I'm genuinely suprised that there hasn't been more inroads into trust chain architecture. The current web works because it requires people to get the semantics - and people are good at working out whether or not they trust a source.
There is a logical seperation between authors. With the semantic web, everything is presented together. How do you know if you can trust it? How do you know that all authors are presenting data the way it is?
It depends what you define an application to be. Something that is loaded by an OS? Well if this is an OS, and it lets you load linux, then by that definition it's an application.
Maybe a better question - Is DOSEMU an application? Are any emulators applications? Is Wine an application (No I know it's not an emulator)?
How do you define an operating system? It's all code. Linux can also be run as a BIOS - does that mean all bios's are operating systems? (Just not very good at it)
Old mate, you said one thing, then when it was argued, you used some physics calculator to argue a different thing. Maybe you should consider a carreer in politics.
you originally said:
Because a billion tons of gravel travelling at 25,000 miles per hour is just as deadly as a billion ton chunk of rock travelling at 25,000 miles per hour. It's not the rock itself that's the problem. It's the kinetic energy from the object's mass that's the problem. Gravel - rock - it's all the same at 25,000 miles per hour...
So i said:
A Billion ton chunk of rock has enough shielding/insulation to not burn up. It would cause a crapload more damage.
And then you give me a crapload of calculations for
so you blow it up in the middle and you get TWO big chunks of say 15 billion tons each and a billion or so tons of gravel and million ton objects.
Holy crap you're a tripper. All I'm saying is that one big rock would cause more damage than gravel. And, just like a politician, you've argued some different point.
In the end you try and get back to some sort of resemblance of a point - good on you. Yeah the kenetic energy would be converted to heat. No it wouldn't be a nice thing. But i'm not arguing if it would be a nice thing of not. I'm just saying that one chunk of rock would cause more damage than the same mass of gravel.
If you want some physics involved, Take one of the most pure physics laws we know. Newtons law. Force = Mass x Acceleration. If the mass of the object is less, then the force is less if the acceleration remains constant. I don't need pages to show you that gravel has less force than a big rock.
Now can you quit with your whining - you're giving nerds a bad name.
Gravel at 25,000 miles an hour would burn up in the earth's atmosphere.
A Billion ton chunk of rock has enough shielding/insulation to not burn up. It would cause a crapload more damage.
How about we try an experiment. We put a tonne of sand above my house in a geosyncronous orbit. And we put a rock that weighs a tonne in a geosyncronous orbit above your house. Then we put them both straight down.
XUL is a great concept, but it will be a while before it's ready to develop decent apps in.
At the moment, there are bugs preventing things as simple as an itunes style datasheet - where you can click in a field and an editbox appears for editing. Even the nearest workaround has a bug attached to it.
XUL's a great theory, but if it were as great as the hype machine says it is, we'd be seeing a crapload more XUL apps by now.
The standard home users will abandon the internet, and the.com boom will shrivel more and more until computers will be reserved for direct communication and business purposes.
No way dude. If worst came to the worst, people will just use trusted VPNS.
What a lot of people don't realise, is the internet is alive. Yeah it's a symbiotic relationship (to humans), but it now is evolving at a faster rate than people need it to. It's evolution is governed by our collaboration but at the same time it's our collaboration medium.
So we're using it to collaborate, and it improves by us collaborating.
Okay you got me. A computer isn't a totally closed environment.
But anything that happens in software, can be reproduced in software. And everything that happens in a computer, is software. Interrupts are a slightly different story, but in modern operating systems they are virtualised enough to make the distinction irrelivant.
Give me a situation that cannot be reproduced due to physical limitations of computers. If it happens once, you should be able to reproduce it. You CAN reproduce it, but current tools don't let you.
I have myself coded quite large multi-threaded systems. The one in particular i'm thinking of was basically an automated testing system, which commonly (depending on the testing scenario) spawned 150 threads, each with it's own RTP (Real time protocol) client. It had a bug at one stage which involved sometimes (rarely) clearing a read mutex after already reading the data associated with it. Because the data had already been sent, but the read mutex was clear, the thread would deadlock - the data had arrived but it had no way of notifying anyone. Due to the multi-threaded nature of this application, most of the time it would be okay. But when context switches were at the wrong times, deadlock!
Yeah, this seems like random behaviour. But it's not. If the operating system's scheduling algorithm was simpler, say if it only context switched after the mutex was clear, it would be sweet.
We (programmers) have total control over our computers, yet we continue to debug in an environment that isn't controlled. If the bug is context switch/timing related, why can't we just run it a few times until the bug occurs, and then replay it with the same timing of context switches etc.?
