Re:Perhaps the SCM Solution is not the problem
on
Linus Drops BitKeeper
·
· Score: 2, Interesting
If the linux process is so poor, why is it more popular than *BSD?
Just as Windows...
the Linux apprach says: show me the code, convince us it's better. The *BSD approach says: Show me the code, convince us it's better, get somone on the core team to sponser it, shedule it for the next (or prehaps one after next) release, and wait.
What is the real difference here? In both cases, code submitters have to get approval by maintainers: with Linux, there's only one mainainer. With BSD, there are many committers. Actually, BSDs approach seems even more failproof than Linux', because you don't have to worry about Linus going on vacation (or some BitKeeper issues). You just ask another committer to submit your code to review.
Thus, Linux has a lot more mobility, and can utilise new approaches sooner than *BSD
Hmmm, the -CURRENT branches are very mobile too, just as with Linux. There's one important difference though: BSD kernel interfaces (external and internal) tend to be a lot more stable than Linux' interfaces. This is both good and bad: with Linux, you can experiment with new interfaces and paradigms. With BSD you can do the same. But with Linux, the experiments quickly make it into the public kernel, even if they have to be retracted later. With BSD, this happens too, but it's very rare.
Let them be different - choose your favourite model, and be happy with it.
Exactly. Hosting the Linux source code repository was their ONLY publicity asset. How many people in the IT industry ever heard about them if they were not connected to the Linux project?
I'd be tempted to say: Good riddance! but it's not so easy: if people were actively trying to mimic BK's software, sooner than later BK would they have been out of business anyway. Someone in Mgmt is just trying to stop a sinking ship from taking water: they are doomed to fail.
Ask any librarian about categorizations. It's already difficult to group books in good categories. How much more difficult would it be to group everything in meaningful categories? The Hitchhiker's Guide would get it right, but ICANN is not Ursa Minor! Furthermore, categorizations are inherently dynamic: a web site that was in CAT A would be in a few months in CAT Z. And add to this that many sites fit into many categories at the same time. Should they use multiple TLDs?
The only reasonable model would be to get rid of categories at all and use a flat namespace. Currently,.com/.net/.org, but especially.com (and in some countries their ccTLD) are already playing the role of a flat namespace. Getting rid of ccTLDs and gTLDs in favor of a real flat namespace (10^9 names in the root!) is not such a real big step at all.
who "innovated" DOS, gui computing, windowed applications, mouse based ui, menus, word processor, spreadsheets, email client, address book, database... you get the picture.
DOS: developed from CP/M, not an OSS project.
gui, windows, mouse: developed at XEROX PARC as a research project. Not OSS, but paid for commercial research.
menus (pull-down menus?): SAA architecture, developed by IBM. Not open source.
word processors: multiple vendors. Even Emacs came as a replacement for a non-OSS editor.
spreadsheets: VisiCalc et. al.: not OSS.
databases: Cobol-based ISAM databases and first SQL-based RDBMS were certainly NOT open source! Again IBM et. al.
When the US government decides on global political issues, the whole world normally follows suit. So even if they don't legally govern the world, when the US government sets some global policy, it will be accepted by most other governments in the world. Of course, there are exceptions to this rule, but not so many, if you really come to think about it.
Therefore, the US government using the global.gov TLD is ironically not really that far from the political reality.
The telephone system in most countries already works that way: it maps an arbitrary telephone number to a real physical slot number. That's the reason people can keep their telephone numbers even when moving to a new location.
The shortest TLDs are 2-letter ccTLDs. Why don't we consider 1 letter TLDs as well:.a,.b, etc...?
Incidentally, the german telecom, which changes names like a chameleon every now and then, called itself something like t... a few years ago (or they used this as a logo?), and they seriously asked if they could register 't...' as a domain name. Oh well, they've changed names again since then.
.net was originally meant to be used by ISPs and network infrastructure providers..isp would be just redundant. Oh,.net became meaningless and open the the general public? Too bad...
