Re: Community Input Sought on SOWs for RFC Production Center and RFC Publisher
The point w.r.t. MIB module checking was that during editing phase, even a small typo in a double quote or some such would render the MIB module invalid/non-compilable (i.e. invalid SYNTAX). So if RPC does not touch the text at all, then there is no need for them to check. But if they DO touch it (even for pagination or some such) I would recommend doing the check. For MIB modules that should be pretty easy, simple and not cause much extra work. Of course the MIB module needs to be checked (SYNTAX that is) BEFORE it gets submitted to AD/IESG even. Bert On 8/17/13 2:09 PM, Jeffrey Hutzelman wrote: While I have no objection in principle to the RPC doing this check, it seems likely that the burden of doing so would be significant and, at least for IETF-stream documents, not worthwhile for the relatively small gains realized. This sort of check should be done before a document is ever submitted to the IESG, let alone the RPC.
Academic and open source rate (was: Charging remote participants)
Hi Hadriel, At 12:31 16-08-2013, Hadriel Kaplan wrote: I may be misunderstanding you, but I'm proposing we charge large corporations with large travel budgets slightly *more* than others.[1] I'm not suggesting an overhaul of the system. I'm not proposing they get more attention, or more weight, or any such thing. That sounds like the ability to pay. It might be worth considering changing the student rate to an academic and open source rate and doubling the rate. I am not getting into a definition of academic or open source [1]. It is left to the organization to determine whether it is a good idea to be honest or try the weasel words [2] approach. Regards, -sm 1. If the IETF is serious about running code (see RFC 6982) it would try to encourage open source developers to participate more effectively in the IETF. 2. weasel words give the impression of taking a firm position while avoiding commitment to any specific claim.
Re: Academic and open source rate (was: Charging remote participants)
On Sun, Aug 18, 2013 at 2:51 PM, SM s...@resistor.net wrote: Hi Hadriel, At 12:31 16-08-2013, Hadriel Kaplan wrote: I may be misunderstanding you, but I'm proposing we charge large corporations with large travel budgets slightly *more* than others.[1] I'm not suggesting an overhaul of the system. I'm not proposing they get more attention, or more weight, or any such thing. That sounds like the ability to pay. It might be worth considering changing the student rate to an academic and open source rate and doubling the rate. I am not getting into a definition of academic or open source [1]. It is left to the organization to determine whether it is a good idea to be honest or try the weasel words [2] approach. Regards, -sm 1. If the IETF is serious about running code (see RFC 6982) it would try to encourage open source developers to participate more effectively in the IETF. 2. weasel words give the impression of taking a firm position while avoiding commitment to any specific claim. +1 on opensource. Especially in the application / RAI area space. There are several implementers who could benefit from the interaction as well as contribute to making standards better. Standards can be written in ways that can make implementation easier. I have seen several instances where RFCs have unnecessary complex and larger / longer than they should be. Having more implementers in the WG session room is always welcome as it will lead to better implementations and adoptions. Also since so many opensource contributors work on their own time and money (though not all of them), it would be welcome to give them a concessional rate. So much of the daily software we use / write depends on open source libraries and apps, I think will be useful to have them be a bigger part of the standards process. A guarded +1 on academics as well. The IRTF has been doing a good job of involving academics. Would love to hear more on their experiences before commenting. -- Vinayak
Re: Academic and open source rate (was: Charging remote participants)
