I'm not buying a Bluray movie until I can rip them. Why?
(1) I want to play in a player which will skip region coding and
UOP crap.
(2) I want to play them on any device, and a frequently rip DVDs and
move them around at the moment.
Am I a dirty "pirate"? No. In fact I've only ever made an unauthorized
copy of one movie, and that was because the movie is
unavaiable through any other means. (Great film by the way...)
I have stacks and stacks of purchased DVDs at home.
This is not some tiny, insignificant business decision. This is a major decision - they've got some hardware, the hardware is
no good without an operating system, do they use Linux or BSD or license a commercial operating system or write their own OS?
Licensing the commercial operating system or writing an alternative would have cost them $100millions. They saved all that money by using
Linux in this case.
Imagine also Microsoft's reaction if they'd
just downloaded Windows "for free" off the net somewhere and used it on their routers - it wouldn't have taken them 5 years to find out the consequences of that decision.
Cisco / Linksys set themselves up for a fall here. If they wanted code they could just rip off and use whilst largely ignoring the license, why on earth didn't they just use BSD code?
These are large companies, presumably they have lawyers.
But they're acting like some kid who downloads an image from Google Image Search and uses it on their webpage - "I downloaded it off the web for free so I can just use it right?"
That's thousands and thousands of man-hour of work, sold at $5
So what? Really, so what?
I spent thousands of hours cataloguing my navel fluff by length and color and making a giant encyclopedia of it,
so everyone should pay me thousands of $$$ for it...
If users will only pay $5 a copy for your software, you'd better find a cheaper way to produce it, or just not do it at all. No one owes you a living.
Actually, reference-counted memory allocations are a terrible idea. Reference counting is the slow child of
the garbage collection family - inefficient, and very intrusive. Once everything is ref-counted you might as well
speed things up by using real GC.
That's actually nonsense. Most addicts will be able to have children before the addiction kills them.
Proof of this is that there are addicts around. There has been plenty of time when humans were
consuming toxic, addictive substances for this to
evolve out of human beings if it was going to.
If they're addicted to something
like cannabis, then the addicition won't kill them at all.
So if C is too slow, what language do you use when speed matters? Literally just curious
To be fair, I didn't actually say that C is too slow. I questioned whether glib's awful abstractions slowed things down, and asked
if anyone had benchmarks.
Nevertheless, I'll plug OCaml here. It's about as fast as C
and far nicer to use. As I mentioned in another comment, you can use C to fine-tune the layout of OCaml structures in memory
to get maximum performance.
You want to make sure that the layout is efficient and well localized, so that you're not stalling the CPU while it's fetching the next piece of data, but once you've used it you won't need it again for a long time
Exactly right! In the past I used C code to manually lay out OCaml objects to optimize access, so I could still use a nice high-level language, but get the speed advantage.
The data size was approx 34 GB, stored in an mmap segment backed by a file.
Section 6 in this document describes it.
Are C/C++ really fast? Have you got figures to back it up? I'm not just talking about using C directly, but using C with some heavyweight library like Gtk which does its own
very inefficient implementation of objects (glib), uses reference counting, and adds tons of asserts (which in a true HLL could be eliminated by the compiler).
As Wikipedia would say... [citation needed].
Good, but I'd be happier if the people involved had been arrested. Surely there must be enough information out there to trace the controllers of this bot net by now.
So now to the big worry-- how are developers going to make money? I'm not sure. There will be demand for software development, and where there's demand, there's money to be made.
Agreed 100%. It's just like being, say, a builder. Is it a terrible thing
if you build a house and then let the public live in it without paying you a fee
every time they enter? Is that putting
honest builders out of business? Will builders starve? Erm, no, because
new houses are constantly needed, and old houses are constantly repaired and replaced.
Well first of all, Red Hat employs a lot of Fedora developers (like me), which is certainly worth a lot more than a lousy coffee cup.
Secondly your premise is wrong: Fedora developers - those not employed by Red Hat - keep working
on Fedora despite not getting any rewards. Why is that? Well, some work for reputation instead of
money. A large number work for other companies who benefit from the mutual sharing of
code. The vast majority, however, don't work for / on Fedora at all. They work for Ubuntu, GNOME, Apache, and
a thousand other upstream projects, and we in Fedora and Red Hat package up those projects. (Packaging, while
an important activity, is only a tiny part of the process of writing free software).
what silverlight seeks to achieve that isnt currently offered in the web browsing experience?
There is some interesting stuff in Silverlight which isn't in the current "web browsing experience"
(or at least, not easily). Whether it's a good idea I'll leave for you to decide...
The interesting bit is XAML
which is an XML-based format for describing the user interfaces of standard applications, eg. text boxes,
forms, inputs and so on. Much richer and more fully featured than standard HTML forms, but arguably
HTML + JS + canvas gives you something similar if you're prepared to invest a lot of effort. Anyway. For client applications, it's a bit like Gnome's Glade XML descriptions.