We have.gov, instead of.gov.us, because the US Government is effectively governing the whole world (save for some few stubborn "renegate countries" they would like to govern as well). Other countries would use.gov.ccTLD to designate their *local* governments. Oh, and the french would not accept the english "gov" and would opt for.gouv.fr;-)
As for.mil: it's a wonder we don't have the TLDs.usaf,.army and.navy as well, considering how much they are competing against each other!
There's an important difference here: regular, small-scale and small-intensity anti-social behavior is not always coupled with the desire to make profit using equipment and resources that are not your own. A public drunk harms only himself, but a spam operative consumes HUGE amounts of resources (bandwidth and time) for the purpose of making money. This is very different intent and the damage to the society is much higher. THAT's why big spammers ought to get much higher penalties than a poor, mostly harmless troublemaker.
This guy should be removed from the internet, he SHOULDN'T be in prison.
Not every spammer is an IT geek. They can use (or have people use) software written by others. A spammer is essentially a businessman who uses (or induces people to use) IT technology. Even if you forbid internet access, that would not change anything, because that guy would still use his money to pay other people (with access) to spread this spam.
register-level descriptions of the hardware, which unlike the POSIX API are not public.
Yes, that's true. But it is not that hard to discover this empirically and publish a semi-official description, which would then be used to program a driver.
What would prevent nVidia and their ilk to release their drivers in source code, yet not on GPL terms? After all, they are only coding against a published (yet, unfortunately, moving) driver API. They don't have to GPL their code after all. If they released the source, it could be compiled just fine under ppc, arm or whatnot.
Commercials take up roughly 25% of the time in television programming
You can always decide to watch TV or not. If you're using Email for a living, you don't have a choice. By watching TV, you (reluctantly) agree to have 25% of your time diverted to commercials. If you don't like this, get pay TV commercials-free channels, or rent DVDs. But how would you "opt out" of spam other than by switching e-mail addresses every so often?
His crime was not a violent one, he shouldn't go to jail for 9 years.
There's something wrong with this kind of reasoning. Even murder would affect only one victim, and their relatives. Say, 10 people or so. Mass-spamming affects hundreds of millions of people every day. It steals a lot of their time from them.
If you steal from one person, you get convicted to a small sentence. But do it to millions of people, and you'll be a repeat offender and would be prosecuted by every law enforcement agency. Spamming is nothing more than repeat offense on a gigantic scale.
Killers affect only a few people, and society gets all berserk about it, going to extremes like capital punishment in some states. Spammers affect a whole lot more of people (though, certainly not as "strikingly" as killers, but nonetheless), and they should not even serve 9 years in jail?
By the same token: white collar crime (Enron, anyone?) is not a violent crime. But willfully bankrupting corporations affects also a lot of people much more than a puny murderer. Should white collar crime not be prosecuted harshly, just because it is non-violent?
Not only money, but precious time from every victim. Assuming that everyone spends just 10 seconds a day scanning spam or throwing that junk away: how many wasted centuries does that amount to?
That one guy sending out 10,000,000 emails a day has probably stolen many, many combined lifetimes from his victims. The punishment should really fit the crime, and 9 years of jail is really way too low a sentence!
Wouldn't a better punishment be somethign vaguely like what they did to Mitnick?
Hmmm... not every spammer is necessarily and automatically a computer geek. Spammers may pay some geeks to do the dirty programming job for them, but they are (slimy) businessmen before everything else. Preventing them from touching a computer won't stop them at all and would be totally useless. A heavy jail sentence is much better for this kind of antisocial behaviour.
some of the hard-liners who write the kernel apparently prefer bad hardware support to non-GPLd drivers
Most binary-only drivers are x86 only. What about other arches? There are a lot of good reasons to keep GPLed driver sources in the tree, even if they are not always as good (?) as their binary-only equivalents!
We'll get off our lazy asses, stop whining, and start doing some serious reverse-engineering to find out how that piece of hardware works and how it is supposed to be accessed!
Seriously: just because manufacturers provide binary drivers and no specs, doesn't mean that we are not responsible for writing our own drivers! Currently, a lot of people behave like spoiled kids: bhwahh! big bad vendor took my toy away! whine, cry, throw tantrums....
