[pgrouting-dev] shortest_path_shooting_star

[Daniel Kastl]

>
>
> On Saturday, June 9, 2012, Stephen Woodbridge wrote:
>
>> Hi all,
>>
>> In all likelihood, we will abandon the shooting star code unless someone
>> want to try and figure out the implementation and try to fix the code.
>> Currently the TRSP code does everything that shooting star does, so there
>> is not loss of functionality AND it is significantly faster something like
>> 2+ times faster, and it does not have any bugs in it.
>>
>
Hi Steve,

I agree that TRSP is a good alternative, and the only issue now is that
it's not part of the latest release and documentation isn't on the website.
What I like on Shooting Star is, that it routes from road link to road link
instead of point to point. But it's right that it should be fixed first.

I'm happy to add some TRSP documentation/tutorial to the pgRouting website.
But it would help me if you could post a rough draft version or point me to
the right email threads. I will then review it and bring it into a nice
format. I also have a bit limited spare time right now.



>
>> On thought I had was that we could wire in the TRSP algorithm to the
>> Shooting Star function signature, but give that we would like to
>> rationalize all the function APIs potentially in 2.0 if we can ever get the
>> time to do that release, I'm not sure what the value of having two function
>> signatures, trsp and shooting star both using the same code.
>>
>
I'm not sure this is a good idea or not, but in fact all shortest path
algorithms do more or less the same. So one function might be better,
called "PGR_shortest_path" for example, and the library either is smart
enough to choose the algorithms depending on the passed parameters (if each
algorithm has different ones) or the first parameter could specify the
algorithm.

It's difficult to concentrate on a new release without sponsorship or
funding, I think. But you can be sure, that in case there is a project
opportunity that would allow to work on a new release this would be high
priority for Georepublic.

Personally (as I'm mainly using Ubuntu) I would like to wait a bit until
Debian/Ubuntu GIS will have PostGIS 2.0 packaged. I guess that PostGIS 2.x
and PostreSQL 9.1 will soon become a standard and pgRouting 2.0 would be
better a PostgreSQL extension and make use of PostGIS concave hull function
to be able to drop CGAL dependency.

Daniel



>
>> Anton was the last developer the tried to work on it and he is gone. We
>> would be happy to take patches if someone can figure out the Boost code.
>>
>> Thanks,
>>  -Steve
>>
>> On 6/8/2012 7:15 PM, Sanak wrote:
>>
>>> Hi Steve,
>>> I have not enough knowledge about Shooting Star algorithm and Boost
>>> Graph Library,
>>> but I think that the bug occurred in "core / src /
>>> shooting_star_boost_wrapper.**cpp" after 1.0.3 release(2008/09/04).
>>> [History for core-src-shooting_star_boost_**wrapper.cpp -
>>> pgRouting-pgrouting]
>>> https://github.com/pgRouting/**pgrouting/commits/master/core/**
>>> src/shooting_star_boost_**wrapper.cpp<https://github.com/pgRouting/pgrouting/commits/master/core/src/shooting_star_boost_wrapper.cpp>
>>> And, old issue #190 may be related.
>>> [#190 (First edge and last edge don't respect the one way restriction) –
>>> pgRouting]
>>> ftp://ftp.remotesensing.org/**pgrouting/forum/pgrouting.**
>>> postlbs.org/ticket/190.html<ftp://ftp.remotesensing.org/pgrouting/forum/pgrouting.postlbs.org/ticket/190.html>
>>> Isn't there any person who have enough knowledge about Shooting Star
>>> argorithm and Boost Graph Library?
>>> Regards,
>>> 2012/6/9 Steve Horn <steve at stevehorn.cc <mailto:steve at stevehorn.cc>>
>>>
>>>    Was just wondering if anyone has had any time to address the bug in
>>>    the shortest_path_shooting_star algorithm where it ignores one way
>>>    street designations?
>>>
>>>    --
>>>    Steve Horn
>>>
>>>
>>>    ______________________________**_________________
>>>    pgrouting-dev mailing list
>>>    pgrouting-dev at lists.osgeo.org <mailto:pgrouting-dev at lists.osgeo.org>
>>>    http://lists.osgeo.org/**mailman/listinfo/pgrouting-dev<http://lists.osgeo.org/mailman/listinfo/pgrouting-dev>
>>>
>>>
>>>
>>>
>>> ______________________________**_________________
>>> pgrouting-dev mailing list
>>> pgrouting-dev at lists.osgeo.org
>>> http://lists.osgeo.org/**mailman/listinfo/pgrouting-dev<http://lists.osgeo.org/mailman/listinfo/pgrouting-dev>
>>>
>>
>> ______________________________**_________________
>> pgrouting-dev mailing list
>> pgrouting-dev at lists.osgeo.org
>> http://lists.osgeo.org/**mailman/listinfo/pgrouting-dev<http://lists.osgeo.org/mailman/listinfo/pgrouting-dev>
>>
>
>
> --
> Steve Horn
>
>
> _______________________________________________
> pgrouting-dev mailing list
> pgrouting-dev at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/pgrouting-dev
>
>


-- 
Georepublic UG & Georepublic Japan
eMail: daniel.kastl at georepublic.de
Web: http://georepublic.de
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/pgrouting-dev/attachments/20120610/45a835f6/attachment-0001.html>