In context of a buffer overflow bug, why can't we just store what's in memory to a file, in such a way that we can see what structures are where, and then inspect it at will? Yeah i've used memory inspectors - they could be several orders of magnitude more user friendly.
That's what I mean. It's computationally possible, but we don't do it. We are still in the dark ages of computer programming.
While I wholeheartedly agree to the spirit of your comments, I think they need addressing.
1) Yes it's a hack. Yes it's better to actually find the cause of the problem. I hate hacks as much as the next guy - but to the user it would solve all these "reflow bugs" easily and simply. There is nothing to stop them being fixed in the background and for this hack to eventually be taken out. But to a user, if a page looks screwed up, and it doesn't in internet explorer, it's firefox's fault.
Make it a preference that defaults to on, make all pages render correctly today, and worry about performance improvements later.
Functionality beats performance anytime.
2) I would suggest the refresh be double buffered, and just swap buffers when it's done. The page will take the same time to load, but after x (the time it takes it refresh the page) ms, the page will also look right.
Further to my last comment - We doesn't firefox force a redraw when the page has finished loading?
If the fix is as simple as changing the text size etc. to get it to redraw, why can't we just automatically do this when all elements have loaded?
Couldn't they just make a proxy that records the times that each packet is recieved, and then use this timing everytime?
so the packets would be in the same order, and the reflow bugs would be reproducable?
I just find it hard to understand how we can't reproduce things that happen in such a closed environment (a computer). It's like debugging is still in the dark ages. In science, you need to make experiments reproducable or you get laughed at.
Yeah I guess I was mostly just playing devil's advocate.:)
IMHO Linux needs to make stack protection default. We can't rely on programmers to be trained correctly. We can't rely on them to use bounds checking compilers (unless it's inheret in the language design - and that's the only language used on that computer). We can't rely on our peers to be interested enough in our code to find all our bugs.
Anyway I don't have a particular point, but i do agree with your ideals.:)
In a perfect world, there shouldn't be any wars. Rapists should be able to get off on porn and hungry people should just grow their own food.
It's really a quite simple concept.
Who cares? People shouldn't be giving advice on what should happen. We have to plan for the worst. Because the worst is just as possible.
Noone should be promising that something isn't possible when it is.
RMS, I love how you always start treads as anonymous coward, and reply to them yourself so that we like you more.
It tricked me for years but it was because i wasn't stepping into your shoes. Now I am as paranoid as you, and i can see what you're doing.
Rock on.
Re:a series of well planned coincidences.
on
Dive Into Python
·
· Score: 1
Okay, then what do you want to code? What do you want to know?
All you have to do, is try to do it. Focus on a task at a time.
What is a task you want to do, that you have (or nearly have) all the skills to do. Then do it. You'll learn heaps from that - programming never goes as smoothly as anyone would like.
You'll get ideas for improvments and additions, and the next thing you know, you'll have learnt heaps more than you thought you would have.
If you run out of ideas, try implementing an RFC. Make an ident daemon. Pretty easy stuff, but you'll learn heaps.
if you're ever in a situation that you need a ditch and you can't operate a shovel, then you'll have two choices. 1) Pay someone to dig you a ditch. 2) Learn to use a shovel first.
Now we're going hell offtopic here, but i think he's to traditional art what post-modenism is to modernism.
His philosphy took what already existed, and saw more into it, thus incorporating the old philosphy in his new one.
It's hard to make a new era past post-modernism. Because it's just a wider world view, incorporating anything we might do.
It's like making a better methodology to science.
Andy warhol did something close, by making art just marketting, but duchamp had that beat cause he's saying art is simply pointing something out. Yeah DUH, if art is a message then you need to tell it to someone!
Man. That was so close. The thing about these question with celebs is, you need to phrase it in a question. Kinda like jeopardy. You would have won the question for sure! Just write it something like;
Who would I piss on only if they weren't in a fire?
and then Rob Glaser gets to answer. Good luck, I hope you get mod points.
I hate when people demean art by claiming that their homespun crap is on par with the work of true artists. I'm looking right at you, Thomas Kinkade!
Ever heard of Duchamp? or the readymade? Duchamp did things like put a toilet in an exhibition. Duchamp didn't demean art in the same way that the photograph didn't/doesn't demean art.
Lookup the whole Dada movement. Art is about it's own destruction as much as it is about anything else.
Slashdot Mirror
I couldn't agree with the parent more.
I'm genuinely suprised that there hasn't been more inroads into trust chain architecture. The current web works because it requires people to get the semantics - and people are good at working out whether or not they trust a source.
