On 7/8/19 16:26, Eric Rescorla wrote: > > > On Wed, Aug 7, 2019 at 6:22 AM Fernando Gont <fgont@xxxxxxxxxxxxxxx > <mailto:fgont@xxxxxxxxxxxxxxx>> wrote: > > On 7/8/19 16:13, Eric Rescorla wrote: > > > > > > On Wed, Aug 7, 2019 at 4:49 AM Stephen Farrell > > <stephen.farrell@xxxxxxxxx <mailto:stephen.farrell@xxxxxxxxx> > <mailto:stephen.farrell@xxxxxxxxx > <mailto:stephen.farrell@xxxxxxxxx>>> wrote: > > [...] > > > > > > It's worth noting that the way we present errata significantly > decreases > > their usefulness, as they are not really that obvious when one > reads the > > RFC. The inline errata display project would significantly > increase the > > value of errata, and at least for me, would have increased the > > importance of processing them. > > While the non-IETFer may be unaware about the errata system, and simply > work from the RFC repo, > > > Not just the non-IETFer. I routinely work from the downloaded RFCs and even > when I go to the Web site, I don't click through to the errata. I routinely work from downloaded RFCs. However, when approaching a new spec, it is nice to be able to check which errors have been reported and verified, as opposed to having to go through them myself, and figure out if they make sense or not -- particularly if I'm new to the spec. > processing the errata is important for at least > two reasons: > > 1) It relieves the folk that *Are* aware about the system to have to > figure out by themselves which of the reported errata make sense, and > which don't. > > 2) It serves as a basis for working on bis documents, or even triggering > bis documents. > > > Having prepared a number of bis documents, I am of the opinion that this > is not a major part of the effort in doing so. Having worked on a number of this documents, a starting point for me has been to address verified/hfdu errata, among other things. Also, if you happen to have a document with tons of verified errata, that might be an indication that a bis is warranted. > Like many things, this is a matter of cost and benefit, and if the > benefit were > larger, then it would make the cost more worth bearing. I agree with you that the tool could be valuable and useful. That said, if we have this: <https://www.ietf.org/blog/iesg-processing-rfc-errata-ietf-stream/> , does it make sense for ADs to discretionally decide whether they want to process errata? If one is going to have long queues of unprocessed errata, and accept that happily, then shut down the errata system, or don't pretend that filled errata have a "status" (make them all "reported"). That's my point. (fwiw, no, i don't think that's the way to go). Thanks, -- Fernando Gont SI6 Networks e-mail: fgont@xxxxxxxxxxxxxxx PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