But Linux and BSD were developed despite USLs opposition to releasing source code. It happened anyway, because some great bright minds were not ready to put up with lock-in attempts. They wanted to be free and they took their freedom in their own hands. The same will happen, if hardware companies stopped producing binary drivers for linux: it would just motivate and spurn real hackers to write their own.
Actually, vendor's binary-only drivers are worse than no-drivers, because it stiffles motivation from developers: "why should we spend time developing our own driver, when NVIDIA provides their own already? It works, right? So what do you want?" If there were no drivers at all, hackers would say: "hey, that's a pretty nice piece of hardware: I'd like to use it! So let's discover what the Windows driver does with it, and let's mimic that too!"
That's the mindset that helped grow Linux to a well-respected OS and it is about time to go back to this kind of thinking!
It runs very well on my laptop too. But what about hibernation and other ACPI power saving modes? How well are they supported by FreeBSD right now? FreeBSD's laptop readiness is comparable to most Linux distros'. It's the overall laptop friendliness of F/OSS operating systems that could still be a tad better; linux or bsd alike. Missing open specs for laptop peripheral chipsets are still the biggest stumbling block.
It's really just a matter of time before some malware gets shipped with binary-only drivers. It could be piggy-backed on legit drivers, or it could even be required operation of the driver itself.
Imagine a graphics adapter driver, that catches all DRMed stuff, and phones home to the MPAA that YOUR machine is displaying film XYZ for the 5th time (so it HAS to be illegal, right!?). Or a wireless card driver that opens a back door to DHS and other agencies?
You may be willing to sacrifice security for convenience, but personally, I'd prefer to have the source code right there, so I (and many others) can check the integrity of the driver.
Oh yeah: in many security-aware companies, binary-only drivers is an absolute no-no. They are going open source exactly because their operations are security-sensitive, and if they now kept adding untrusted binary black boxes to their systems, even their firewalls wouldn't protect them long enough against all kind of espionage or even sabotage.
Just say no to binary-only drivers! You never really know what you get! If manufacturers don't want to disclose their source code, they may have something to hide as well. How trustworthy are they really then?
Slashdot Mirror
If the linux process is so poor, why is it more popular than *BSD?
Just as Windows...
the Linux apprach says: show me the code, convince us it's better. The *BSD approach says: Show me the code, convince us it's better, get somone on the core team to sponser it, shedule it for the next (or prehaps one after next) release, and wait.
What is the real difference here? In both cases, code submitters have to get approval by maintainers: with Linux, there's only one mainainer. With BSD, there are many committers. Actually, BSDs approach seems even more failproof than Linux', because you don't have to worry about Linus going on vacation (or some BitKeeper issues). You just ask another committer to submit your code to review.
Thus, Linux has a lot more mobility, and can utilise new approaches sooner than *BSD
Hmmm, the -CURRENT branches are very mobile too, just as with Linux. There's one important difference though: BSD kernel interfaces (external and internal) tend to be a lot more stable than Linux' interfaces. This is both good and bad: with Linux, you can experiment with new interfaces and paradigms. With BSD you can do the same. But with Linux, the experiments quickly make it into the public kernel, even if they have to be retracted later. With BSD, this happens too, but it's very rare.
Let them be different - choose your favourite model, and be happy with it.
Right!
Exactly. Hosting the Linux source code repository was their ONLY publicity asset. How many people in the IT industry ever heard about them if they were not connected to the Linux project?
I'd be tempted to say: Good riddance! but it's not so easy: if people were actively trying to mimic BK's software, sooner than later BK would they have been out of business anyway. Someone in Mgmt is just trying to stop a sinking ship from taking water: they are doomed to fail.
One day we'll listen to him BEFORE things start to go all runny
Just like... avoiding binary drivers?
Ask any librarian about categorizations. It's already difficult to group books in good categories. How much more difficult would it be to group everything in meaningful categories? The Hitchhiker's Guide would get it right, but ICANN is not Ursa Minor! Furthermore, categorizations are inherently dynamic: a web site that was in CAT A would be in a few months in CAT Z. And add to this that many sites fit into many categories at the same time. Should they use multiple TLDs?
