that adjusts the screen depth? Took me 6 months to find out on Linux. On MacOS and Windows it takes a minute.
> Linux not only costs less...
...you can get hardware that will run it for next-to-nothing...
only if you exclude hassle cost. You plug stuff in and it works. If your time is worth money, macs are cheaper.
>
but if your time is worth money, that's really going to be expensive.
> The article's notation of a proliferation of ibooks at this or that conference is close to meaningless. Those conferences are damn expensive!
People whose time is worth money, buy Macs and go to conferences, because they are well paid, and their companies are well funded. If your computers doubled in price, but your paycheck also doubled at the same time, wouldn't that be a deal?
gimme a pc catalog. OK. The dell you got ... ok change that to a compaq. here's one for $1599, 1ghz. 14.1" screen. dvd/cdrw 128mb/20gb. 10/100 ethernet. Free printer. Mid-priced in its class.
Titanium for $2994, has 15.2" megawide screen. 512mb ram/30gb. Firewire port. DVD-ROM, oops. 10/100/1000 ethernet. Free Printer. Sleep works correctly. This is the top of the line mac.
Not the same beast.
OK here's a mac ibook, $1694. 600mhz. 128mb./20gb. 12.1" screen, 10/100 ethernet. dvd-rom/cdrw. firewire. free printer. Sleep works correctly. That's more like the compaq, and only $100 more. We're talking 7% more expensive for a Mac.
I'm having a hard time finding that 4.2ghz laptop. Can't even find a 1.7ghz pentium laptop. hmmm... Keep in mind that 1 powerpc mhz is about 1.5 or 2mHz for pentium. And your cutting edge speeds don't show up in laptops.
> having the software I am running and rely on constantly
> changing does not help me get my job done
me too. Takes me a long time to get a setup that's not broken. Then I don't mess with it.
Currently typing this on a 1998 machine with a 1999 OS version. With vid monitors from 1991 and a tape drive from 1992.
The browser iS recent: iCab cuz Netscape and Mozilla are both broken. iCab 2.6 cuz 2.5.3 had a bug. I'm sticking with 2.6.
>And 5 minutes later your box is trojaned because you left the
/customerdata to the network" are good. Systems with two simultaneous malfunctions, with similar symptoms, are a disaster, because even if the beginner stumbles onto one fix, the other malfunction will fool them into thinking they should try something else. (better error handling helps.)
>shares world-writeable. That is the problem with Windows:
>By oversimplifying things that are complex by nature,
>it sets up its users for disaster.
The problem is not the oversimplifying. Or, at least, it doesn't have to be.
In this specific case, I think the devil is Microsoft's inability to think in terms of security. [don't get me started... a topic for another megabyte of commentary]
On MacOS, the same barn door is defaulted to CLOSED. (OK, part of that is the fact that the macos TCP/IP stack does not inherently route.) I think the best default is something that's easy to open up to neighboring machines but hard to open up to the internet at large.
Defaults should be set for the MOST NAIVE user to have SOME KIND OF SUCCESS with the MINIMUM DANGER.
MOST NAIVE - the more advanced users know where the config settings are, and can change them to suit. Or they can find them if they don't know where at the start. The naive users, and THOSE EXPERTS who DON'T HAVE THE TIME to mess with it, are stuck with the defaults. So often, in the Unix world, the defaults are set for the convenience of the experts who wrote the software, when the defaults should be for the beginner. Therefore, for instance, stty should set the default backspace-delete character to BOTH del and bs, cuz beginners never know how to change the setting it's at, and often don't even realize what's going on, what the problem is.
SOME KIND OF SUCCESS - without success that is obvious, an inexperienced user will get discouraged, or will stumble onto the right thing but wander away when it's not confirmed. Microsoft gets this right with their network config... well, not "right" but not bad. Relatively easy success. Messages that confirm "you are now serving
MINIMUM DANGER - the newbie also does not know how to recover from mishaps. This also partly goes for the expert who doesn't want to hassle. On the "rm" comand, for instance, the -i option (ask before deleting) should be ON by default. Those experts who know how, will turn it off after they get sick of answering the same damn question over and over. This is what microsoft messes up all the time with security - the barn door is once again left open, and it's the user's fault if they don't know enough to close it.