On Aug 18, 2013, at 5:21 AM, SM s...@resistor.net wrote: 1. If the IETF is serious about running code (see RFC 6982) it would try to encourage open source developers to participate more effectively in the IETF. Define open source developers. Technically quite a lot of developers at my employer develop open source, as do many at many of the corporations which send people to the IETF. Heck, even I personally submit code to Wireshark now and then. Distinguishing between Self-paying vs. Expensing is pretty easy. Open source vs. Closed source is a big can of worms. I'd love to get more developers in general to participate - whether they're open or closed source doesn't matter. But I don't know how to do that, beyond what we do now. The email lists are free and open. The physical meetings are remotely accessible for free and open. To attend the physical meetings in person takes real money, but the registration fee is dwarfed by the travel+food+lodging costs. The most successful open-source conferences I've seen are ones that only last a couple days, and located where many of them live. (which in the US would be silicon valley area, in terms of largest concentration) But you can't just have it there once every few years - you have to have it there repeatedly to really succeed at that. And it does cost the IETF lots of money to host the physical meetings, and that cost is directly proportional to the number of physical attendees. More attendees = more cost. Remote participation cost isn't nearly as linear nor as high, afaik. -hadriel
Re: Academic and open source rate (was: Charging remote participants)
--On Sunday, 18 August, 2013 08:33 -0400 Hadriel Kaplan hadriel.kap...@oracle.com wrote: ... And it does cost the IETF lots of money to host the physical meetings, and that cost is directly proportional to the number of physical attendees. More attendees = more cost. I had promised myself I was finished with this thread, but I can't let this one pass. (1) If IETF pays separately for the number of meeting rooms, the cost is proportionate to the number of parallel sessions, not the number of attendees. (2) If IETF gets the meeting rooms (small and/or large) for free, the costs are borne by the room rates of those who stay in the hotel and are not proportionate to much of anything (other than favoring meetings that will draw the negotiated minimum number of attendees who stay in that hotel). (3) Equipment costs are also proportional to the number of meetings we run in parallel. Since IASA owns some of the relevant equipment and has to ship it to meetings, there are some amortization issues with those costs and shipping costs are dependent on distance and handling charges from wherever things are stored between meetings (I assume somewhere around Fremont, California, USA). If that location was correct and we wanted to minimize those charges, we would hold all meetings in the San Francisco area or at least in the western part of the USA. In any event the costs are in no way proportionate to the number of attendees. (4) The costs of the Secretariat and RFC Editor contracts and other associated contracts and staff are relatively fixed. A smaller organization, with fewer working groups and less output, might permit reducing the size of those contracts somewhat, but that has only the most indirect and low-sensitively relationship to the number of attendees, nothing near proportional. (5) If we have to pay people in addition to Secretariat staff to, e.g., sit at registration desks, that bears some monotonic relationship to the number of attendees. But the step increments in that participate function are quite large, nothing like directly proportional. (6) The cost of cookies and other refreshments may indeed be proportional to the number of attendees but, in most facilities, that proportionality will come in large step functions. In addition, in some places, costs will rise with the number of unusual dietary requirements. The number of those requirements might increase with the number of attendees, but nowhere near proportionately. Unusual is entirely in the perception of the supplier/facility but, from a purely economic and cost of meetings standpoint, the IETF might be better off if people with those needs stayed home or kept their requirements to themselves. So, meeting cost directly proportional to the number of physical attendees? Nope. best, john p.s. You should be a little cautious about a charge the big companies more policy. I've seen people who make the financial decisions as to who comes say things like we pay more by virtue of sending more people, if they expect us to spend more per person, we will make a point by cutting back on those we send (or requiring much stronger justifications for each one who wants to go). I've also seen reactions that amount to We are already making a big voluntary donation that is much higher than the aggregate of the registration fees we are paying, one that small organizations don't make. If they want to charge us more because we are big, we will reduce or eliminate the size of that donation. Specific company examples on request (but not on-list), but be careful what you wish for.