The only reasonable model would be to get rid of categories at all and use a flat namespace. Currently, .com/.net/.org, but especially .com (and in some countries their ccTLD) are already playing the role of a flat namespace. Getting rid of ccTLDs and gTLDs in favor of a real flat namespace (10^9 names in the root!) is not such a real big step at all.
Not that I'm seriously advocating it!
.cum is anglophone, .xxx is language neutral. That's also the reason why .xxx has been proposed instead of .sex.
who "innovated" DOS, gui computing, windowed applications, mouse based ui, menus, word processor, spreadsheets, email client, address book, database... you get the picture.
When the US government decides on global political issues, the whole world normally follows suit. So even if they don't legally govern the world, when the US government sets some global policy, it will be accepted by most other governments in the world. Of course, there are exceptions to this rule, but not so many, if you really come to think about it.
Therefore, the US government using the global .gov TLD is ironically not really that far from the political reality.
The telephone system in most countries already works that way: it maps an arbitrary telephone number to a real physical slot number. That's the reason people can keep their telephone numbers even when moving to a new location.
The shortest TLDs are 2-letter ccTLDs. Why don't we consider 1 letter TLDs as well: .a, .b, etc...?
Incidentally, the german telecom, which changes names like a chameleon every now and then, called itself something like t... a few years ago (or they used this as a logo?), and they seriously asked if they could register 't...' as a domain name. Oh well, they've changed names again since then.
.aslongasthednsspecificationallowswhichisalot
.net was originally meant to be used by ISPs and network infrastructure providers. .isp would be just redundant. Oh, .net became meaningless and open the the general public? Too bad...
We have .gov, instead of .gov.us, because the US Government is effectively governing the whole world (save for some few stubborn "renegate countries" they would like to govern as well). Other countries would use .gov.ccTLD to designate their *local* governments. Oh, and the french would not accept the english "gov" and would opt for .gouv.fr ;-)
As for .mil: it's a wonder we don't have the TLDs .usaf, .army and .navy as well, considering how much they are competing against each other!
Let the UN take ICANN over and start selling global trademarks.
The UN as a merchant? Oh no! Everything else would be far better than this!
If the UN starts selling things, shouldn't they be using the un.com domain instead of the .int tld?
There's an important difference here: regular, small-scale and small-intensity anti-social behavior is not always coupled with the desire to make profit using equipment and resources that are not your own. A public drunk harms only himself, but a spam operative consumes HUGE amounts of resources (bandwidth and time) for the purpose of making money. This is very different intent and the damage to the society is much higher. THAT's why big spammers ought to get much higher penalties than a poor, mostly harmless troublemaker.
This guy should be removed from the internet, he SHOULDN'T be in prison.
Not every spammer is an IT geek. They can use (or have people use) software written by others. A spammer is essentially a businessman who uses (or induces people to use) IT technology. Even if you forbid internet access, that would not change anything, because that guy would still use his money to pay other people (with access) to spread this spam.
register-level descriptions of the hardware, which unlike the POSIX API are not public.
Yes, that's true. But it is not that hard to discover this empirically and publish a semi-official description, which would then be used to program a driver.
What would prevent nVidia and their ilk to release their drivers in source code, yet not on GPL terms? After all, they are only coding against a published (yet, unfortunately, moving) driver API. They don't have to GPL their code after all. If they released the source, it could be compiled just fine under ppc, arm or whatnot.
Commercials take up roughly 25% of the time in television programming
You can always decide to watch TV or not. If you're using Email for a living, you don't have a choice. By watching TV, you (reluctantly) agree to have 25% of your time diverted to commercials. If you don't like this, get pay TV commercials-free channels, or rent DVDs. But how would you "opt out" of spam other than by switching e-mail addresses every so often?
His crime was not a violent one, he shouldn't go to jail for 9 years.
There's something wrong with this kind of reasoning. Even murder would affect only one victim, and their relatives. Say, 10 people or so. Mass-spamming affects hundreds of millions of people every day. It steals a lot of their time from them.