> ...hardware gets adopted by Linux distros pretty well. ...
... I said to myself, if it works this badly in the dealer's lot, what happens when I get it out onto the highway? Far cheaper was to pay $30 and get a Caldera distro that just worked. Cheaper in terms of my time.
>...a better and better job recognizing and installing hardware
>without me having to go in and recompile or anything.
I once tried to install a Debian distro. Ethernet wasn't working, dunno why. After messing with that, and other hassles, for a day and a half, I was confronted with downloading source for a replacement driver, building it up, redoing the kernel,
> what you did was a bit unnecessary. Since you're running
> RH 6.2, you need to grab the packages that are built for
> 6.2, wu-ftpd-2.6.1-0.6x.21.i386.
in other words, "blah blah blah, blah blah blah, blah blah". Exactly where did you learn the fact that If "you're running RH 6.2, you need to grab the packages that are built for 6.2" ??!?! Does this rule extend to version 7.2? To version 5.2? How do you know that it does? Where did you learn it? My mom doesn't know it. It is not instinctive.
"Oh of course RPM didn't exist before redhat version blah... therefore it's a moot point". Notice how a significant part of the facts needed to understand how to use stuff, is history: this came before that, that's why it's the way it is, then this other guy, really arrogant, he did something else that way... so if you start using Linux in 2001 you will be at a disadvantage cuz of all the things you didn't live through.
These are the details that make the difference. These are the facts that new users don't know, or aren't sure of.
> "setup.exe" is no easier or harder than double-clicking on an RPM
....
Last time I double clicked on an RPM, well, first, it started doing something on the first click, not the second, what browser was I using?
Then some program came up, there was a window with lots of doodads around. We looked at each other and said "what next" (me and that program). Maybe there was a button that I thought meant "install". I cliked it, something happened. Did it install? dunno. Try "installing" it again, see if it does something different, like says it's already installed. Clicked again. Something happened again, no difference. One way or another, my software was not installed. I think. No way to tell.
I generally install RPMs on the command line. (I'm a mac guy, mind you.) You read the man page for rpm and it goes on and on about the theory of the RPM system and all of the commands to build, update and make sure these options are all mutually exclusive except you can combine the -X option with -T if you haven't specified a default registration directory in the system-wide config file blah blah blah blah.
Scanning down the man page, you find out that the command to actually INSTALL one of these rpm files (think: 99% of the use rpm will have is a simple install) you just say:
install -i filename.rpm
So I memorized that. That's why I use the command line in Linux: stuff works. It's the user interface that the programmers themselves use.
why didn't they put that at the beginning of the man page? why didn't they make that the default? You unzip a zip file like this:
unzip filename.zip
PS: zip was made by pc people. yes i'm trolling, sorry i'll stop.
You untar a tar file like this:
tar -xvf filename.tar
what's with the xvf? Oh by the way you don't need the minus if it's the first option and you can pipe the input from stdout by including the blah blah blah options and bada bada bada bada
You ungzip a gzipped file and it actually EATS and REPLACES THE PREEXISTING FILE. Holy disappearing data batman! what if something goes wrong?
> People do make it out to be harder than it really is.
No.
It's that we linux users have lost track of all of the facts and details we've learned over the years and months that we use all the time to make it work. So it SEEMS easy to us. It's just like knowing or not knowing a foreign language. It seems so easy for the ones who know, but for learners, all it takes is one detail that you don't know and you're paralyzed.
> It seems that the difficult part is being able to empathize
> with whomever you are attempting to help. In my experience,
> the ability to truly visualize things from another person's
> mindset is rather rare.
YEAH! This ability, to put yourself into the mind of the other person, is important for:
- doing tech support
- writing documentation
- writing user interfaces
- teaching
- social interactions in general
If a programmer is bad at it, they are going to make bad user interfaces. Knowing how to fling the GUI API does not make good user interfaces.
It is mental work, it does require thinking, but unfortunately this thinking is not perceived among technical people as being very merit-worthy. The programmers with the fame are the ones who hack the kernel or fix the driver, not the ones who adapt the kernel to a human being. The academics with fame are the ones who figure out relativity or DNA, not the ones who teach it to new students, even though the mental effort is similar in magnitude. It just doesn't seem "macho" enough. (Handful of exceptions: Einstein, Feynmann, RM Stallman and Don Knuth were/are excellent at teaching/writing, they do both sides well.)