Re: Academic and open source rate (was: Charging remote participants)
I've been told, though obviously I don't know, that the costs are proportional. I assume it's not literally a if we get one additional person, it costs an additional $500. But I assume SM wasn't proposing to get just one or a few more open source developer attendees. If we're talking about just a few people it's not worth arguing about... or doing anything about. It would only be useful if we got a lot of such attendees. -hadriel On Aug 18, 2013, at 10:01 AM, John C Klensin john-i...@jck.com wrote: --On Sunday, 18 August, 2013 08:33 -0400 Hadriel Kaplan hadriel.kap...@oracle.com wrote: ... And it does cost the IETF lots of money to host the physical meetings, and that cost is directly proportional to the number of physical attendees. More attendees = more cost. I had promised myself I was finished with this thread, but I can't let this one pass. (1) If IETF pays separately for the number of meeting rooms, the cost is proportionate to the number of parallel sessions, not the number of attendees. (2) If IETF gets the meeting rooms (small and/or large) for free, the costs are borne by the room rates of those who stay in the hotel and are not proportionate to much of anything (other than favoring meetings that will draw the negotiated minimum number of attendees who stay in that hotel). (3) Equipment costs are also proportional to the number of meetings we run in parallel. Since IASA owns some of the relevant equipment and has to ship it to meetings, there are some amortization issues with those costs and shipping costs are dependent on distance and handling charges from wherever things are stored between meetings (I assume somewhere around Fremont, California, USA). If that location was correct and we wanted to minimize those charges, we would hold all meetings in the San Francisco area or at least in the western part of the USA. In any event the costs are in no way proportionate to the number of attendees. (4) The costs of the Secretariat and RFC Editor contracts and other associated contracts and staff are relatively fixed. A smaller organization, with fewer working groups and less output, might permit reducing the size of those contracts somewhat, but that has only the most indirect and low-sensitively relationship to the number of attendees, nothing near proportional. (5) If we have to pay people in addition to Secretariat staff to, e.g., sit at registration desks, that bears some monotonic relationship to the number of attendees. But the step increments in that participate function are quite large, nothing like directly proportional. (6) The cost of cookies and other refreshments may indeed be proportional to the number of attendees but, in most facilities, that proportionality will come in large step functions. In addition, in some places, costs will rise with the number of unusual dietary requirements. The number of those requirements might increase with the number of attendees, but nowhere near proportionately. Unusual is entirely in the perception of the supplier/facility but, from a purely economic and cost of meetings standpoint, the IETF might be better off if people with those needs stayed home or kept their requirements to themselves. So, meeting cost directly proportional to the number of physical attendees? Nope. best, john p.s. You should be a little cautious about a charge the big companies more policy. I've seen people who make the financial decisions as to who comes say things like we pay more by virtue of sending more people, if they expect us to spend more per person, we will make a point by cutting back on those we send (or requiring much stronger justifications for each one who wants to go). I've also seen reactions that amount to We are already making a big voluntary donation that is much higher than the aggregate of the registration fees we are paying, one that small organizations don't make. If they want to charge us more because we are big, we will reduce or eliminate the size of that donation. Specific company examples on request (but not on-list), but be careful what you wish for.
Re: Academic and open source rate (was: Charging remote participants)
On Sun, Aug 18, 2013 at 8:33 AM, Hadriel Kaplan hadriel.kap...@oracle.comwrote: On Aug 18, 2013, at 5:21 AM, SM s...@resistor.net wrote: 1. If the IETF is serious about running code (see RFC 6982) it would try to encourage open source developers to participate more effectively in the IETF. Define open source developers. Technically quite a lot of developers at my employer develop open source, as do many at many of the corporations which send people to the IETF. Heck, even I personally submit code to Wireshark now and then. Distinguishing between Self-paying vs. Expensing is pretty easy. Open source vs. Closed source is a big can of worms. +1 I suspect we have all done the open source thing at some point. Whether open source makes sense as a business strategy depends on your position in the ecosystem. Folk like the 10gen (MongoDB) people can't compete against Oracle for the closed source DB market so an open source plus proprietary service strategy is completely logical for them. Following the most a logical business model for your product is hardly a point of moral superiority. I am currently putting a large amount of my private code onto SourceForge as open source, should my employer get a discount for this? Should my employer pay a premium rate to allow discounts to others? Should the fact that my employer provides open source products that facilitate consuming a proprietary product count? I really don't think this makes any sense at all. Open Source is not Free Software though some people conflate the two. -- Website: http://hallambaker.com/
Re: Anyone having trouble submitting I-Ds?
