[OSM-talk] talk Digest, Vol 32, Issue 72

[lewispusey]

Nothing most closely represents whats happening on the ground, its an 
abstraction. If your going to abstract the entities that wind up being the 
most logical for any number of reasons are lines, ponts and polygons. Ways 
are a further abstraction that cause headaches further down the line, 
especially when they occlude an underlying seg line polygon ordering. ASCII 
txt ==> is in itself an attempt at a linear representation. Just because 
industry standard GIS mapping software that can remain nameless has 
implemented a rigorous type of ordering of that nature was because that 
winds up being the logical foundation no matter how many time you argue in a 
"linear" circle. What would drive the definition of "ways", ground truth? 
how would you define ground truth, what types of abstraction?
    With forking ways what type of scripts do you want to develope to I.D. a 
problem: with that type of definition where a way is any type of branchy 
thingy is a "way" hence no problem  logical  analysis becomes impossible I 
expect.
Lewis
----- Original Message ----- 
From: <talk-request at openstreetmap.org>
To: <talk at openstreetmap.org>
Sent: Thursday, April 26, 2007 4:56 AM
Subject: talk Digest, Vol 32, Issue 72


> Send talk mailing list submissions to
> talk at openstreetmap.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
> or, via email, send a message with subject or body 'help' to
> talk-request at openstreetmap.org
>
> You can reach the person managing the list at
> talk-owner at openstreetmap.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of talk digest..."
>
>
> Today's Topics:
>
>   1. maplint vs. uncontinuous streets? (Ulf Lamping)
>   2. Re: 12am Freeze (Grant Slater)
>   3. Re: 12am Freeze (Dirk-L?der Kreie)
>   4. Re: maplint vs. uncontinuous streets? (D Tucny)
>   5. Re: Dyce airport (Lester Caine)
>   6. Re: Being Constructive (Joerg Ostertag (OSM Munich/Germany))
>   7. Re: 12am Freeze (Andy Robinson)
>   8. Re: maplint vs. uncontinuous streets? (Andy Allan)
>   9. Re: osmarender slippy map rendering (Lars Aronsson)
>  10. Re: maplint vs. uncontinuous streets? (80n)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 26 Apr 2007 02:03:00 +0200
> From: Ulf Lamping <ulf.lamping at web.de>
> Subject: [OSM-talk] maplint vs. uncontinuous streets?
> To: OSM-Talk <talk at openstreetmap.org>
> Message-ID: <462FEC34.8070607 at web.de>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Hi List!
>
> Using the maplint layer, e.g. from
> http://www.informationfreeway.org/?lat=6349372.81236&lon=1235169.55894&zoom=16&layers=00B0
>
> helped me a lot to fix obvious bugs in the way I've mapped various 
> streets.
>
> However, I've noticed:
>
> a) uncontinuous ways marked as "unordered" - ways simply lacking
> "missing" segments "in between"
> b) ways that "forks" - are spreading "subways" (forking) are also marked
> as "unordered"
>
> Are there ways to tag such ways right that I don't see?
>
> Regards, ULFL
>
>
>
> ------------------------------
>
> Message: 2
> Date: Thu, 26 Apr 2007 01:05:52 +0100
> From: Grant Slater <openstreetmap at firefishy.com>
> Subject: Re: [OSM-talk] 12am Freeze
> To: robert.wyatt at mail.utexas.edu, OSM <talk at openstreetmap.org>
> Message-ID: <462FECE0.2040204 at firefishy.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Robert T Wyatt wrote:
>> Out of curiosity, which 12 midnight are we talking about? UTC? I'll
>> endeavor to avoid it. :-)
>>
>>
>
> British Summer Time
>
> UTC+1
>
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 26 Apr 2007 03:22:50 +0200
> From: Dirk-L?der Kreie <osm-list at deelkar.net>
> Subject: Re: [OSM-talk] 12am Freeze
> To: Tom Hughes <tom at compton.nu>
> Cc: talk at openstreetmap.org
> Message-ID: <462FFEEA.4080807 at deelkar.net>
> Content-Type: text/plain; charset=ISO-8859-1
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Tom Hughes schrieb:
>
>> Well it's 00:35 now and the DB is still non-responsive, as it has
>> been most nights recently. Maybe they are taking a bit more than
>> five minutes now?
>>
>> Tom
>
> in the last few days it has taken up to an hour to recover from a
> sort-of backlog that builds up around that time. not sure this is
> explainable by the statsrun alone.