There is a logical seperation between authors. With the semantic web, everything is presented together. How do you know if you can trust it? How do you know that all authors are presenting data the way it is?
What if they just interpret the data differently?
It depends what you define an application to be. Something that is loaded by an OS? Well if this is an OS, and it lets you load linux, then by that definition it's an application.
Maybe a better question - Is DOSEMU an application? Are any emulators applications? Is Wine an application (No I know it's not an emulator)?
How do you define an operating system? It's all code. Linux can also be run as a BIOS - does that mean all bios's are operating systems? (Just not very good at it)
Maybe you should get away from the bull.
Old mate, you said one thing, then when it was argued, you used some physics calculator to argue a different thing. Maybe you should consider a carreer in politics.
you originally said:
So i said:
And then you give me a crapload of calculations for
Holy crap you're a tripper. All I'm saying is that one big rock would cause more damage than gravel. And, just like a politician, you've argued some different point.
In the end you try and get back to some sort of resemblance of a point - good on you. Yeah the kenetic energy would be converted to heat. No it wouldn't be a nice thing. But i'm not arguing if it would be a nice thing of not. I'm just saying that one chunk of rock would cause more damage than the same mass of gravel.
If you want some physics involved, Take one of the most pure physics laws we know. Newtons law. Force = Mass x Acceleration. If the mass of the object is less, then the force is less if the acceleration remains constant. I don't need pages to show you that gravel has less force than a big rock.
Now can you quit with your whining - you're giving nerds a bad name.
Gravel at 25,000 miles an hour would burn up in the earth's atmosphere.
A Billion ton chunk of rock has enough shielding/insulation to not burn up. It would cause a crapload more damage.
How about we try an experiment. We put a tonne of sand above my house in a geosyncronous orbit. And we put a rock that weighs a tonne in a geosyncronous orbit above your house. Then we put them both straight down.
You supply the funding, i'll supply the rock.
umm, No way dude. There are no such thing as dumb linux users.
Is this Score 5, Interesting because it's interesting that a)
as in it's interesting cause he's telling the truth but we can relate or b)
i'm confused but i don't mean to troll.
Thank you for sheding some common sense on this nonsense.
XUL is a great concept, but it will be a while before it's ready to develop decent apps in.
At the moment, there are bugs preventing things as simple as an itunes style datasheet - where you can click in a field and an editbox appears for editing. Even the nearest workaround has a bug attached to it.
XUL's a great theory, but if it were as great as the hype machine says it is, we'd be seeing a crapload more XUL apps by now.
checkout http://wiki.mozilla.org/XULDev?NeilsWishlist to see some of the things XUL needs to be ready for the primetime.
No way dude. If worst came to the worst, people will just use trusted VPNS.
What a lot of people don't realise, is the internet is alive. Yeah it's a symbiotic relationship (to humans), but it now is evolving at a faster rate than people need it to. It's evolution is governed by our collaboration but at the same time it's our collaboration medium.
So we're using it to collaborate, and it improves by us collaborating.
Life's soul purpose is to survive.
The internet is dead - just like BSD.
Okay you got me. A computer isn't a totally closed environment.
But anything that happens in software, can be reproduced in software. And everything that happens in a computer, is software. Interrupts are a slightly different story, but in modern operating systems they are virtualised enough to make the distinction irrelivant.
Give me a situation that cannot be reproduced due to physical limitations of computers. If it happens once, you should be able to reproduce it. You CAN reproduce it, but current tools don't let you.
I have myself coded quite large multi-threaded systems. The one in particular i'm thinking of was basically an automated testing system, which commonly (depending on the testing scenario) spawned 150 threads, each with it's own RTP (Real time protocol) client. It had a bug at one stage which involved sometimes (rarely) clearing a read mutex after already reading the data associated with it. Because the data had already been sent, but the read mutex was clear, the thread would deadlock - the data had arrived but it had no way of notifying anyone. Due to the multi-threaded nature of this application, most of the time it would be okay. But when context switches were at the wrong times, deadlock!
Yeah, this seems like random behaviour. But it's not. If the operating system's scheduling algorithm was simpler, say if it only context switched after the mutex was clear, it would be sweet.
We (programmers) have total control over our computers, yet we continue to debug in an environment that isn't controlled. If the bug is context switch/timing related, why can't we just run it a few times until the bug occurs, and then replay it with the same timing of context switches etc.?
In context of a buffer overflow bug, why can't we just store what's in memory to a file, in such a way that we can see what structures are where, and then inspect it at will? Yeah i've used memory inspectors - they could be several orders of magnitude more user friendly.
That's what I mean. It's computationally possible, but we don't do it. We are still in the dark ages of computer programming.
