[OSM-dev] Multiple Overpass dispatchers/databases on the same machine?

2012-08-02 Thread Samat K Jain
Is there a way to install multiple instances of Overpass on the same
machine?

I'd like to have an API endpoint for the world (w/out metadata), and
then another API endpoint for local use (w/ metadata). E.g.
http://example.com/world/interpreter?data=… and
http://example.com/local/interpreter?=data=….

I'm mostly everything figured out (e.g. database imports) except running
multiple instances of the dispatcher. And if I got that running, I
haven't figured out how to get the CGI scripts to communicate to that
other dispatcher.

Communication between CGI script and dispatcher happens over a shared
memory segment (in /dev/shm), but the name of these appear to be
hardcoded in src/overpass_api/core/settings.cc. Is there a way to
override these so I can run multiple copies of Overpass without having
to compile different copies?

-- 
Samat K Jain http://samat.org/ ▪ GPG: 0x4A456FBA

-- 
Samat K Jain http://samat.org/ ▪ GPG: 0x4A456FBA

___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] Merkaator port to Android

2011-12-22 Thread Samat K Jain
On Friday, December 02, 2011 12:56:02 PM Jaak Laineste wrote:
 Hi,
  as you may now, during GSoC QGIS was ported to Android tablets
 (http://hub.qgis.org/projects/quantum-gis/wiki/QGIS_Mobile_GSoC_2011).
 Merkaator has technically same base elements: Qt, GEOS, PROJ.4 etc, so
 based on this experience porting of Merkaator could be also possible
 with much smaller fuss. It should be even smaller work than getting
 JOSM working under Android.
 
  Question: do you know anyone who really would need and use it?

FYI, there is a Mobile Merkaartor in it's source repo:

  https://gitorious.org/merkaartor/main/trees/master/mobilemerk

No idea on it's status, though. I remember seeing something about a Merkaartor 
port to Lighthouse (Qt for Android) but can't find it now.

-- 
Samat K Jain http://samat.org/ ▪ GPG: 0x4A456FBA

I don't know with what weapons World War III will be fought, but World War IV 
will be fought with sticks and stones.
— Albert Einstein (267)

This e-mail is: [ ] bloggable [x] ask first [ ] private


___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] ogr2osm.py error message

2011-03-11 Thread Samat K Jain
 here is what's around line 594 in the script:
 
 print
 
 print Simplifying line segments
 
 for line in lineSegments.values():
 
 #print line
 
 for segmentID in line:# No need to check the last segment, it could
 not be simplyfied
 
 #print segmentID
 
 #print segmentNodes[segmentID]
 
 for nodeID in segmentNodes[segmentID]:
 
 simplifyNode(nodeID)
 
 #simplifyNode(segmentNodes[segmentID][0])# last node in
 segment
 
 
 Thanks for any help!

I've seen this a couple times before too. I haven't cared enough to investigate 
what's exactly going on…

But to make the error go away, change that 'for' statement on line 594 to:

  for nodeID in segmentNodes.get(segmentID, []):

The side effects depend on your data. Make sure to check for problems.

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

Adults are just obsolete children, and too hell with them.
-- Theodor Seuss Geisel (Dr. Seuss) (362)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] Let's put more wikipedia-tags in OSM WAS Re: Some Questions about the Collaboration of OpenStreetMap and Wikipedia

2011-02-24 Thread Samat K Jain
On Thursday, February 24, 2011 10:16:22 AM M∡rtin Koppenhoefer wrote:
 I invite everybody to look in his region for features (not only
 place-features) to see, if they are already linked to wikipedia. In
 case they are not I suggest to add a tag in the form:
 wikipedia=language-code:article-name
 as suggested here:
 http://wiki.openstreetmap.org/wiki/Key:wikipedia
 
 e.g.
 wikipedia=en:London
 or
 wikipedia=it:Hong_Kong

AFAIK, do *NOT* include underscores in the key name… as is mentioned on the 
aforementioned wiki page.

E.g. that should be:

wikipedia=it:Hong Kong

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

A fractal is by definition a set for which the Hausdorff Besicovitch dimension 
strictly exceeds the topological dimension.
-- Mandelbrot, The Fractal Geometry of Nature (402)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] OSM formats optimised for client-side vector rendering?

2011-02-17 Thread Samat K Jain
On Thursday, February 17, 2011 03:20:30 AM Nick Whitelegg wrote:
 However I'm wondering if there is any consensus on a standard OSM data 
 format optimised for vector rendering. There seems to be the OSM Mobile 
 Binary Protocol though by the looks of things, that's only used in one 
 application: the Symbian app whereami. So does everyone just use their own 
 or are there any formats used by multiple apps?