See article in the recent Wired Magazine about asperger's syndrome. Quick sumary: Asperger's syndrome is like mild Autism. These people have trouble visualizing what's going on in other people's brains, among other symptoms. And it's rampant in Silicon Valley.
Yeah, I remember the first time I got Linux running on my laptop, I was desperate to change the display resolution. My screen was 1024x768, but the installer let me set it higher, what's up with that? So I did. Well, I got a "stepping out" style screen which I hated, gotta drag to the edge to see the pixels beyond the edge. Battling seasickness, I struggled to find the "monitors control panel" or the "display control panel" in vain. IN a moment of reflection, my mouse rested on a titlebar, and a tooltip came up, "To move this window around, drag it around in the title bar" or something. DUUU!!!
observation: the tooltip help items are generally worthless on Linux. Any time your tooltip say something the user already knows, that's worthless. Another example: see item that says "foobar". The tooltip says "foobar". Exact same spelling. Zero new information. Doesn't say WHAT foobar is or what it does or why the heck you would use it. People who write software like this deserve a pie in the face.
I didn't know at the time that there isn't one "control panel" directory, or any central place, there's like a half dozen different "control panel" programs and configuration programs, they all work different. They are not intuitively organized, they are organized according to the subsystems which you have yet to learn, and the programmers who worked on each. EG hardware like mice are not configured in a section called "hardware" or a control panel named "GUI settings", it's in XFree86Config (of course!) You can search the entire file system for the word "mouse" and find tons of stuff, none of which tell you to go to this config program.
I ended up reinstalling the whole thing. The only control panel I ever saw that controlled the screen resolution was in the installer, so it was easier to redo that than to keep spinning my wheels.
I reinstalled it about two dozen times. I was also trying to multiboot NT and Win98, so that added to the complexity, but I was determined to learn it. Being able to blow away the hard disk allows lots of educational experiments.
By contrast, MacOS and Windows pretty much install the first time. Their control panels are all in one place so after 15 seconds you've found what you need. That's why bill gates is laughing all the way to the bank.
I think that Linux can be made "easy for newbies" in a sortof abstract sense. If you take someone and keep them in the watered-down gui tools that are there, you can think that they have everything they need. If you try to live like that, yourself, however, you realize the folly of that.
The problem is that stuff goes wrong. If you don't have the skills to fix it, your computer just turned into a big paperweight. And really the only way to be able to fix stuff in Linux is:
1) learn the command line (big time committment)
2) read the docs (big time committment)
MacOS and to a lesser extent Windows are designed by people who actually live in the gui tools that they build for themselves, and they make them work, on a power user level. I sit there in macos, grab a hundred source files, throw them at the editor and a hundred windows open, boom boom boom. Then I do global search and replaces on them, in unison. (and macos was doing this in the late 80's.) Try that with vi.
When things go wrong, I fix it with gui tools - Resourcerer, the finder, and just regular control panels and other facilities that (surprise!) have error messages that actually say what they mean. Stuff just works better in general, so you don't have to go around fixing stuff as often.
Example: install Linux. Ethernet isn't up. The ethernet card IS on the list of approved cards... after you go websurfing for hours learning stuff you never knew before about which ethernet boards use which chips and emulate which other chips and how well they emulate them. OK now spend more hours struggling to find the reason why it isn't working. Try downloading another driver. Try rebuilding the kernel. Read the random error messages that squirt out of each of these processes, in each case, make a guess as to whether the error message is (a) important (b) unimportant (c) correct (d) incorrect. When you get to the end and it still doesn't work, reevaluate your decision on each of those error messages to see if maybe you should have paid attention anyway.
Example: get digital camera for older Mac. "The usb link with this camera will only work on a usb that comes native with your mac, not on an add-on usb adapter". I have an add-on usb adapter. I try it anyway. It works. It works the first time, in fact. No error messages, no questions as to whether I'm doing this right or that right, obviously it's right, or right enough.
this is right. I had a Z80A in my homebrew computer (think plywood) and did some Sinclair ZX-81 work.