>
> - --
>
> Dirk-L?der "Deelkar" Kreie
> Bremen - 53.0952?N 8.8652?E
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFGL/7qFUbODdpRVDwRAjWiAJ9Z9HmudPYcaT72C6x8a2brCkgOAwCfRkgP
> EAKvy9arKoFbD2XOQfH01UY=
> =B8Bh
> -----END PGP SIGNATURE-----
>
>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 26 Apr 2007 10:50:30 +0800
> From: "D Tucny" <d at tucny.com>
> Subject: Re: [OSM-talk] maplint vs. uncontinuous streets?
> To: "Ulf Lamping" <ulf.lamping at web.de>
> Cc: OSM-Talk <talk at openstreetmap.org>
> Message-ID:
> <1c1b6e840704251950h23f90cbbv98cf592c467848ef at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> On 26/04/07, Ulf Lamping <ulf.lamping at web.de> wrote:
>>
>> Hi List!
>>
>> Using the maplint layer, e.g. from
>>
>> http://www.informationfreeway.org/?lat=6349372.81236&lon=1235169.55894&zoom=16&layers=00B0
>>
>> helped me a lot to fix obvious bugs in the way I've mapped various
>> streets.
>>
>> However, I've noticed:
>>
>> a) uncontinuous ways marked as "unordered" - ways simply lacking
>> "missing" segments "in between"
>> b) ways that "forks" - are spreading "subways" (forking) are also marked
>> as "unordered"
>>
>> Are there ways to tag such ways right that I don't see?
>
>
> The correct way to tag these is as multiple ways... A way should just be a
> continuous string of ordered segments in a single direction... This can be
> less than ideal and is not most intuitive, however, the super-ways
> proposal/discussion is aimed at providing a way of linking related
> 'sub-ways' into logical groups that share attributes...
>
> d
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: 
> http://lists.openstreetmap.org/pipermail/talk/attachments/20070426/18ee6abc/attachment-0001.htm
>
> ------------------------------
>
> Message: 5
> Date: Thu, 26 Apr 2007 06:40:41 +0100
> From: Lester Caine <lester at lsces.co.uk>
> Subject: Re: [OSM-talk] Dyce airport
> To: "talk at openstreetmap.org >> OSM Talk" <talk at openstreetmap.org>
> Message-ID: <46303B59.5000706 at lsces.co.uk>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Frederik Ramm wrote:
>>>> Dyce airport seems to be exactly 1 degree to the east of it's true
>>>> location:
>> [...]
>>>> and hence ends up in the middle of the ocean.
>>
>>> Of cause the fact that it's not called Dyce Airport would be worth
>>> correcting as well ;)
>>> It was Aberdeen Airport when I went through last night :)
>>
>> What, on a seaplane? ;-)
>
> I was going to make some comment about the taxi fare being cheap if that 
> was
> how far we went :)
>
> -- 
> Lester Caine - G8HFL
> -----------------------------
> Contact - http://home.lsces.co.uk/lsces/wiki/?page=contact
> L.S.Caine Electronic Services - http://home.lsces.co.uk
> MEDW - http://home.lsces.co.uk/ModelEngineersDigitalWorkshop/
> Firebird Foundation Inc. - http://www.firebirdsql.org/index.php
>
>
>
> ------------------------------
>
> Message: 6
> Date: Thu, 26 Apr 2007 08:04:59 +0200
> From: "Joerg Ostertag (OSM Munich/Germany)"
> <openstreetmap at ostertag.name>
> Subject: Re: [OSM-talk] Being Constructive
> To: talk at openstreetmap.org
> Message-ID: <200704260805.00028.openstreetmap at ostertag.name>
> Content-Type: text/plain;  charset="iso-8859-1"
>
> On Wednesday 25 April 2007 23:11, Schuyler Erle wrote:
>> * On 25-Apr-2007 at 10:29AM PDT, SteveC said:
>> > >   Clever chap, that Jesus fellow. (ref Matthew 7:12 - yes, I know he
>> > >   wasn't the first to say it!)
>> >
>> > Doesn't fit with game theory though :-O
>> >
>> > http://en.wikipedia.org/wiki/Prisoner%
>> > 27s_dillema#The_iterated_prisoner.27s_dilemma
>>
>> On the other hand, the social dynamics of OSM development (or any
>> similar distributed, collaborative endeavor) are substantially more
>> complex than the iterated Prisoner's Dilemma...
>>
>> I would prefer to stay away from ethical platitudes or game theory,
>> and simply and with all humility encourage everyone to do their best
>> to put the welfare of the project ahead of personal satisfaction. :)
>
> Yes, Please !!
