<div dir="ltr"><div>Thanks Steve will add a few lines to my proposal explaining the reason behind opting Astar.<br><br><br><br><br></div>Mukul<br></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Apr 25, 2013 at 8:55 PM, Stephen Woodbridge <span dir="ltr"><<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">On 4/25/2013 11:06 AM, Mukul priya wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
<br>
How much of the network do you have to explore using each algorithm?<br>
<br>
This will be reduced significantly while using Astar as we will have a<br>
fair bit of idea about the direction towards which we should proceed.<br>
<br>
How does this impact the number of grids you have to load?<br>
<br>
Once we have an idea about the direction towards which we should proceed<br>
, only those grids will be loaded. This will result in loading of less<br>
number of grids.<br>
<br>
How does this impact your proposal?<br>
The basic motivation behind the proposal is to make shortest path<br>
computation faster. Number of database querry will be reduced as we will<br>
be fetching lesser number of grids. This will have a positive effect on<br>
computation time.<br>
</blockquote>
<br></div>
Right, so in your proposal, you want to be clear that the benefit will be achieved using astar or another algorithm with a heuristic and not dijkstra. I understand that "shortest path" is a generic reference to all of these algorithms, but the only ones that will benefit from this approach will be ones with a heuristic that allow us to explore only a subset the overall bounding box of edges that we might otherwise use as input. So it is best to be clear on these points.<br>

<br>
-Steve<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">
On Thu, Apr 25, 2013 at 7:07 PM, Stephen Woodbridge<br></div><div class="im">
<<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.com</a> <mailto:<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.<u></u>com</a>>> wrote:<br>
<br>
    On 4/25/2013 12:26 AM, Mukul priya wrote:<br>
<br>
        Thanks Steve ,I am familar with astar ,It uses a heuristic to guide<br>
        itself to the destination, there was huge focus on it in our<br>
        game theory<br>
        course.<br>
<br>
        Although do point out the advantages it has in real life road<br>
        network<br>
        and its importance to my proposal.<br>
<br>
<br>
    How much of the network do you have to explore using each algorithm?<br>
    How does this impact the number of grids you have to load?<br>
    How does this impact your proposal?<br>
<br>
    -Steve<br>
<br>
        On Thu, Apr 25, 2013 at 7:43 AM, Stephen Woodbridge<br>
        <<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.com</a> <mailto:<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.<u></u>com</a>><br></div><div class="im">

        <mailto:<a href="mailto:woodbri@swoodbridge." target="_blank">woodbri@swoodbridge.</a>__<u></u>com<br>
        <mailto:<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.<u></u>com</a>>>> wrote:<br>
<br></div><div class="im">
             Hi Mukul,<br>
<br>
             This is a good write up. It largely mirrors out discussion. One<br>
             thing that needs to change in this is that dijkstra should be<br>
             replaced with astar.<br>
<br>
             Are you familiar with how these are different?<br></div>
        <a href="http://theory.stanford.edu/~____amitp/GameProgramming/____AStarComparison.html" target="_blank">http://theory.stanford.edu/~__<u></u>__amitp/GameProgramming/____<u></u>AStarComparison.html</a><br>
        <<a href="http://theory.stanford.edu/%7E__amitp/GameProgramming/__AStarComparison.html" target="_blank">http://theory.stanford.edu/%<u></u>7E__amitp/GameProgramming/__<u></u>AStarComparison.html</a>><br>
<br>
        <<a href="http://theory.stanford.edu/%__7Eamitp/GameProgramming/__AStarComparison.html" target="_blank">http://theory.stanford.edu/%_<u></u>_7Eamitp/GameProgramming/__<u></u>AStarComparison.html</a><div><div class="h5">
<br>
        <<a href="http://theory.stanford.edu/%7Eamitp/GameProgramming/AStarComparison.html" target="_blank">http://theory.stanford.edu/%<u></u>7Eamitp/GameProgramming/<u></u>AStarComparison.html</a>>><br>
<br>
<br>
             Do you see why? and specifically why it matter for this<br>
        project?<br>
<br>
             -Steve<br>
<br>
             On 4/24/2013 9:10 PM, Mukul priya wrote:<br>
<br>
                 ***Name*:Mukul Priya<br>
<br>
                 *Country:*India<br>
<br>
                 *School and degree*: International Institute Of Information<br>
<br>
                 Technology-Hyderabad ,<br>
<br>
                                                  B.Tech + Masters in<br>
        computer<br>
                 Science And<br>
                 Engineering<br>
<br>
                 *Email*:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a><br>
        <mailto:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a>> <mailto:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a><br>
        <mailto:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a>>><br>
                 <mailto:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a><br>
        <mailto:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a>> <mailto:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a><br>
        <mailto:<a href="mailto:mukul2047@gmail.com" target="_blank">mukul2047@gmail.com</a>>>><br>
<br>
<br>
                 *Phone*:+91 9885991061<br>
<br>
                 *OSGeo project(s*):pgRouting<br>
<br>
                 *Title:* A partitioned approach to on demand increment<br>
        graph<br>
                 assembly<br>
                 for pgRouting.<br>
<br>
                 *Describe your idea*<br>
                 *1. Introduction*<br>
<br>
<br>
                          pgRouting has been working towards providing<br>
        routing<br>
                 functionality on a PostGis/PostgreSql  Geo spatial<br>
<br>
                     database. It already has support for shortest path<br>
        algorithm<br>
                 like<br>
                 astar ,dijkstra , tdsp and trsp .But for a very<br>
<br>
                     large network, routing becomes time consuming. Network<br>
                 partitioning<br>
                 is one such way which can prove to<br>
<br>
                     be helpful in improving the overall time efficiency<br>
        of routing<br>
                 querries. The main idea  here is to first partition<br>
<br>
                     the whole network using a Quad tree approach and<br>
        when the<br>
                 shortest<br>
                 path is computed these partitions are<br>
<br>
                     loaded on demand. hence , there is an on demand<br>
        incremental<br>
                 graph<br>
                 generation.<br>
<br>
<br>
                          The project aims at designing and implementing a<br>
                 Shortest Path<br>
                 algorithm on an on demand incremental<br>
<br>
                       Graph using a network partitioning approach.<br>
<br>
<br>
<br>
                 *2. Background*<br>
<br>
<br>
                          Considering the present status where pgRouting has<br>
                 support for<br>
                 shortest path algorithm like astar etc.<br>
<br>
                    Looking at their implementation details we can<br>
        observe that<br>
                 the graph<br>
                 is configured dynamically for all<br>
<br>
                    of them before computation.My proposal will also be<br>
        on the<br>
                 same track<br>
                 and for very large networks<br>
<br>
                    where  the distance between source and destination<br>
        can be<br>
                 very large ,<br>
                 the response time will be<br>
<br>
                    significantly lesser and memory wise too it will be<br>
        lot more<br>
                 efficient. Presently they don't use any partitioning<br>
<br>
                    approach so it will prove to be a good addition to their<br>
                 support for<br>
                 shortest past algorithms.<br>
<br>
<br>
                 *3. The idea*<br>
<br>
<br>
                      There are two major components of my idea .<br>
<br>
<br>
                    *<br>
<br>
                      *Network Partitioning*<br>
<br>
<br>
                          For this part we can use a quad tree approach.<br>
        Say , we<br>
                 start<br>
                 with a square and a condition like maximum<br>
<br>
                          number of nodes that can reside in a square .<br>
        if the<br>
                 number of<br>
                 nodes in the square is greater than the max<br>
<br>
                          condition we further quarter it into four<br>
        squares and<br>
                 allot the<br>
                 nodes appropriately to each of them.All these<br>
<br>
                          squares can be called grids and they all will<br>
        be addressed<br>
                 uniquely using a grid cell number .Each node<br>
<br>
                          will be assigned a grid cell number based on<br>
        the grid<br>
                 they are<br>
                 lying inside.<br>
<br>
<br>
                                 We will have data structures to address<br>
        edges as<br>
                 they can<br>
                 remain contained in either one grid cell<br>
<br>
                           or span across a number of grid cells.the<br>
        idea is to<br>
                 first flag<br>
                 such edges and store the grid cell numbers<br>
<br>
                           of the grids that the edge crosses/intersects.<br>
<br>
<br>
                    *<br>
<br>
                      *On demand graph generation and Routing.*<br>
<br>
<br>
                                  The idea here is to first identify the<br>
        grid cell<br>
                 initially and then fetch the edges that are associated with<br>
<br>
                          that grid. These are the edges that will get<br>
        appended<br>
                 to the<br>
                 present graph and the graph will keep growing<br>
<br>
                          dynamically this way. we will have appropriate<br>
        database<br>
                 tables<br>
                 addressing the above issue such that<br>
<br>
                          we are able to fetch the required edges<br>
        quickly using  a<br>
                 database querry.<br>
<br>
<br>
                           The implementation details for the above are :<br>
<br>
                          we will first partition the whole graph using<br>
        the quad tree<br>
                 approach and each node/vertex will be assigned<br>
<br>
                          a grid cell number so we can have  database<br>
        table for the<br>
                 assigned nodes and edges like :<br>
<br>
                          CREATE TABLE vertex{<br>
<br>
                           id Node_id                               \\<br>
          unique<br>
                 for each<br>
                 node.<br>
<br>
                           cell grid_cell_number;               \\ the<br>
        grid in<br>
                 which the<br>
                 node lies.<br>
<br>
                            geometry;<br>
<br>
                            }<br>
<br>
<br>
                            CREATE TABLE edge  {<br>
<br>
                             int id ;<br>
<br>
                             Int node_a;<br>
<br>
                              int node_b;        \\ the connecting nodes<br>
<br>
                       //     we can have other parameters like<br>
        traversal cost<br>
                 and return<br>
                 cost.<br>
<br>
                            }<br>
<br>
                             using the tables we can then fetch the<br>
        edges form<br>
                 edge table<br>
                    using simple database querry once<br>
<br>
                    we are provided with the grid cell number very easily.<br>
<br>
                    In short the summary of the whole idea is :<br>
<br>
<br>
                 Step1: Fetch the start node , get the related Grid cell<br>
        number.<br>
<br>
<br>
                 Step 2: Increment the graph by fetching all the items<br>
        related to<br>
                 that grid.<br>
<br>
<br>
                 Step3 : Check for boundary nodes  ( this is done while<br>
        we are<br>
                 partitioning) or target node .<br>
<br>
                 Step 4: On hitting a boundary node<br>
<br>
                                  {<br>
<br>
                                    check if the connected grid is<br>
        loaded and<br>
                 continue if<br>
                 it is<br>
<br>
                                    or we extend the graph with the new<br>
        grid and<br>
                 continue<br>
                 with the routing;<br>
<br>
                                   }<br>
<br>
                              or On hitting the target Node<br>
<br>
                                {<br>
<br>
                                           stop;<br>
<br>
                                 }<br>
<br>
<br>
<br>
                 *4. Project plan*<br>
<br>
<br>
                        I will have about 12 weeks to implement the<br>
        project. The<br>
                 tentative<br>
                 schedule is as follows:<br>
<br>
                        Before 17th June : Get familiar with the development<br>
                 environment<br>
                 of pgRouting and test some demos.<br>
<br>
                        week 1-2: Discuss and Define the various data<br>
        structures<br>
                 and data<br>
                 table that will be required.Prepare the<br>
<br>
                                        Overall implementation<br>
          architecture.<br>
<br>
<br>
                        week 3-4-  Discuss and Implement network<br>
        partitioning<br>
                 using quad<br>
                 tree.<br>
<br>
                        week 5-8 -Start coding for on demand graph<br>
        generation and<br>
                 routing<br>
                 using Dijkstra. In between Prepare mid-term<br>
<br>
                                          report.<br>
<br>
<br>
                        week 9 - Integration with pgrouting .<br>
<br>
                        week 10-12- Testing and Bug fixing. Draft Final<br>
        report and<br>
                 documentation.<br>
<br>
<br>
                 *<br>
                    5. Future ideas / How can your idea be expanded? *<br>
<br>
<br>
<br>
                         It can be integrated with other shortest path<br>
        algorithms<br>
                 like<br>
                 astar etc. which pgrouitng provides.For time dependent<br>
<br>
                       shortest path computation, it will greatly reduce the<br>
                 updating cost<br>
                 as we will be updating the cost of only those edges<br>
<br>
                       that are contained in the grid cells that are loaded.<br>
<br>
                        In route nearest neigbour querries can also be<br>
                 implemented using<br>
                 the partitioning approach.Overall we can<br>
<br>
                        expand this approach to various other algorithms.<br>
<br>
                 *Explain how your SoC task would benefit the OSGeo member<br>
                 project and<br>
                 more generally the OSGeo Foundation as a whole:*<br>
<br>
<br>
                       The proposed idea will significantly improve the<br>
                 performance of<br>
                 shortest path finding algorithms.This will have a<br>
<br>
                     positive effect on the user base of pgRouting. For<br>
        various<br>
                 software<br>
                 paltforms or applications that are using this library ,<br>
<br>
                     the response time will significantly improve.<br>
<br>
                 *<br>
<br>
                 Please provide details of general computing experience:<br>
                 *<br>
<br>
<br>
                 I generally use ubuntu ( 12.04) for all my academic<br>
        purposes and<br>
                 windows<br>