Probably not useful to you, since it appears to be proprietary, but it's worth 
mentioning—the format used by MapDroyd/NavDroyd for Android is fantastic:

  http://www.mapdroyd.com/

Reasonable file sizes, and very quick/responsive vector rendering on even 
first-gen Android phones.

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

I'm not saying there should be a capital punishment for stupidity, but why 
don't we just take the safety labels off of everything and let the problem 
solve itself?
-- None (680)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] OpenID for OpenStreetMap?

2011-02-15 Thread Samat K Jain
On Friday, February 11, 2011 12:10:42 AM Frederik Ramm wrote:
 The forum and help system can already use the standard OSM login. The 
 only thing where you need an extra login is the wiki.

Thanks for the clarification.

 The area where I see most potential benefit for OpenID is an use case 
 that has popped up on these lists at least twice: Third-party system N 
 has an existing user base who have logins at the site, and wants these 
 users to be able to make certain edits in OSM. Since N believes that 
 their users cannot be asked to go to the OSM site, sign up there, and 
 then use OAuth to permit A's application to edit the map, N is tempted 
 to use one single account to submit all edits under (to make it easier 
 for users).
 
 With a well-integrated OpenID/OAuth system, which we don't yet have, N 
 could act as an OpenID provider against OSM, redirect their users to OSM 
 once, where the user would only have to fill in a single web page that 
 says yes, I am User U and I want to create an OSM account and allow N 
 to submit edits, and they're all set.

That's a very good use case!

 Other than that - even if I had a Google account, which I don't, I 
 would not want Google to know when I log in to OSM. I think that if we 
 should allow OpenID logins, it is also our responsibility to inform the 
 user about the privacy they give away by using it.

I agree (I don't use Google as my OpenID)… but of all the things users need be 
educated about, I can't see this as being high on the list. They're already 
using Google, *the* juggernaut of user profiling on the Internet; whether they 
know that a user logs into OSM is inconsequential to everything else they've 
already given Google. Using that same logic, we could argue that we shouldn't 
allow people with @gmail.com addresses.

We don't, of course; the point is users have a choice and are free to make it.

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

All Bibles are man-made.
-- Thomas Edison (96)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] OpenID for OpenStreetMap?

2011-02-10 Thread Samat K Jain
On Thursday, February 10, 2011 05:02:27 AM Mike N wrote:
  Frankly though a bigger question is whether we want to. I've never been
  that convinced because OpenID just doesn't seem to be getting the
  necessary traction to really work.

That's a misleading way to think about it… with Google, Yahoo, etc acting as 
OpenID providers, almost _everyone_ on the Internet has an OpenID. I don't have 
numbers, but likely it is more than Facebook or Twitter. It *is* the most 
ubiquitious login system on the Internet.

But yes, many sites are dropping OpenID login support (jargon: fewer OpenID 
consumers). However, they're switching to Facebook Connect and Twitter Auth—is 
outsourcing authentication to corporate silos beneficial to OSM? OpenID, where 
people have a choice in outsourcing their auth to a corporate silo, remains the 
best option for an open internet.

I'm not sold on OpenID either - I know the pro arguments, but with 
 OpenID, if you're keylogged or phished, they have full access to all 
 accounts immediately.

We're drifting off-topic here… but centralizing logins into OpenID means that 
you can add more security to the login process, which benefits logins to all 
sites.

With standard username-passwords pairs, there's no universal way to improve the 
process.

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

The ability to procrastinate is what separates us from the machines.
-- Chris Gregory (12)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] OpenID for OpenStreetMap?

2011-02-10 Thread Samat K Jain
On Thursday, February 10, 2011 03:01:21 PM Matt Amos wrote:
 On Thu, Feb 10, 2011 at 8:09 PM, Serge Wroclawski emac...@gmail.com wrote:
  On Thu, Feb 10, 2011 at 2:24 PM, Matt Amos zerebub...@gmail.com wrote:
  i might be missing the point here, but are you suggesting that OSM
  would be an openID provider, consumer or both?
 
  I'm suggesting that instead of taking on a technology, we take on a
  set of requirements, or needs, and address them. And from there I'm
  saying that OpenID itself is a side issue.

The reason I focus on OpenID is that code for the rails port, the help site, 
and wiki (I don't know about other OSM web properties? the forum?) already 
exists.

 ok - so what's the problem, or requirement in corporate-speak, that
 we're trying to address here?

