6 ms·

Why are Internet standards called “Request for Comments”?

85 points by jasoncrawford 13y ago

tptacek 13y ago
It's interesting and a little ironic to observe how the IETF has evolved over the last 30 years. Time was, the IETF was more than the standardization effort for the Internet; it was also an intellectual response to the institutional standards body of the day, the CCITT/ITU-T. Where the ITU was bogged down by process, riven by commercial interests and infighting, and unapproachable by researchers, the IETF was animated by "rough consensus and working code". Clearly, in the contest between the ITU-T (CLNP) and IETF (IP), the ITU-T lost. Presumably, many hundreds of people were involved in telecoms standardization at ITU-T. Where do we suppose those people went? Did they just give up on their work? Or did they instead migrate to the IETF? Either way: the IETF functions more like the ITU-T today than like the IETF of 1994. "Standards" are owned by denizens of the IETF process; new functionality unknown to the Internet is specified in standards documents before it's ever implemented, or, better yet, "standardized" in opposition to working code. I'd tentatively suggest that the IETF has served its purpose, and is now at risk of outliving it.
- jordanb 13y ago
  This is an interesting perspective. I always loved the "Joke" RFCs from the earlier days of the IETF. It's sad to see the organization get away from that ethos. What are your thoughts on the W3C? It seems like they spent the early 2000s going down the path you describe with XHTML, but with HTML5, they've rediscovered "rough consensus and working code," albeit driven entirely by the big browser vendors.
  - icebraining 13y ago
    Take my uninformed opinion with a massive grain of salt, but it seems to me that the W3C is mostly glad to leave the HTML5 nonsense to the WHATWG while they work on the stuff they really care about, such as the Semantic Web.
  - sgtpepper 13y ago
    Why "loved" in the past tense? There are new joke RFCs every year: http://www.rfc-editor.org/rfc/rfc6921.txt http://www.rfc-editor.org/rfc/rfc6921.txt
  - plorkyeran 13y ago
    HTML5 is the product of WHATWG, which was specifically created to work around the fact that the W3C was trying its hardest to be utterly irrelevant. > The WHATWG was founded by individuals of Apple, the Mozilla Foundation, and Opera Software in 2004, after a W3C workshop. Apple, Mozilla and Opera were becoming increasingly concerned about the W3C’s direction with XHTML, lack of interest in HTML and apparent disregard for the needs of real-world authors.
- noselasd 13y ago
  Many of them probably went to do similar work at other bodies such as 3GPP - which have managed to produce successful protocols and networks.
- contingencies 13y ago
  I'd tentatively suggest that the IETF has served its purpose, and is now at risk of outliving it. This matches my own experience, as the author of multiple internet standards drafts and someone who asked them whether it would be possible to open a group to discuss financial networking. What a hornet's nest that stirred up! Key examples of outdated properties include: finnicky document format requirements (straight from the 1970s), no unicode support within documents (PITA), document character width limitations that can prevent the effective presentation of required information (no fix for this), monolingual nature of its website and resources, and its own bureaucracy despite its official stance: "this is not a bureaucracy". It occurs to me that we could replace the IETF RFC process with a git repo or a blockchain. Github should probably do this proactively.
