3 ms·

I wouldn't assume the commenters are young, but if they are, attracting new blood to kernel development is a good thing, right? Especially if the new blood is p

by drewm1980 7y ago

I wouldn't assume the commenters are young, but if they are, attracting new blood to kernel development is a good thing, right? Especially if the new blood is prevented by the language from introducing memory errors :)

jacquesm 7y ago
New blood and the kernel sound like a great combination for problems. There are many more kinds of errors that people could make besides memory errors. Linux powers a good portion of the world, the vetting of kernel contributors should be done accordingly and the processes put in place should be strong enough that even if 'young and new blood' makes contributions that these are reviewed with a keen eye to all the lessons learned over the years, something those new and young people will still need to do. I'm all for including newcomers into important open source projects but the kernel is the one place where I would expect some experience to be a requirement before contributing simply to avoid overloading the people further up the chain with a stream of obvious errors.
- majewsky 7y ago
  Correct me if I'm wrong, but I've been under the impression that kernel contributors are not vetted? Only patches are vetted.
  - jacquesm 7y ago
    That would open up the kernel to malicious contributions. It is always a good idea to know who your counterparty is. There is good precedent to warrant such vigilance: https://www.newscientist.com/article/dn24165-how-nsa-weakens-encryption-to-access-internet-traffic/ https://www.newscientist.com/article/dn24165-how-nsa-weakens... Accidents can never be ruled out but a malicious operator will have far less chance of getting away with something when detected early.
  - rcxdude 7y ago
    You are mostly correct, there's no real barrier to just anyone submitting a patch. However, the kernel does have a rule against anonymous or pseudonymous contributions, mainly for licensing clarity reasons (not that you need to submit any form of proof of identity).
  - StreamBright 7y ago
    Even the patches are not vetted that much, give the track record of the linux kernel from the security point of view.
- psv1 7y ago
  When the first prototypes of Linux were released in 1991 Linus was 21-22 years old.
  - jacquesm 7y ago
    Yes, and nobody cared about whether it was released in one piece or not. And there was plenty wrong with those initial releases, especially compared to what was already out there. Linux is now mission critical, which means the rules of 1991 no longer apply.
    - StreamBright 7y ago
      You mean services built with Linux are mission critical?
  - jononor 7y ago
    And it was not used anywhere. And the code was 1/1000 the size of what it is now. Today it runs business critical operations across the entire globe. It is not the same situation as in 1991.

jacquesm 7y ago

New blood and the kernel sound like a great combination for problems. There are many more kinds of errors that people could make besides memory errors. Linux powers a good portion of the world, the vetting of kernel contributors should be done accordingly and the processes put in place should be strong enough that even if 'young and new blood' makes contributions that these are reviewed with a keen eye to all the lessons learned over the years, something those new and young people will still need to do. I'm all for including newcomers into important open source projects but the kernel is the one place where I would expect some experience to be a requirement before contributing simply to avoid overloading the people further up the chain with a stream of obvious errors.

majewsky 7y ago
Correct me if I'm wrong, but I've been under the impression that kernel contributors are not vetted? Only patches are vetted.
- jacquesm 7y ago
  That would open up the kernel to malicious contributions. It is always a good idea to know who your counterparty is. There is good precedent to warrant such vigilance: https://www.newscientist.com/article/dn24165-how-nsa-weakens-encryption-to-access-internet-traffic/ https://www.newscientist.com/article/dn24165-how-nsa-weakens... Accidents can never be ruled out but a malicious operator will have far less chance of getting away with something when detected early.
- rcxdude 7y ago
  You are mostly correct, there's no real barrier to just anyone submitting a patch. However, the kernel does have a rule against anonymous or pseudonymous contributions, mainly for licensing clarity reasons (not that you need to submit any form of proof of identity).
- StreamBright 7y ago
  Even the patches are not vetted that much, give the track record of the linux kernel from the security point of view.
psv1 7y ago
When the first prototypes of Linux were released in 1991 Linus was 21-22 years old.
- jacquesm 7y ago
  Yes, and nobody cared about whether it was released in one piece or not. And there was plenty wrong with those initial releases, especially compared to what was already out there. Linux is now mission critical, which means the rules of 1991 no longer apply.
  - StreamBright 7y ago
    You mean services built with Linux are mission critical?
- jononor 7y ago
  And it was not used anywhere. And the code was 1/1000 the size of what it is now. Today it runs business critical operations across the entire globe. It is not the same situation as in 1991.