Hi Tim, Good idea.
I've been wondering a bit about what one might call the OHM technical challenges (over & above standard OSM stuff). Aside from the obvious time-slider side of things, there are a couple of others which might be worth mentioning/including in a technical presentation, because they are relevant to OSM but are more apparent with OHMs because of different needs/use-cases : - *Source/provenance stuff*. Most data added to OHM will require a fair degree of research & is not standardly verifiable a la OSM, plus use by digital humanities scholars pretty much mandates a high degree of documentation of the data. I have therefore been thinking a bit about how one might better support this need. A simple way would be to extend the regular OSM schema by introducing a column called *metatags *on nodes, ways, relations. This would behave exactly as the current tags column except items placed in this data would be meta-data (obvious things like source, attribution, fixme & notes tags in OSM). Effectively the current tags column on the changeset table is only meaningful for metadata tags. I dont know how far such a change would affect the overall API, but for editors the one way of implementing might be to split the current advanced tag entry panes into two. The other would be a toggle to flag individual tags as meta or not. (The ODI OpenAddresses project has some interesting ideas about handling provenance, but I dont know whether they fit readily with primarily user-generated data, - *Coastlines*. Currently coastline handling in the OSM renders is a bit of a kludge. This becomes abundantly clear when one wants to handle changes in coastlines (eg. Isle of Thanet in Roman times, Buenos Aires in late 19th C and Hong Kong in late 20th C). - *Map Data Scales*. OSM is broadly converging on accurate map detail around 1 m, and technically can go to about 1 cm or less. OHM is quite likely to contain data which is highly accurate (e.g., accurately surveyed archaelogical data) and very general (continent-wide road networks), and they may need to co-exist (Roman period). I imagine we will continue to handle this at the tagging level, and therefore doesn't fall within the ambit of a technical presentation. However, it may be in the future OSM may want to consider a way of handling data at different scales as a way of coping with the generalisation problem. - *Historical Gazetteer.* Needs mentioning, if only to highlight that retro-fitting temporality into a piece of software is usually hard, whereas if designed in at the outset, the present is only a special case. Anyway if you dont use these notes, I will! Cheers, Jerry On 22 March 2015 at 14:26, Tim Waters <[email protected]> wrote: > Have submitted a talk about the technical side of OHM - and have said it > could be bundled up with other OHM talks if necessary. > > On 22 March 2015 at 01:32, Rob H Warren <[email protected]> wrote: > >> I submitted one today about historical flood mapping using open street >> map and newspapers. >> >> Richard, I added in the notes a request for an OHM session. We'll see >> what happens. >> >> -rhw >> >> > On Mar 21, 2015, at 9:00 AM, [email protected] wrote: >> > >> > Message: 3 >> > Date: Fri, 20 Mar 2015 17:19:24 -0400 >> > From: Richard Welty <[email protected]> >> > To: "[email protected]" <[email protected]> >> > Subject: Re: [OHM] SOTM US proposal deadline >> > Message-ID: <[email protected]> >> > Content-Type: text/plain; charset="utf-8" >> > >> > On 3/20/15 5:01 PM, Jeff Meyer wrote: >> >> Richard - thanks for the reminder! >> >> >> >> Has anyone submitted anything yet? >> >> >> > i submitted mine a couple of days ago. >> > >> > richard >> > >> > -- >> > [email protected] >> > Averill Park Networking - GIS & IT Consulting >> > OpenStreetMap - PostgreSQL - Linux >> > Java - Web Applications - Search >> > >> > -------------- next part -------------- >> >> >> _______________________________________________ >> Historic mailing list >> [email protected] >> https://lists.openstreetmap.org/listinfo/historic >> > > > _______________________________________________ > Historic mailing list > [email protected] > https://lists.openstreetmap.org/listinfo/historic > >
_______________________________________________ Historic mailing list [email protected] https://lists.openstreetmap.org/listinfo/historic