- girvo 13y ago
  I read an amazing article about all that... I think it was focused on TCP/IP, and the big telcos ideas competing. I wish I knew what the link was.
waterlion 13y ago
Is what passes for a blog post? Copy-pasting 4 paragraphs out of a book?
- drakaal 13y ago
  The Internet is all about free exchange of information... Copyright be darned. :-) Worse it isn't really even a good copy paste. I don't even like WikiPedia, and it is a much better explanation on this. http://en.wikipedia.org/wiki/Request_for_Comments http://en.wikipedia.org/wiki/Request_for_Comments I consider this the definitive origin story of RFC: http://www.wired.com/wiredenterprise/2012/05/steve-crocker/all/ http://www.wired.com/wiredenterprise/2012/05/steve-crocker/a...
- SilasX 13y ago
  If you can't beat 'em, join 'em! On your own blog, do a cut-n-paste of what you got from googling a common question, and get the same attention!
tiffani 13y ago
It's interesting that the name "Request for Comments" invokes "Oh, this is a club that I can play in too," as this is the impression I've gotten as I've read increasingly more of them lately and learn how they come to exist at all.
ams6110 13y ago
Makes sense to request comments and input as standards are being developed, but I always wondered why these documents never graduate from "RFC" to "Specification" which is what they ultimately really are.
- KC8ZKF 13y ago
  They can and do become Internet Standards, or STDs. http://en.wikipedia.org/wiki/Internet_standard http://en.wikipedia.org/wiki/Internet_standard
Slackwise 13y ago
I like how hovering over "kudos", in an attempt to understand what it is, automatically performs an action I didn't want to perform. And there is no undo. Great UI there, guys. I like how you focus on aesthetic novelty instead of functionality, but I guess that explains why you're hiding the UI all over the site until you hover over crap [1]. [1]: http://en.wikipedia.org/wiki/Mystery_meat_navigation http://en.wikipedia.org/wiki/Mystery_meat_navigation
- startling 13y ago
  What does it do?
  - Widdershin 13y ago
    It's a like button, effectively. Not a Facebook like button, just a 'I like this post' button.
  - hablahaha 13y ago
    What's really hilarious is they tell you "Don't Move". I made a gif of the behavior in question, so you don't have to try it yourself. I inadvertently gave them like, five more kudos trying to capture it: http://oi39.tinypic.com/24e2kw1.jpg http://oi39.tinypic.com/24e2kw1.jpg Sorry, I never make videos, much less upload them, so I didn't really know what services to use. If anyone has recommendations, I'd love to hear about htem.
    - underwater 13y ago
      imgur.com is the current standard in throw away image hosting.
      - hablahaha 13y ago
        Yeah, but then it's linked to my account then, right? I guess I was looking for more of a throw away picture hosting scheme
wpietri 13y ago
One of the things that I love about RFCs is that they're an existence proof. Sometimes I'll get asked, "How can we have a successful organization without a lot of top-down control of X?" where X is something like architecture or process or coding standards or furniture choice. When people see problems, they imagine solutions pushed through a power structure. And of course, they imagine themselves as the ones in power, forgetting how many bullshit edicts they've had to deal with over the years. The Internet and its RFCs are my favorite existence proof that you don't need centralized control to get good design and reliable systems. Indeed, you could argue that the Internet, beat out the other early networks because it wasn't centrally controlled.