>
>
> -- 
> J?rg (Germany, Munich)
>
> http://www.ostertag.name/
> irc://irc.oftc.net/#osm
> Tel.: +49 89 420950304
> Skype: JoergOstertag
>
>
>
> ------------------------------
>
> Message: 7
> Date: Thu, 26 Apr 2007 09:16:34 +0100
> From: "Andy Robinson" <Andy_J_Robinson at blueyonder.co.uk>
> Subject: Re: [OSM-talk] 12am Freeze
> To: 'Dirk-L?der Kreie' <osm-list at deelkar.net>, "'Tom Hughes'"
> <tom at compton.nu>
> Cc: talk at openstreetmap.org
> Message-ID:
> <!&!AAAAAAAAAAAuAAAAAAAAAOKaD4mR3JBOrEpRon92nMgBANp/H2q5kHFIvKMsnZiQaZAAAAABxJAAABAAAACT5amo6H2vS7Wt+zpHOiWyAQAAAAA=@blueyonder.co.uk>
>
> Content-Type: text/plain; charset="iso-8859-1"
>
> The stats run over the last week has on some nights actually been taking a
> lot longer than the typical 5 or so minutes. Here's the resent list from
> stats I've saved. Last nights run was particularly long.
>
> Date Mins
> 26-Apr 31.42
> 25-Apr 5.58
> 24-Apr 14.49
> 23-Apr 26.26
> 22-Apr 10.56
> 21-Apr 13.66
> 20-Apr 4.77
> 19-Apr 6.39
> 18-Apr 4.49
> 13-Apr 4.65
> 12-Apr 3.81
> 11-Apr 6.21
> 10-Apr 5.75
> 08-Apr 5.98
> 05-Apr 6.68
> 04-Apr 6.02
> 03-Apr 5.85
>
> Cheers
>
> Andy
>
> Andy Robinson
> Andy_J_Robinson at blueyonder.co.uk
>>-----Original Message-----
>>From: talk-bounces at openstreetmap.org [mailto:talk-
>>bounces at openstreetmap.org] On Behalf Of Dirk-L?der Kreie
>>Sent: 26 April 2007 2:23 AM
>>To: Tom Hughes
>>Cc: talk at openstreetmap.org
>>Subject: Re: [OSM-talk] 12am Freeze
>>
>>-----BEGIN PGP SIGNED MESSAGE-----
>>Hash: SHA1
>>
>>Tom Hughes schrieb:
>>
>>> Well it's 00:35 now and the DB is still non-responsive, as it has
>>> been most nights recently. Maybe they are taking a bit more than
>>> five minutes now?
>>>
>>> Tom
>>
>>in the last few days it has taken up to an hour to recover from a
>>sort-of backlog that builds up around that time. not sure this is
>>explainable by the statsrun alone.
>>
>>- --
>>
>>Dirk-L?der "Deelkar" Kreie
>>Bremen - 53.0952?N 8.8652?E
>>
>>-----BEGIN PGP SIGNATURE-----
>>Version: GnuPG v1.4.6 (GNU/Linux)
>>Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>>
>>iD8DBQFGL/7qFUbODdpRVDwRAjWiAJ9Z9HmudPYcaT72C6x8a2brCkgOAwCfRkgP
>>EAKvy9arKoFbD2XOQfH01UY=
>>=B8Bh
>>-----END PGP SIGNATURE-----
>>
>>_______________________________________________
>>talk mailing list
>>talk at openstreetmap.org
>>http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>
>
>
>
>
> ------------------------------
>
> Message: 8
> Date: Thu, 26 Apr 2007 09:33:38 +0100
> From: "Andy Allan" <gravitystorm at gmail.com>
> Subject: Re: [OSM-talk] maplint vs. uncontinuous streets?
> To: "D Tucny" <d at tucny.com>
> Cc: OSM-Talk <talk at openstreetmap.org>
> Message-ID:
> <c4193f8c0704260133o9de9557j4a97c3da11ff16a4 at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> On 4/26/07, D Tucny <d at tucny.com> wrote:
>> The correct way to tag these is as multiple ways... A way should just be 
>> a
>> continuous string of ordered segments in a single direction...
>
> You know, I still disagree with this. I don't see why Y shaped streets
> can't just be a single way. I realise that it's much trickier for
> renderers to handle, but it isn't *incorrect* to be put into the
> database like this, since it most closely represents what is happening
> on the ground. Even if we had superways (which it's worth remembering
> that we don't), splitting this into two ways still has no merit based
> on the facts on the ground. (Superways would be useful for other
> stuff, but they aren't a universal screwdriver.)
>
> I'll continue to use the flexibility offered by the current data
> model, until either the API is changed to prevent it or future
> developments give me a better mechanism. And along those lines, can
> people please refrain from saying that such techniques I and others
> use are "wrong", "incorrect" and so on, and more truthfully suggest
> that multiple ways are "suggested", "a convention" or even
> "preferred", if that's what you want to say.
