Read https://github.com/mkdocs/mkdocs/issues/2108 and just have to say wow...
This worked for me: `mkdocs serve --dev-addr 0.0.0.00:8000` On Tue, Jun 20, 2023 at 7:34 PM Andrew Latham <lath...@gmail.com> wrote: > https://github.com/asterisk/documentation/pull/1 for cookies > > I will look at a 0.0.0.0 binding next > > On Tue, Jun 20, 2023 at 7:20 PM Andrew Latham <lath...@gmail.com> wrote: > >> Maybe file an issue at https://github.com/asterisk/documentation/issues >> >> I just tested and it works on localhost for me. It also prompted me for >> cookies so I will do a PR for that. >> >> On Tue, Jun 20, 2023 at 6:53 PM <aster...@phreaknet.org> wrote: >> >>> On 6/20/2023 8:33 PM, George Joseph wrote: >>> > On Tue, Jun 20, 2023 at 5:06 PM Joshua C. Colp <jc...@sangoma.com >>> > <mailto:jc...@sangoma.com>> wrote: >>> > >>> > On Tue, Jun 20, 2023 at 3:51 PM <aster...@phreaknet.org >>> > <mailto:aster...@phreaknet.org>> wrote: >>> > >>> > On 6/20/2023 10:32 AM, George Joseph wrote: >>> > > The one exception is the auto-generated documentation for >>> > > AMI/ARI/Dialplan. That I'm starting to work on now. >>> > Thanks, George - I see from the README that one can run this >>> > on a local >>> > webserver. Will the auto-generated documentation aspect tie in >>> > with this >>> > as well? I wrote my own xmldoc to HTML generator a while back >>> > so I can >>> > view documentation for out of tree modules. If this can do >>> > that out of >>> > the box, then that would certainly be nice functionality to >>> take >>> > advantage of. Will it just be sourcing from a core xml file, >>> > that we can >>> > point elsewhere if we want, or is that going to remain more >>> > upstream >>> > specific like it was with Confluence? >>> > >>> > >>> > I don't know how George plans to approach it fully, but ultimately >>> > the reference documentation will also end up as markdown and >>> > consumed with mkdocs. I do not expect those markdown files to be >>> > checked into the tree but generated as part of the deploy process. >>> > Any tooling to consume the XML and produce the markdown files will >>> > be available, so if someone wanted it locally they could. >>> > >>> > >>> > Each version of Asterisk generates between 800 and 900 pages of >>> > dynamic docs so it's going to take a bit of thought on how to >>> > integrate them. As Josh noted, we don't want those markdown files >>> > checked into the repo but we do want mkdocs to integrate them >>> > seamlessly into the main docs site, including the search indexing. >>> > We could run a full site build once a night to convert the static and >>> > dynamic pages into html and index them all BUT we don't have >>> > server-side searching available so it's done in the browser and right >>> > now, even without the dynamic pages, the search_index.json file is >>> > 4.1MB. This might make it prudent to create a "virtual" site with its >>> > own index just for the dynamic docs and link to it from the main >>> > site. Maybe even a separate virtual site for each Asterisk version. >>> > In fact, there are tools to create a versioned API site already >>> > available. Kind of like how Python does it with a drop down at the top >>> > of every page to select the Python version you want to see the >>> > documentation for. >>> >>> Thanks, George - that helps! >>> I was a bit surprised by how fast the search results were on the new >>> site. It seems to be a lot better than the old wiki (which doesn't seem >>> to work anymore, either...) >>> >>> There does seem to be an issue with the web hosting. It seems to be >>> running: >>> root@debian11:/usr/src/documentation# mkdocs serve >>> INFO - Building documentation... >>> INFO - Cleaning site directory >>> INFO - Documentation built in 16.96 seconds >>> INFO - [20:42:02] Watching paths for changes: 'docs', 'mkdocs.yml' >>> INFO - [20:42:02] Serving on http://127.0.0.1:8000/ >>> >>> But if I navigate to port 8000 on that machine in my browser, I get >>> nothing... nothing even seems to be listening on that port. >>> It works if I curl localhost on that server, so it seems to be listening >>> on just the loopback address. I don't really see how that's helpful - it >>> should probably be listening on all interfaces, so one can see what it >>> looks like graphically, no? >>> >>> Realistically though, I wouldn't want to run a separate python server >>> anyways, I just want static webpages I can serve in an Apache >>> virtualhost, like my current doc generation process. It seems if I run >>> "mkbuild docs" it does that. So if the dynamic docs have a similar >>> process this seems like it will work great! >>> >>> -- >>> _____________________________________________________________________ >>> -- Bandwidth and Colocation Provided by http://www.api-digital.com -- >>> >>> asterisk-dev mailing list >>> To UNSUBSCRIBE or update options visit: >>> http://lists.digium.com/mailman/listinfo/asterisk-dev >> >> >> >> -- >> - Andrew "lathama" Latham - >> > > > -- > - Andrew "lathama" Latham - > -- - Andrew "lathama" Latham -
-- _____________________________________________________________________ -- Bandwidth and Colocation Provided by http://www.api-digital.com -- asterisk-dev mailing list To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-dev