I wasn't very clear in my original e-mail, sorry… in short, numbered form:

1. Make it easier for users to sign up, and for existing users to login.
2. Eliminate extra, confusing, and unnecessary logins for other OSM websites 
(help, wiki, forum, etc).
3. A single sign on system—login into one OSM site, and you're logged into all 
(i.e. what Canonical's Launchpad does).

1 and 2, as I understand it, can be realized tomorrow by turning OpenID on—the 
devil, of course, is UX details. 3 will require a bit more thinking and 
programming, if it's an important problem to solve.

  That sounds like it makes non-web based editors hard to develop.
 
 nope. the user has to log into the website to authorize an OAuth token
 - what does it matter whether they log in via the usual username and
 password or OpenID? certainly it doesn't matter to the editor, even if
 it's doing some sort of ugly hack like JOSM's fully automatic mode.

OpenID does _nothing_ for authentication with 3rd party clients (i.e. 
editors). Bringing it up is confounding different issues. I should reiterate 
that I am no way suggesting that OpenID _replace_ OSM usernames/passwords, but 
I am suggesting it augment them.

It's worth pointing out, however, that Google has an OpenID/OAuth hybrid 
protocol:

  
http://googledataapis.blogspot.com/2009/01/bringing-openid-and-oauth-together.html

But this still aimed at the web only.

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

I fell asleep reading a dull book, and I dreamt that I was reading on, so I 
woke up from sheer boredom.
-- None (496)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


Re: [OSM-dev] OpenID for OpenStreetMap?

2011-02-10 Thread Samat K Jain
On Thursday, February 10, 2011 04:00:09 PM Tom Hughes wrote:
 No, actually it has not bit rotted at all - it is still working fine.
 
 Not only was it still working fine, but I have today merged the latest 
 code from the trunk and resolved the small number of merge conflicts 
 that resulted.

Neat! So, I guess… I need to convince as many as people as possible so we can 
get it deployed?

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

There is a theory which states that if ever anyone discovers exactly what the 
Universe is for and why it is here, it will instantly disappear and be replaced 
by something even more bizarre and inexplicable. There is another theory which 
states that this has already happened.
-- Douglas Adams, The Hitchhiker's Guide to the Galaxy (635)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev


[OSM-dev] OpenID for OpenStreetMap?

2011-02-09 Thread Samat K Jain
Recent discussions on IRC have been making me want to bring up—why does 
OpenStreetMap not yet support OpenID?

Apparently, OpenID is already (mostly) implemented in a dev install of the 
rails port:

  http://openid.dev.openstreetmap.org/

It doesn't appear to work anymore, however; at least I'm unable to login with 
my normal OSM credentials and associate my OpenID with them.

OpenStreetMap now has (at least) 3 sets of logins, just for the web:

 * OpenStreetMap.org
 * Help.OpenStreetMap.org, though this has an SSO mechanism
 * Wiki.OpenStreetMap.org

This situation is only going to get worse. OpenStreetMap supporting OpenID (at 
least, as an OpenID consumer) for its web properties alleviates this problem 
for the far majority who have an OpenID (use GMail? Yahoo Mail? you've an 
OpenID! You can also setup your own trivially[1]).

Just to reiterate the problem OpenID will solve: instead of having multiple 
username/password pairs for each of these services, users will only need to use 
their single OpenID (or many, should they want to associate multiple OpenIDs 
with their account, which is a feature the best OpenID consumers have).

I should make it clear: it's unlikely that OpenID replace the normal 
OpenStreetMap.org credentials that we already have. But it does alleviate the 
need for separate logins for all these other sites.

OpenID also provides a platform to build an SSO (single-sign on) system in the 
future E.g. see Ubuntu and how they tie Launchpad, the Community Wikis, etc 
together with a single username/password pair. Annoyingly, Launchpad doesn't 
let you associate your existing OpenID with their login system.

I haven't been able to find much previous discussion on the mailing lists, and 
the discussion on the wiki[2] is less discussion and more a few persons' 
opinions. I'd like to open up the debate again—especially when it's just a 
matter of merging code that's already written.

  [1]: http://intertwingly.net/blog/2007/01/03/OpenID-for-non-SuperUsers
  [2]: http://wiki.openstreetmap.org/wiki/Talk:Single_sign_on

Samat

-- 
Samat K Jain http://samat.org/ | GPG: 0x4A456FBA

For the world is divided, experience teaches, into two parts: things to be 
murdered, and things that would hinder the murder of things...
-- Grendel, John Gardner (60)


signature.asc
Description: This is a digitally signed message part.
___
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev