Mikel -- that all makes sense to me too. And agree with Jeff that we should try using it til it breaks.
One question though about the format of the date_start and date_end values: Since ISO 8601 allows so many different ways to indicate a time (for instance, time offset from UTC, week numbering, etc), would we need to determine one strict interpretation of the format and require it to be used? In other words, 'if you want to upload your data, you can't have week numbers in your timestamps, even though ISO8601 allows it', etc. It seems that being strict with this, at least in these early stages, this would keep things simple. Then perhaps later, support for parsing the expanded syntax permitted by ISO8601 could be enabled, to allow uploading of data in any ISO8601-compliant format. Specifically, the ISO standard document (which I'm not ashamed to say I find pretty interesting) goes into much greater detail. Section 4.1.2.3 and 4.3.3 deals with the uncertainty issue (pages 13 and 19), relevant because of earlier conversations here on the mailing list about displaying uncertainty of times and time ranges: http://dotat.at/tmp/ISO_8601-2004_E.pdf - Brad On Wed, Feb 27, 2013 at 1:13 PM, <[email protected]> wrote: > Send Historic mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.openstreetmap.org/listinfo/historic > or, via email, send a message with subject or body 'help' to > [email protected] > > You can reach the person managing the list at > [email protected] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Historic digest..." > > > Today's Topics: > > 1. Re: Dating features (was: Introduction & questions) (Mikel Maron) > 2. Re: Dating features (was: Introduction & questions) (Rob Warren) > 3. Re: Dating features (was: Introduction & questions) (Jeff Meyer) > 4. Re: Dating features (was: Introduction & questions) (Mikel Maron) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 27 Feb 2013 09:52:51 -0800 (PST) > From: Mikel Maron <[email protected]> > To: Rob Warren <[email protected]>, > "[email protected]" <[email protected]> > Subject: Re: [Historic] Dating features (was: Introduction & > questions) > Message-ID: > <[email protected]> > Content-Type: text/plain; charset="iso-8859-1" > > how about > > for a date range: > base-uri/date_start/date_end/zoom/x/y.img > > for a moment: > base-uri/date/zoom/x/y.img > > where date values are ISO 8601 strings? > ? > * Mikel Maron *?+14152835207 @mikel s:mikelmaron > > >>________________________________ >> From: Rob Warren <[email protected]> >>To: [email protected] >>Sent: Wednesday, February 27, 2013 9:41 AM >>Subject: Re: [Historic] Dating features (was: Introduction & questions) >> >> >>No forking yet, we need to figure out time-stamping semi-officially. Rob >> >>On 27-Feb-13, at 10:18 AM, Mikel Maron wrote: >> >>> I strongly prefer encoded into the URL, to stay consistent with how tiles >>> are currently requested, and make it simple to adapt client viewers. >>> >>> Looks like "tile_translate" in mod_tile.c would be the place to start? Have >>> you forked mod_tile on github? >> >> >>_______________________________________________ >>Historic mailing list >>[email protected] >>http://lists.openstreetmap.org/listinfo/historic >> >> >> > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > <http://lists.openstreetmap.org/pipermail/historic/attachments/20130227/886acf5b/attachment-0001.html> > > ------------------------------ > > Message: 2 > Date: Wed, 27 Feb 2013 14:39:46 -0400 > From: Rob Warren <[email protected]> > To: Mikel Maron <[email protected]> > Cc: "[email protected]" <[email protected]> > Subject: Re: [Historic] Dating features (was: Introduction & > questions) > Message-ID: <[email protected]> > Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes > > > Mikel, > > On 27-Feb-13, at 1:52 PM, Mikel Maron wrote: > >> for a date range: >> base-uri/date_start/date_end/zoom/x/y.img > > Makes sense. > >> for a moment: >> base-uri/date/zoom/x/y.img > > By 'moment', I take it you mean day? Or are you thinking of "Historic" > values? > > As long as this works on a relation, I'd be happy with it. > > best, > rhw > > > > > ------------------------------ > > Message: 3 > Date: Wed, 27 Feb 2013 11:07:30 -0800 > From: Jeff Meyer <[email protected]> > To: Rob Warren <[email protected]> > Cc: "[email protected]" <[email protected]> > Subject: Re: [Historic] Dating features (was: Introduction & > questions) > Message-ID: > <caa1ffey971u+8o_mp5wcv4yfje4ahgstckeijpnp5swvcom...@mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > So, let's test two cases: > > * Time Pixels: > > base-uri/YYYY/zoom/x/y.img > > Should this return start, middle, or end of YYYY, or should we enforce > YYYY-MM-DD at a minimum? > > * Web Mercator of Time: > > If the year is -2000 BCE, how should we express that? > > I'm fine with any answer, but we should pick one and use it until it > breaks. Sound ok? I'm not smart enough to suggest the answer... > > - Jeff > > > > On Wed, Feb 27, 2013 at 10:39 AM, Rob Warren <[email protected]>wrote: > >> >> Mikel, >> >> >> On 27-Feb-13, at 1:52 PM, Mikel Maron wrote: >> >> for a date range: >>> base-uri/date_start/date_end/**zoom/x/y.img >>> >> >> Makes sense. >> >> for a moment: >>> base-uri/date/zoom/x/y.img >>> >> >> By 'moment', I take it you mean day? Or are you thinking of "Historic" >> values? >> >> As long as this works on a relation, I'd be happy with it. >> >> best, >> rhw >> >> >> >> ______________________________**_________________ >> Historic mailing list >> [email protected] >> http://lists.openstreetmap.**org/listinfo/historic<http://lists.openstreetmap.org/listinfo/historic> >> > > > > -- > Jeff Meyer > Global World History Atlas > www.gwhat.org > [email protected] > 206-676-2347 > <http://www.openstreetmap.org/user/jeffmeyer> osm: Historical > OSM<http://wiki.openstreetmap.org/wiki/Historical_OSM> > / my OSM user page <http://www.openstreetmap.org/user/jeffmeyer> > t: @GWHAThistory <https://twitter.com/GWHAThistory> > f: GWHAThistory <https://www.facebook.com/GWHAThistory> > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > <http://lists.openstreetmap.org/pipermail/historic/attachments/20130227/0b632e38/attachment-0001.html> > > ------------------------------ > > Message: 4 > Date: Wed, 27 Feb 2013 13:13:08 -0800 (PST) > From: Mikel Maron <[email protected]> > To: Jeff Meyer <[email protected]>, Rob Warren > <[email protected]> > Cc: "[email protected]" <[email protected]> > Subject: Re: [Historic] Dating features (was: Introduction & > questions) > Message-ID: > <[email protected]> > Content-Type: text/plain; charset="iso-8859-1" > >> base-uri/YYYY/zoom/x/y.img >>? >> Should this return start, middle, or end of YYYY, or should we enforce >> YYYY-MM-DD at a minimum? > ? > Usual way I've seen this dealt with is to assume "start" of the year (or > month or day, or whatever fidelity of the given input date). > >>?If the year is -2000 BCE, how should we express that? > > -2000? > http://en.wikipedia.org/wiki/ISO_8601#Years > > > It might actually need to be -1999, not so smooth. > >>?I'm fine with any answer, but we should pick one and use it until it breaks. >>Sound ok? I'm not smart enough to suggest the answer... > > Yea, my suggestion too. This might not be perfect, but we can adjust once we > have the basic flow going. > As you can see, I'm smart enough to use wikipedia only ;) >> By 'moment', I take it you mean day? Or are you thinking of "Historic" >> values? > > ? > Sorry, I was simply meant a single datetime value. > >>?As long as this works on a relation, I'd be happy with it. > > That would all depend on how the relation is processed by osm2pgsql. > > -Mikel > > * Mikel Maron *?+14152835207 @mikel s:mikelmaron > > >>________________________________ >> From: Jeff Meyer <[email protected]> >>To: Rob Warren <[email protected]> >>Cc: Mikel Maron <[email protected]>; "[email protected]" >><[email protected]> >>Sent: Wednesday, February 27, 2013 2:07 PM >>Subject: Re: [Historic] Dating features (was: Introduction & questions) >> >> >>So, let's test two cases: >> >> >>* Time Pixels: >> >> >>base-uri/YYYY/zoom/x/y.img >> >> >>Should this return start, middle, or end of YYYY, or should we enforce >>YYYY-MM-DD at a minimum? >> >> >>* Web Mercator of Time: >> >> >>If the year is -2000 BCE, how should we express that? >> >> >>I'm fine with any answer, but we should pick one and use it until it breaks. >>Sound ok? I'm not smart enough to suggest the answer... >> >> >>- Jeff >> >> >> >> >> >>On Wed, Feb 27, 2013 at 10:39 AM, Rob Warren <[email protected]> >>wrote: >> >> >>>Mikel, >>> >>> >>>On 27-Feb-13, at 1:52 PM, Mikel Maron wrote: >>> >>> >>>for a date range: >>>>base-uri/date_start/date_end/zoom/x/y.img >>>> >>> > Makes sense. >>> >>> >>>for a moment: >>>>base-uri/date/zoom/x/y.img >>>> >>>By 'moment', I take it you mean day? Or are you thinking of "Historic" >>>values? >>> >>>As long as this works on a relation, I'd be happy with it. >>> >>>best, >>>rhw >>> >>> >>> >>>_______________________________________________ >>>Historic mailing list >>>[email protected] >>>http://lists.openstreetmap.org/listinfo/historic >>> >> >> >> >>-- >>Jeff Meyer >>Global World History Atlas >>www.gwhat.org >>[email protected] >>206-676-2347 >>?osm:?Historical OSM?/?my OSM user page >>?t:?@GWHAThistory >>?f:?GWHAThistory >> >> >> >> >> >> >> > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > <http://lists.openstreetmap.org/pipermail/historic/attachments/20130227/271d8a11/attachment.html> > > ------------------------------ > > _______________________________________________ > Historic mailing list > [email protected] > http://lists.openstreetmap.org/listinfo/historic > > > End of Historic Digest, Vol 7, Issue 6 > ************************************** _______________________________________________ Historic mailing list [email protected] http://lists.openstreetmap.org/listinfo/historic