85 points by jasoncrawford 13y ago

It's interesting and a little ironic to observe how the IETF has evolved over the last 30 years. Time was, the IETF was more than the standardization effort for the Internet; it was also an intellectual response to the institutional standards body of the day, the CCITT/ITU-T. Where the ITU was bogged down by process, riven by commercial interests and infighting, and unapproachable by researchers, the IETF was animated by "rough consensus and working code". Clearly, in the contest between the ITU-T (CLNP) and IETF (IP), the ITU-T lost. Presumably, many hundreds of people were involved in telecoms standardization at ITU-T. Where do we suppose those people went? Did they just give up on their work? Or did they instead migrate to the IETF? Either way: the IETF functions more like the ITU-T today than like the IETF of 1994. "Standards" are owned by denizens of the IETF process; new functionality unknown to the Internet is specified in standards documents before it's ever implemented, or, better yet, "standardized" in opposition to working code. I'd tentatively suggest that the IETF has served its purpose, and is now at risk of outliving it.

jordanb 13y ago
This is an interesting perspective. I always loved the "Joke" RFCs from the earlier days of the IETF. It's sad to see the organization get away from that ethos. What are your thoughts on the W3C? It seems like they spent the early 2000s going down the path you describe with XHTML, but with HTML5, they've rediscovered "rough consensus and working code," albeit driven entirely by the big browser vendors.
- icebraining 13y ago
  Take my uninformed opinion with a massive grain of salt, but it seems to me that the W3C is mostly glad to leave the HTML5 nonsense to the WHATWG while they work on the stuff they really care about, such as the Semantic Web.
- sgtpepper 13y ago
  Why "loved" in the past tense? There are new joke RFCs every year: http://www.rfc-editor.org/rfc/rfc6921.txt http://www.rfc-editor.org/rfc/rfc6921.txt
- plorkyeran 13y ago
  HTML5 is the product of WHATWG, which was specifically created to work around the fact that the W3C was trying its hardest to be utterly irrelevant. > The WHATWG was founded by individuals of Apple, the Mozilla Foundation, and Opera Software in 2004, after a W3C workshop. Apple, Mozilla and Opera were becoming increasingly concerned about the W3C’s direction with XHTML, lack of interest in HTML and apparent disregard for the needs of real-world authors.
noselasd 13y ago
Many of them probably went to do similar work at other bodies such as 3GPP - which have managed to produce successful protocols and networks.
contingencies 13y ago
I'd tentatively suggest that the IETF has served its purpose, and is now at risk of outliving it. This matches my own experience, as the author of multiple internet standards drafts and someone who asked them whether it would be possible to open a group to discuss financial networking. What a hornet's nest that stirred up! Key examples of outdated properties include: finnicky document format requirements (straight from the 1970s), no unicode support within documents (PITA), document character width limitations that can prevent the effective presentation of required information (no fix for this), monolingual nature of its website and resources, and its own bureaucracy despite its official stance: "this is not a bureaucracy". It occurs to me that we could replace the IETF RFC process with a git repo or a blockchain. Github should probably do this proactively.
girvo 13y ago
I read an amazing article about all that... I think it was focused on TCP/IP, and the big telcos ideas competing. I wish I knew what the link was.

waterlion 13y ago

Is what passes for a blog post? Copy-pasting 4 paragraphs out of a book?

drakaal 13y ago
The Internet is all about free exchange of information... Copyright be darned. :-) Worse it isn't really even a good copy paste. I don't even like WikiPedia, and it is a much better explanation on this. http://en.wikipedia.org/wiki/Request_for_Comments http://en.wikipedia.org/wiki/Request_for_Comments I consider this the definitive origin story of RFC: http://www.wired.com/wiredenterprise/2012/05/steve-crocker/all/ http://www.wired.com/wiredenterprise/2012/05/steve-crocker/a...
SilasX 13y ago
If you can't beat 'em, join 'em! On your own blog, do a cut-n-paste of what you got from googling a common question, and get the same attention!

tiffani 13y ago

It's interesting that the name "Request for Comments" invokes "Oh, this is a club that I can play in too," as this is the impression I've gotten as I've read increasingly more of them lately and learn how they come to exist at all.

ams6110 13y ago

Makes sense to request comments and input as standards are being developed, but I always wondered why these documents never graduate from "RFC" to "Specification" which is what they ultimately really are.

KC8ZKF 13y ago
They can and do become Internet Standards, or STDs. http://en.wikipedia.org/wiki/Internet_standard http://en.wikipedia.org/wiki/Internet_standard

Slackwise 13y ago

I like how hovering over "kudos", in an attempt to understand what it is, automatically performs an action I didn't want to perform. And there is no undo. Great UI there, guys. I like how you focus on aesthetic novelty instead of functionality, but I guess that explains why you're hiding the UI all over the site until you hover over crap [1]. [1]: http://en.wikipedia.org/wiki/Mystery_meat_navigation http://en.wikipedia.org/wiki/Mystery_meat_navigation

startling 13y ago
What does it do?
- Widdershin 13y ago
  It's a like button, effectively. Not a Facebook like button, just a 'I like this post' button.
- hablahaha 13y ago
  What's really hilarious is they tell you "Don't Move". I made a gif of the behavior in question, so you don't have to try it yourself. I inadvertently gave them like, five more kudos trying to capture it: http://oi39.tinypic.com/24e2kw1.jpg http://oi39.tinypic.com/24e2kw1.jpg Sorry, I never make videos, much less upload them, so I didn't really know what services to use. If anyone has recommendations, I'd love to hear about htem.
  - underwater 13y ago
    imgur.com is the current standard in throw away image hosting.
    - hablahaha 13y ago
      Yeah, but then it's linked to my account then, right? I guess I was looking for more of a throw away picture hosting scheme

wpietri 13y ago

One of the things that I love about RFCs is that they're an existence proof. Sometimes I'll get asked, "How can we have a successful organization without a lot of top-down control of X?" where X is something like architecture or process or coding standards or furniture choice. When people see problems, they imagine solutions pushed through a power structure. And of course, they imagine themselves as the ones in power, forgetting how many bullshit edicts they've had to deal with over the years. The Internet and its RFCs are my favorite existence proof that you don't need centralized control to get good design and reliable systems. Indeed, you could argue that the Internet, beat out the other early networks because it wasn't centrally controlled.