Microsoft extended XAML so you could send it over the web, to a Silverlight plugin in a browser,
so you can turn Windows client apps into fully remote web applications.
I should add
that nothing here is innovative -- this is exactly how Mozilla XUL
works, although Mozilla didn't really ever get their act together to make XUL work
properly for local clients (you can do it, but it's a giant PITA, unreliable and
changes constantly).
As a goverment type , i think a good idea to try might be technocracy : decisions are made according to what the best solution the problem is , based on scientifical approach and simulation models , that can veryfied by everyone.
Good luck getting the religious know-nothings to agree to that.
Actually, the practical experience with direct democracy (for example from Switzerland) says the exact opposite.
Yes and no. Switzerland, clearly yes, their practical direct democracy works well.
However for the "mother of all direct democracies" you have to look at the ancient Greek city-states,
Athens being the prime example. Direct democracy in Athens resulted in the legalized
slaughter and enslavement of many Athenian enemies.
Interestingly in Athens, every government position was directly elected except for the really
important ones like military commanders.
Firstly, no one is suggesting that Firefox should mark self-signed certified sites as "secure". What's wrong is the giant error message it shows instead. Why
not show the same error message and require > 4 clicks for unencrypted pages too, since they are clearly worse than self-signed pages.
It should mark these pages just the same way
as it does with unencrypted sites.
Secondly, Orange could perform a man-in-the-middle attack (with expensive hardware) but you can detect this after the fact by
checking signatures.
The threat is not MITM attacks, it's impossible-to-detect dragnet surveillance on unencrypted traffic.
Slashdot Mirror
I'm not buying a Bluray movie until I can rip them. Why? (1) I want to play in a player which will skip region coding and UOP crap. (2) I want to play them on any device, and a frequently rip DVDs and move them around at the moment.
Am I a dirty "pirate"? No. In fact I've only ever made an unauthorized copy of one movie, and that was because the movie is unavaiable through any other means. (Great film by the way ...)
I have stacks and stacks of purchased DVDs at home.
Rich.
This is not some tiny, insignificant business decision. This is a major decision - they've got some hardware, the hardware is no good without an operating system, do they use Linux or BSD or license a commercial operating system or write their own OS? Licensing the commercial operating system or writing an alternative would have cost them $100millions. They saved all that money by using Linux in this case.
Imagine also Microsoft's reaction if they'd just downloaded Windows "for free" off the net somewhere and used it on their routers - it wouldn't have taken them 5 years to find out the consequences of that decision.
Rich.
Cisco / Linksys set themselves up for a fall here. If they wanted code they could just rip off and use whilst largely ignoring the license, why on earth didn't they just use BSD code? These are large companies, presumably they have lawyers. But they're acting like some kid who downloads an image from Google Image Search and uses it on their webpage - "I downloaded it off the web for free so I can just use it right?"
That's thousands and thousands of man-hour of work, sold at $5
So what? Really, so what?
I spent thousands of hours cataloguing my navel fluff by length and color and making a giant encyclopedia of it, so everyone should pay me thousands of $$$ for it ...
If users will only pay $5 a copy for your software, you'd better find a cheaper way to produce it, or just not do it at all. No one owes you a living.
Rich.
Actually, reference-counted memory allocations are a terrible idea. Reference counting is the slow child of the garbage collection family - inefficient, and very intrusive. Once everything is ref-counted you might as well speed things up by using real GC.
Rich.
Imaginary things are now real!
It's working out really well for "intellectual property" ...
Rich.
That's actually nonsense. Most addicts will be able to have children before the addiction kills them. Proof of this is that there are addicts around. There has been plenty of time when humans were consuming toxic, addictive substances for this to evolve out of human beings if it was going to.
If they're addicted to something like cannabis, then the addicition won't kill them at all.
Rich.
I just finished Ben Goldacre's book "Bad Science" and I can highly recommend it.
Rich.
Good tip. Do you know why on earth Wikipedia doesn't implement https://en.wikipedia.org/? That page, ie. the 'obvious' secure page, doesn't exist.
So if C is too slow, what language do you use when speed matters? Literally just curious
To be fair, I didn't actually say that C is too slow. I questioned whether glib's awful abstractions slowed things down, and asked if anyone had benchmarks.
Nevertheless, I'll plug OCaml here. It's about as fast as C and far nicer to use. As I mentioned in another comment, you can use C to fine-tune the layout of OCaml structures in memory to get maximum performance.
Rich.
(My original comment up there was modded down to -1, but it's interesting to read the answers anyway)
You want to make sure that the layout is efficient and well localized, so that you're not stalling the CPU while it's fetching the next piece of data, but once you've used it you won't need it again for a long time
Exactly right! In the past I used C code to manually lay out OCaml objects to optimize access, so I could still use a nice high-level language, but get the speed advantage. The data size was approx 34 GB, stored in an mmap segment backed by a file. Section 6 in this document describes it.
