I can say from the AERO/OMNI experience that there is no substitute for having hands-on involvement with the actual code. Architectural concepts such as “send an update if anything changes” can be simple in theory but become difficult or completely intractable to code – and there is no way of knowing that without actually sitting down with the code and putting the theories into practice. You might be able to team with good coders who you can hand parts of the spec to and have them implement it, but unless they fully share your passion and unless you carefully review the code you can end up with junk code that needs to be either rewritten or tossed out completely and start from scratch. And, only through actual hand-on experience with the code yourself can you separate the wheat from the chaff from what other people contribute. So, while I agree that it is not necessary for the architect to write 100% of the code themselves, I believe it is mandatory that they have an in-depth knowledge of the framework to the point that they can recognize the good from the bad and make informed decisions when admitting/rejecting code offered by others. Better yet is to have a larger team with a good mix of architectural and software engineering skills. But , the team needs to work very closely together and have the right give-and-take to accept the good stuff and reject the bad stuff. A mutual respect and a shared passion working toward a common goal are also mandatory.
Fred From: Int-area [mailto:[email protected]] On Behalf Of Khaled Omar Sent: Thursday, September 17, 2020 9:17 AM To: Stewart Bryant <[email protected]>; Ted Lemon <[email protected]> Cc: int-area <[email protected]> Subject: [EXTERNAL] Re: [Int-area] IPv10 draft (was Re: FW: [v6ops] v6ops - New Meeting Session Request for IETF 109 - IPv10) >> It takes many types of engineer to design the Internet, some will be good at >> somethings, others will have complementary skills. What is important is that >> as a team we create a design that meets the requirement. And this is what I ask for. >> To suggest that absence of a particular skill disqualifies an individual >> from participating is one of the most counter-divestity things I have heard >> on this list. Shouldn’t be one man show. We cannot repeat that millions of time, someone should run STP please ☺ Khaled Omar From: Int-area <[email protected]<mailto:[email protected]>> On Behalf Of Stewart Bryant Sent: Thursday, September 17, 2020 6:12 PM To: Ted Lemon <[email protected]<mailto:[email protected]>> Cc: int-area <[email protected]<mailto:[email protected]>> Subject: Re: [Int-area] IPv10 draft (was Re: FW: [v6ops] v6ops - New Meeting Session Request for IETF 109 - IPv10) If you can’t write code, what business do you have proposing standards? Proposing standards is an additional skill on top of network programming, not a separate skillset. That is an opinion. I am not sure it is correct, or necessary to express it. It is not one that I agree with. It takes many types of engineer to design the Internet, some will be good at somethings, others will have complementary skills. What is important is that as a team we create a design that meets the requirement. To suggest that absence of a particular skill disqualifies an individual from participating is one of the most counter-divestity things I have heard on this list. My concern at that remark should not be taken as an opinion one way or the other on the protocol proposal we are discussing. - Stewart
_______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