While I wholeheartedly agree to the spirit of your comments, I think they need addressing.
1) Yes it's a hack. Yes it's better to actually find the cause of the problem. I hate hacks as much as the next guy - but to the user it would solve all these "reflow bugs" easily and simply. There is nothing to stop them being fixed in the background and for this hack to eventually be taken out. But to a user, if a page looks screwed up, and it doesn't in internet explorer, it's firefox's fault.
Make it a preference that defaults to on, make all pages render correctly today, and worry about performance improvements later.
Functionality beats performance anytime.
2) I would suggest the refresh be double buffered, and just swap buffers when it's done. The page will take the same time to load, but after x (the time it takes it refresh the page) ms, the page will also look right.
Further to my last comment - We doesn't firefox force a redraw when the page has finished loading? If the fix is as simple as changing the text size etc. to get it to redraw, why can't we just automatically do this when all elements have loaded?
Couldn't they just make a proxy that records the times that each packet is recieved, and then use this timing everytime?
so the packets would be in the same order, and the reflow bugs would be reproducable?
I just find it hard to understand how we can't reproduce things that happen in such a closed environment (a computer). It's like debugging is still in the dark ages. In science, you need to make experiments reproducable or you get laughed at.
When will computer science be a real science?
Yeah I guess I was mostly just playing devil's advocate. :)
:)
IMHO Linux needs to make stack protection default. We can't rely on programmers to be trained correctly. We can't rely on them to use bounds checking compilers (unless it's inheret in the language design - and that's the only language used on that computer). We can't rely on our peers to be interested enough in our code to find all our bugs.
Anyway I don't have a particular point, but i do agree with your ideals.
In a perfect world, there shouldn't be any wars. Rapists should be able to get off on porn and hungry people should just grow their own food. It's really a quite simple concept. Who cares? People shouldn't be giving advice on what should happen. We have to plan for the worst. Because the worst is just as possible. Noone should be promising that something isn't possible when it is.
if the light you were emitting was of infinite intensity, then yes. But I can't see that happening realistically.
RMS, I love how you always start treads as anonymous coward, and reply to them yourself so that we like you more.
It tricked me for years but it was because i wasn't stepping into your shoes. Now I am as paranoid as you, and i can see what you're doing.
Rock on.
Okay, then what do you want to code? What do you want to know?
All you have to do, is try to do it. Focus on a task at a time.
What is a task you want to do, that you have (or nearly have) all the skills to do. Then do it. You'll learn heaps from that - programming never goes as smoothly as anyone would like.
You'll get ideas for improvments and additions, and the next thing you know, you'll have learnt heaps more than you thought you would have.
If you run out of ideas, try implementing an RFC. Make an ident daemon. Pretty easy stuff, but you'll learn heaps.
if you're ever in a situation that you need a ditch and you can't operate a shovel, then you'll have two choices.
1) Pay someone to dig you a ditch.
2) Learn to use a shovel first.
That applies to all of life.
Now we're going hell offtopic here, but i think he's to traditional art what post-modenism is to modernism.
His philosphy took what already existed, and saw more into it, thus incorporating the old philosphy in his new one.
It's hard to make a new era past post-modernism. Because it's just a wider world view, incorporating anything we might do.
It's like making a better methodology to science.
Andy warhol did something close, by making art just marketting, but duchamp had that beat cause he's saying art is simply pointing something out. Yeah DUH, if art is a message then you need to tell it to someone!
Man. That was so close. The thing about these question with celebs is, you need to phrase it in a question. Kinda like jeopardy. You would have won the question for sure! Just write it something like;
Who would I piss on only if they weren't in a fire?
and then Rob Glaser gets to answer. Good luck, I hope you get mod points.
Ever heard of Duchamp? or the readymade? Duchamp did things like put a toilet in an exhibition. Duchamp didn't demean art in the same way that the photograph didn't/doesn't demean art.
Lookup the whole Dada movement. Art is about it's own destruction as much as it is about anything else.
can you, kind sir please have this out with this guy? He says the opposite.
unless you're a politician.
Even if there is some architectural problem with extensions that make it require a restart, it would be nice if something like session saver: http://extensionroom.mozdev.org/more-info/sessions aver
When you install an extension, it could automatically save your session, restart your browser and restore your session on reload.
but yeah, having to restart your browser is a bit of a pain!
How long will it be until they add this functionality on the XBox?
Thousands of kids, not even needing to leave home to spend their pocket money!
I'm actually suprised this isn't a opening feature of their shop - xbox conectivity.
I wonder how long it will be before their is a credit card scanner attachment for the xbox.