>
> Andy
>
>
>
> ------------------------------
>
> Message: 9
> Date: Thu, 26 Apr 2007 10:42:23 +0200 (CEST)
> From: Lars Aronsson <lars at aronsson.se>
> Subject: Re: [OSM-talk] osmarender slippy map rendering
> To: OSM <talk at openstreetmap.org>
> Message-ID: <Pine.LNX.4.64.0704261029480.15069 at localhost.localdomain>
> Content-Type: TEXT/PLAIN; charset=utf-8
>
> 80n wrote:
>
>> This is an intentional style.  As you zoom in place-names get
>> bigger but also become semi-transparent.  At some level they
>> become invisible, ideally when the extent of the place is larger
>> than the viewport.
>
> I like this, but there are two things to wish for:
>
> 1. The current definition of a place=city is one with more than
> 100,000 inhabitants.  This Malm?, Helsinki, Cambridge, Oxford and
> a lot of places.  But we should need another level for places
> bigger than 1 million inhabitants like Copenhagen, Riga, London
> and Berlin.  The current city names look small for these places.
>
> 2. Tiles at home renders place=city as grey text at zoom=12, which
> means it should be black at zoom=11.  But t at h doesn't go that far.
> Instead, zoom=11 is created by lowzoom.pl by scaling down the
> zoom=12 image.  The result is that at zoom=11, =10 and =9 you see
> the town names (population 10,000 - 100,000) a lot better than
> city names.
>
> In combination, these wishes suggest that a place=megacity should
> be printed with a very large font, spanning very many tiles.
> This is of course problematic with respect to performance.
> Changing the megacity name would cause many tiles to be rendered.
> And rendering one tile would require a search of a large
> neighborhood (outside the current tile) for any megacities, parts
> of the names of which could touch this tile.  Rendering megacity
> names would also only affect zoom levels 7--11, which today aren't
> covered by the proper tiles at home, but by lowzoom.pl.  But perhaps
> a list of megacities (there really aren't that many in this world)
> could be extracted from the weekly planet.osm and supplied in svn
> as part of lowzoom.pl ?
>
> By the way, how are place=country names treated today?
>
>
> -- 
>  Lars Aronsson (lars at aronsson.se)
>  Aronsson Datateknik - http://aronsson.se
>
>
>
> ------------------------------
>
> Message: 10
> Date: Thu, 26 Apr 2007 09:56:26 +0100
> From: 80n <80n80n at gmail.com>
> Subject: Re: [OSM-talk] maplint vs. uncontinuous streets?
> To: "Andy Allan" <gravitystorm at gmail.com>
> Cc: OSM-Talk <talk at openstreetmap.org>
> Message-ID:
> <8fcd02310704260156y6bc3749eq1fccfb7959b12457 at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Without a doubt, if/when we get superways, then it will be necessary to
> convert non-linear and non-contiguous ways into simple linear ways, so
> there's no harm in continuing to do this.
>
> For the moment the only real benefit of not making branching ways is that
> you get better street name rendering.
>
> In Osmarender street names are just written along a way as if it was a
> single continuous path.  This can result in all kinds of jumbled up and
> fragmented labelling.
>
> This is fixable and, one day, will be.  Osmarender already breaks complex
> ways up into a series of sub-paths for some purposes, but not yet for
> rendering names.   The street name should be written once on each of the
> sub-paths but only if the name is shorter than the length of the sub-path.
>
> 80n
>
> On 4/26/07, Andy Allan <gravitystorm at gmail.com> wrote:
>>
>> On 4/26/07, D Tucny <d at tucny.com> wrote:
>> > The correct way to tag these is as multiple ways... A way should just 
>> > be
>> a
>> > continuous string of ordered segments in a single direction...
>>
>> You know, I still disagree with this. I don't see why Y shaped streets
>> can't just be a single way. I realise that it's much trickier for
>> renderers to handle, but it isn't *incorrect* to be put into the
>> database like this, since it most closely represents what is happening
>> on the ground. Even if we had superways (which it's worth remembering
>> that we don't), splitting this into two ways still has no merit based
>> on the facts on the ground. (Superways would be useful for other
>> stuff, but they aren't a universal screwdriver.)
>>
>> I'll continue to use the flexibility offered by the current data
>> model, until either the API is changed to prevent it or future
>> developments give me a better mechanism. And along those lines, can
>> people please refrain from saying that such techniques I and others
>> use are "wrong", "incorrect" and so on, and more truthfully suggest
>> that multiple ways are "suggested", "a convention" or even
>> "preferred", if that's what you want to say.
>>
>> Andy
>>
>> _______________________________________________
>> talk mailing list
>> talk at openstreetmap.org
>> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: 
> http://lists.openstreetmap.org/pipermail/talk/attachments/20070426/46a86dc4/attachment.htm
>
> ------------------------------
>
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>
>
> End of talk Digest, Vol 32, Issue 72
> ************************************
>