- From: Juan Caballero <virtualofficehours@gmail.com>
- Date: Fri, 8 Dec 2023 17:39:04 -0800
- To: public-swicg@w3.org
- Message-ID: <3b0fad2e-f9f1-47ea-9e79-34d9d4f4cf8f@gmail.com>
Hehe, @nightpool makes an interesting point about the publication that never happened. It's worth noting that on April of 2017, that confusing "and/or" was added along with the link to the future post-publication URL <https://github.com/w3c/activitystreams/commit/947960ed6b0929547a67db1dfe3dedc33a9af82c#diff-83b7041d8f15bf75c990054f5e1558ef418159825d4cb87ae455affc5968c0b7R6853>, which would imply that maybe it was uncertain at that time whether the ontology would be finished and worth publishing before the "stopwatch" ran out on CR. Clever wording, future-proofed against timeline and staffing problems! As to official versus unofficial and workitem versus... idunno, paraphenalia, I understood it to be derivative of the more canonical, consensus-hardened @Context file which is to be taken as canonical wherever the two differ; as such, it might perhaps be on the same level as testing tools donated to the CG, right? Cristiano's characterization on another fork of this thread sounds accurate to me: this more exacting expression of the consensus vocab could be a useful tool for testing implementations' processing/validation against that vocab... but only after someone takes the time to obsessively iterate it as part of a broader systematic conformance effort. Maybe worth returning to after there is more of a conformance regime in place? Also on another fork of this thread, Sarven points out that our CG doesn't have much in the way of bespoke process docs. Those couldn't hurt! I know Bengo proposed some language for "rough consensus" via mailinglist, but other than that proposal, which might not have been approved, I think we are a touch lacking in explicit/specified processes. Maybe now's a good time to pick that up again and jot down some minimum-viable/good-enough process docs for decisions like these? If anyone's feeling inspired, perhaps some kind of workflow or lifecycle for "supplemental work items" (e.g., testing tools and derived artefacts like an alternate expression of something that IS a work item) could also be folded into that CG mini-charter thingy, if it alleviated anxieties about what's happening with vocabs and testing artefacts? Glad to edit and share work if others have opinions and cycles to donate, but don't really wanna pop it out from my own head Athena-like. Thanks, __bumblefudge On 12/8/2023 2:28 PM, nightpool wrote: > Will note for the record that I don't think the OWL document ever got > published at "the namespace" as Appendix A indicates though: > > image.png > > But it does appear to be published on > https://www.w3.org/ns/activitystreams-owl > > I'm not aware of any implementers who might be using this (as Ben > mentioned, go-fed maintains its own because the official file is so > incomplete) but I agree that if it's referenced in the spec we > probably do have a responsibility to do something about it. > > On Fri, Dec 8, 2023 at 4:53 PM Evan Prodromou <evan@prodromou.name> wrote: > > Well, shucks. I've got bad news. It /is/ actually part of the spec > in a way, shape, and/or form. > > Appendix A of the Activity Vocabulary refers to it as a > non-normative definition: > > https://www.w3.org/TR/activitystreams-vocabulary/#non-normative-ontology-definition > > I can't believe I forgot this section was there. > > I think as long as we have this section we have some > responsibility to keep the file up to date. > > Evan > > On 2023-12-08 12:12 p.m., nightpool wrote: >> I agree with all of the points mentioned here, as I've said many >> times on GitHub, the OWL should be maintained/improved by the >> community of users that are utilizing it, it shouldn't be blocked >> on Editor or CG approval to get changes made. (But also, >> conversely, it shouldn't continue to have the "imprimatur" of an >> official document just because it happens to have been included >> in the repo, since it's not part of the spec in any way shape or >> form) >> >> On Fri, Dec 8, 2023, 10:42 AM Evan Prodromou >> <evan@prodromou.name> wrote: >> >> I should have explained: OWL is the Web Ontology Language. >> >> https://en.wikipedia.org/wiki/Web_Ontology_Language >> >> Evan >> >> On 2023-12-08 11:38 a.m., Evan Prodromou wrote: >> > The original development file for AS2 was an OWL file. >> > >> > I don't think it was ever edited after James Snell created >> the first >> > JSON-LD context file. >> > >> > But it's been in the Activity Streams 2.0 repository on >> GitHub since >> > it was created. >> > >> > We get occasional requests to make changes to it to bring >> it into line >> > with the JSON-LD context doc. Some LinkedData developers >> seem to >> > prefer using it. >> > >> > We had a new issue filed this week about it, and on the >> issue triage >> > call we came up with a novel solution: move the file to its >> own >> > repository in the SWICG namespace, and let people who are >> interested >> > in using and maintaining it work on the project. >> > >> > The new repository is here: >> > >> > https://github.com/swicg/activitystreams2-owl >> > >> > Big appreciation to Emelia Smith for getting the process >> rolling. >> > We've already had one PR applied. >> > >> > I had removed the file from the w3c/activitystreams repo, >> but Ben >> > Goering pointed out that it probably needed more consensus >> and a >> > discussion here. >> > >> > So, let's discuss! >> > >> > I'd love to see this unofficial file maintained and >> updated. I think >> > moving it to a repo where people in the LD community can >> maintain it >> > is a great solution. >> > >> > Evan >> > >> > >> On 12/8/2023 2:28 PM, nightpool wrote: > Will note for the record that I don't think the OWL document ever got > published at "the namespace" as Appendix A indicates though: > > image.png > > But it does appear to be published on > https://www.w3.org/ns/activitystreams-owl > > I'm not aware of any implementers who might be using this (as Ben > mentioned, go-fed maintains its own because the official file is so > incomplete) but I agree that if it's referenced in the spec we > probably do have a responsibility to do something about it. > > On Fri, Dec 8, 2023 at 4:53 PM Evan Prodromou <evan@prodromou.name> wrote: > > Well, shucks. I've got bad news. It /is/ actually part of the spec > in a way, shape, and/or form. > > Appendix A of the Activity Vocabulary refers to it as a > non-normative definition: > > https://www.w3.org/TR/activitystreams-vocabulary/#non-normative-ontology-definition > > I can't believe I forgot this section was there. > > I think as long as we have this section we have some > responsibility to keep the file up to date. > > Evan > > On 2023-12-08 12:12 p.m., nightpool wrote: >> I agree with all of the points mentioned here, as I've said many >> times on GitHub, the OWL should be maintained/improved by the >> community of users that are utilizing it, it shouldn't be blocked >> on Editor or CG approval to get changes made. (But also, >> conversely, it shouldn't continue to have the "imprimatur" of an >> official document just because it happens to have been included >> in the repo, since it's not part of the spec in any way shape or >> form) >> >> On Fri, Dec 8, 2023, 10:42 AM Evan Prodromou >> <evan@prodromou.name> wrote: >> >> I should have explained: OWL is the Web Ontology Language. >> >> https://en.wikipedia.org/wiki/Web_Ontology_Language >> >> Evan >> >> On 2023-12-08 11:38 a.m., Evan Prodromou wrote: >> > The original development file for AS2 was an OWL file. >> > >> > I don't think it was ever edited after James Snell created >> the first >> > JSON-LD context file. >> > >> > But it's been in the Activity Streams 2.0 repository on >> GitHub since >> > it was created. >> > >> > We get occasional requests to make changes to it to bring >> it into line >> > with the JSON-LD context doc. Some LinkedData developers >> seem to >> > prefer using it. >> > >> > We had a new issue filed this week about it, and on the >> issue triage >> > call we came up with a novel solution: move the file to its >> own >> > repository in the SWICG namespace, and let people who are >> interested >> > in using and maintaining it work on the project. >> > >> > The new repository is here: >> > >> > https://github.com/swicg/activitystreams2-owl >> > >> > Big appreciation to Emelia Smith for getting the process >> rolling. >> > We've already had one PR applied. >> > >> > I had removed the file from the w3c/activitystreams repo, >> but Ben >> > Goering pointed out that it probably needed more consensus >> and a >> > discussion here. >> > >> > So, let's discuss! >> > >> > I'd love to see this unofficial file maintained and >> updated. I think >> > moving it to a repo where people in the LD community can >> maintain it >> > is a great solution. >> > >> > Evan >> > >> > >>
Received on Saturday, 9 December 2023 01:39:15 UTC