On Fri, Aug 16, 2013 at 7:27 PM, Benjamin Kaduk ka...@mit.edu wrote: On Fri, 16 Aug 2013, Benjamin Kaduk wrote: My web submission told me Your submission is pending email authentication. An email has been sent you with instructions. more than an hour ago, but I haven't seen such a mail. I sent a note to internet-drafts@ mentioning my experience, but wanted to check if anyone else was seeing this behavior. It turns out this was user error -- I was not listed as an author on the previous version of the document, even though I did the actual upload for that version. The anti-hijacking feature causes the confirmation email to only go to the authors listed on the previous version of the document, so mail was not sent to me and things are working as expected. This behavior is not documented to the user when they submit the document and is therefore a bug. -- Website: http://hallambaker.com/
Re: Academic and open source rate (was: Charging remote participants)
In article 01672754-1c4f-465b-b737-7e82dc5b3...@oracle.com you write: I've been told, though obviously I don't know, that the costs are proportional. I assume it's not literally a if we get one additional person, it costs an additional $500. But I assume SM wasn't proposing to get just one or a few more open source developer attendees. If we're talking about just a few people it's not worth arguing about... or doing anything about. It would only be useful if we got a lot of such attendees. My trip to the Berlin IETF cost me about $3300, of which the registration fee was only $650. (The plane ticket was expensive, since I flew from upstate NY, but the hotel was cheap because I booked at a place a block away with a prepaid rate back in May.) If we're going to provide financial inducements for people to come, whether open source developers or anyone else, unless they happen to live in the city where we're meeting, we'll need to give them cash travel grants, not just waive the fee. The IRTF brings winners of their research prize to the meetings to present the winning papers, so we can look at those numbers to see what it costs.
Re: Anyone having trouble submitting I-Ds?
The anti-hijacking feature causes the confirmation email to only go to the authors listed on the previous version of the document, so mail was not sent to me and things are working as expected. This behavior is not documented to the user when they submit the document and is therefore a bug. It's sort of documented somewhere, but I agree that it's a bug that it doesn't tell the submitter what happened. I reported it as a bug a while ago, dunno where it is in the tracker.
Re: Academic and open source rate (was: Charging remote participants)
Hi Hadriel, At 05:33 18-08-2013, Hadriel Kaplan wrote: Define open source developers. Technically quite a lot of developers at my employer develop open source, as do many at many of the corporations which send people to the IETF. Heck, even I personally submit code to Wireshark now and then. Distinguishing between Self-paying vs. Expensing is pretty easy. Open source vs. Closed source is a big can of worms. I'd love to get more developers in general to participate - whether they're open or closed source doesn't matter. But I don't know how to do that, beyond what we do now. The email lists are free and open. The physical meetings are remotely accessible for free and open. On reading the second paragraph of the above message I see that you and I might have a common objective. You mentioned that you don't know how to do that beyond what is done now. I suggested a rate for people with an open source affiliation. I did not define what open source means. I think that you will be acting in good faith and that you will be able to convince your employer that it will not make you look good if you are listed in a category which is intended to lessen the burden for open source developers who currently cannot attend meetings or who attend meetings on a very limited budget. We can discuss about whether a few hundred United States dollars makes a significant difference or we can sit by a pool and discuss about more interesting things. Your colleagues will probably wonder why you brought more value to your company compared to them. You could tell them that it is because you like strawberry ice cream as it is something that wills the void between rational discussion and all-out thermonuclear war. :-) At 08:50 18-08-2013, Hadriel Kaplan wrote: I've been told, though obviously I don't know, that the costs are proportional. I assume it's not literally a if we get one additional person, it costs an additional $500. But I assume SM wasn't proposing to get just one or a few more open source developer attendees. If we're talking about just a few people it's not worth arguing about... or doing anything about. It would only be useful if we got a lot of such attendees. What I proposed might have an impact on just one or a few more persons. The rest is left to the imagination of the reader. :-) Regards, -sm