Rich.
Are C/C++ really fast? Have you got figures to back it up? I'm not just talking about using C directly, but using C with some heavyweight library like Gtk which does its own very inefficient implementation of objects (glib), uses reference counting, and adds tons of asserts (which in a true HLL could be eliminated by the compiler). As Wikipedia would say ... [citation needed].
Rich.
Erm, you want vino which has shipped as standard in GNOME (ie. all distros) for ages, at least 3 years AFAIK.
Rich.
Good, but I'd be happier if the people involved had been arrested. Surely there must be enough information out there to trace the controllers of this bot net by now.
Rich.
So now to the big worry-- how are developers going to make money? I'm not sure. There will be demand for software development, and where there's demand, there's money to be made.
Agreed 100%. It's just like being, say, a builder. Is it a terrible thing if you build a house and then let the public live in it without paying you a fee every time they enter? Is that putting honest builders out of business? Will builders starve? Erm, no, because new houses are constantly needed, and old houses are constantly repaired and replaced.
Rich.
Well first of all, Red Hat employs a lot of Fedora developers (like me), which is certainly worth a lot more than a lousy coffee cup.
Secondly your premise is wrong: Fedora developers - those not employed by Red Hat - keep working on Fedora despite not getting any rewards. Why is that? Well, some work for reputation instead of money. A large number work for other companies who benefit from the mutual sharing of code. The vast majority, however, don't work for / on Fedora at all. They work for Ubuntu, GNOME, Apache, and a thousand other upstream projects, and we in Fedora and Red Hat package up those projects. (Packaging, while an important activity, is only a tiny part of the process of writing free software).
Rich.
Red Hat have a large engineering facility in Westford which is about 30 miles away from Boston/Cambridge.
you consider the browser you used in 2002 to be no different than the one you use now.
It's better - it now pops up helpful little adverts to guide me through the internets.
I can't say the same for my internet provider though. It's just so much slower than it used to be ...
what silverlight seeks to achieve that isnt currently offered in the web browsing experience?
There is some interesting stuff in Silverlight which isn't in the current "web browsing experience" (or at least, not easily). Whether it's a good idea I'll leave for you to decide ...
The interesting bit is XAML which is an XML-based format for describing the user interfaces of standard applications, eg. text boxes, forms, inputs and so on. Much richer and more fully featured than standard HTML forms, but arguably HTML + JS + canvas gives you something similar if you're prepared to invest a lot of effort. Anyway. For client applications, it's a bit like Gnome's Glade XML descriptions. Microsoft extended XAML so you could send it over the web, to a Silverlight plugin in a browser, so you can turn Windows client apps into fully remote web applications.
I should add that nothing here is innovative -- this is exactly how Mozilla XUL works, although Mozilla didn't really ever get their act together to make XUL work properly for local clients (you can do it, but it's a giant PITA, unreliable and changes constantly).
Rich.
It seems about right to me. Remembering how slow my ZX81 was at floating point, it might even have been < 320 flops.
(Cue fanbois saying the C64 was better ...)
Rich.
you can and are moving from a democracy to an oligarchy.
Nah, that would require the EU to be a democracy in the first place, which it most certainly is not.
Rich.
As a goverment type , i think a good idea to try might be technocracy : decisions are made according to what the best solution the problem is , based on scientifical approach and simulation models , that can veryfied by everyone.
Good luck getting the religious know-nothings to agree to that.
Rich.
Actually, the practical experience with direct democracy (for example from Switzerland) says the exact opposite.
Yes and no. Switzerland, clearly yes, their practical direct democracy works well. However for the "mother of all direct democracies" you have to look at the ancient Greek city-states, Athens being the prime example. Direct democracy in Athens resulted in the legalized slaughter and enslavement of many Athenian enemies.
Interestingly in Athens, every government position was directly elected except for the really important ones like military commanders.
Rich.
Firstly, no one is suggesting that Firefox should mark self-signed certified sites as "secure". What's wrong is the giant error message it shows instead. Why not show the same error message and require > 4 clicks for unencrypted pages too, since they are clearly worse than self-signed pages. It should mark these pages just the same way as it does with unencrypted sites.
Secondly, Orange could perform a man-in-the-middle attack (with expensive hardware) but you can detect this after the fact by checking signatures.
The threat is not MITM attacks, it's impossible-to-detect dragnet surveillance on unencrypted traffic.
Rich.
Nonsense, you can easily detect if someone forges certs in a man-in-the-middle attack by comparing signatures after the fact.
But you can't detect massive dragnet surveillance of the type actually being carried out by government organizations.
Let's not confuse some theoretical, hard-to-do, impossible-to-get-away-with attack from what is actually happening in the real world now.
Rich.