If you steal from one person, you get convicted to a small sentence. But do it to millions of people, and you'll be a repeat offender and would be prosecuted by every law enforcement agency. Spamming is nothing more than repeat offense on a gigantic scale.
Killers affect only a few people, and society gets all berserk about it, going to extremes like capital punishment in some states. Spammers affect a whole lot more of people (though, certainly not as "strikingly" as killers, but nonetheless), and they should not even serve 9 years in jail?
By the same token: white collar crime (Enron, anyone?) is not a violent crime. But willfully bankrupting corporations affects also a lot of people much more than a puny murderer. Should white collar crime not be prosecuted harshly, just because it is non-violent?
t translates into real money lost
Not only money, but precious time from every victim. Assuming that everyone spends just 10 seconds a day scanning spam or throwing that junk away: how many wasted centuries does that amount to?
That one guy sending out 10,000,000 emails a day has probably stolen many, many combined lifetimes from his victims. The punishment should really fit the crime, and 9 years of jail is really way too low a sentence!
Wouldn't a better punishment be somethign vaguely like what they did to Mitnick?
Hmmm... not every spammer is necessarily and automatically a computer geek. Spammers may pay some geeks to do the dirty programming job for them, but they are (slimy) businessmen before everything else. Preventing them from touching a computer won't stop them at all and would be totally useless. A heavy jail sentence is much better for this kind of antisocial behaviour.
some of the hard-liners who write the kernel apparently prefer bad hardware support to non-GPLd drivers
Most binary-only drivers are x86 only. What about other arches? There are a lot of good reasons to keep GPLed driver sources in the tree, even if they are not always as good (?) as their binary-only equivalents!
what are you going to do?
We'll get off our lazy asses, stop whining, and start doing some serious reverse-engineering to find out how that piece of hardware works and how it is supposed to be accessed!
Seriously: just because manufacturers provide binary drivers and no specs, doesn't mean that we are not responsible for writing our own drivers! Currently, a lot of people behave like spoiled kids: bhwahh! big bad vendor took my toy away! whine, cry, throw tantrums....
But Linux and BSD were developed despite USLs opposition to releasing source code. It happened anyway, because some great bright minds were not ready to put up with lock-in attempts. They wanted to be free and they took their freedom in their own hands. The same will happen, if hardware companies stopped producing binary drivers for linux: it would just motivate and spurn real hackers to write their own.
Actually, vendor's binary-only drivers are worse than no-drivers, because it stiffles motivation from developers: "why should we spend time developing our own driver, when NVIDIA provides their own already? It works, right? So what do you want?" If there were no drivers at all, hackers would say: "hey, that's a pretty nice piece of hardware: I'd like to use it! So let's discover what the Windows driver does with it, and let's mimic that too!"
That's the mindset that helped grow Linux to a well-respected OS and it is about time to go back to this kind of thinking!
It runs very well on my laptop too. But what about hibernation and other ACPI power saving modes? How well are they supported by FreeBSD right now? FreeBSD's laptop readiness is comparable to most Linux distros'. It's the overall laptop friendliness of F/OSS operating systems that could still be a tad better; linux or bsd alike. Missing open specs for laptop peripheral chipsets are still the biggest stumbling block.
It's really just a matter of time before some malware gets shipped with binary-only drivers. It could be piggy-backed on legit drivers, or it could even be required operation of the driver itself.
Imagine a graphics adapter driver, that catches all DRMed stuff, and phones home to the MPAA that YOUR machine is displaying film XYZ for the 5th time (so it HAS to be illegal, right!?). Or a wireless card driver that opens a back door to DHS and other agencies?
You may be willing to sacrifice security for convenience, but personally, I'd prefer to have the source code right there, so I (and many others) can check the integrity of the driver.
Oh yeah: in many security-aware companies, binary-only drivers is an absolute no-no. They are going open source exactly because their operations are security-sensitive, and if they now kept adding untrusted binary black boxes to their systems, even their firewalls wouldn't protect them long enough against all kind of espionage or even sabotage.
Just say no to binary-only drivers! You never really know what you get! If manufacturers don't want to disclose their source code, they may have something to hide as well. How trustworthy are they really then?