                 7 for entertainment only. I am familiar with Fedora also.<br>
<br>
                 C++ is the language that i use most ( programming<br>
        purposes). For<br>
                 implementing various  course projects i have used C++<br>
        ,python<br>
                 ,Java and<br>
                 matlab.<br>
<br>
<br>
                 I have no experience with hardware. The only experience<br>
        that i<br>
                 had with<br>
                 networking was during my computer network course.<br>
<br>
                 *<br>
                 Please provide details of previous GIS experience:*<br>
<br>
<br>
<br>
                 I am a part of "Lab for Spatial Informatics" which is<br>
        the only<br>
                 lab in<br>
                 India devoted to GIS applications and learning. I am<br>
        familiar<br>
                 with and<br>
                 have used almost all the open source GIS platforms like<br>
        Quantum ,<br>
                 OpenJump, ILwis,Grass.  We have our own rendering<br>
        platform that was<br>
                 developed by our lab.(LSI viewer)<br>
<br>
                 *Please provide details of any previous involvement<br>
        with GIS<br>
                 programming<br>
                 and other software programming:*<br>
<br>
<br>
<br>
                 I am very much familiar with Gdal/OGR  as i used it<br>
        extensively<br>
                 while<br>
                 implementing an outsourced project of Honeywell<br>
<br>
                 Technology . I have played a lot with ESRI shape files.<br>
<br>
                 *Please tell us why you are interested in GIS and open<br>
        source<br>
                 software:*<br>
<br>
<br>
<br>
                 I was very impressed with google earth when it was<br>
        launched way<br>
                 back in<br>
                 2004.GIS platforms like quantum and Openjump were<br>
        introduced  to<br>
                 me when<br>
                 i was in my second year and then i joined Lab for spatial<br>
                 informatics<br>
                 that is in our college based on my interest in this area.<br>
<br>
<br>
                 open source platforms provide an opportunity to learn .<br>
        The best<br>
                 part is<br>
                 you can always create or develop something that you are<br>
                 interested in<br>
                 and there are people who will always be there to help<br>
        you if you<br>
                 are stuck.<br>
<br>
                 *<br>
<br>
                 Please tell us why you are interested in working for<br>
        OSGeo and the<br>
                 software project you have selected:*<br>
<br>
<br>
<br>
                 It will be huge learning experience  since i have been<br>
        using<br>
                 platforms<br>
                 like Qgis , Gdal ,PostGis etc which come under Osgeo<br>
<br>
<br>
                 I chose pgrouting because the algorithms that they have<br>
                 implemented are<br>
                 related to my masters topic and it will be great if i<br>
        get an insight<br>
                 about how these were implemented on a much larger scale.<br>
<br>
                 *Please tell us why you are interested in your specific<br>
        coding<br>
                 project:*<br>
<br>
<br>
<br>
                 Implementing routing algorithms for real world networks is<br>
                 actually very<br>
                 challenging.its a challenge to come up with good<br>
        approaches so<br>
                 that the<br>
                 computation cost and response time is both reduced.<br>
<br>
                 *Would your application contribute to your ongoing<br>
                 studies/degree? If<br>
                 so, how?*<br>
<br>
<br>
                 I am pursuing my Masters and my topic is closely<br>
        related to my<br>
                 proposal<br>
                 . If the results are impressive , i might get a<br>
        publication in<br>
                 this area.<br>
<br>
<br>
                 *Please explain how you intend to continue being an<br>
        active member of<br>
                 your project and/or OSGeo AFTER the summer is over:*<br>
<br>
<br>
<br>
                 I was always interested in open source programming but<br>
        never<br>
                 knew where<br>
                 to start . GSOC  provides an opportunity for beginners<br>
        like me<br>
                 to take a<br>
                 dive in the area  open source programming. i have other<br>
        ideas like<br>
                 implementing IRNN ( In route nearest neighbour<br>
        querries) which<br>
                 can be<br>
                 implemented within pgRouting.I have other intentions like<br>
                 working for<br>
                 bigger projects like Quantum Gis or Grass.I will be<br>
        contributing<br>
                 to the<br>
                 community forum and will always try to improve things.<br>
<br>
                 *Do you understand this is a serious commitment,<br>
        equivalent to a<br>
                 full-time paid summer internship or summer job?*<br>
<br>
<br>
<br>
                 Yes, I understand that this is a serious commitment and<br>
        will try<br>
                 to give<br>
                 my best and work in a professional manner.<br>
<br>
<br>
                 *Do you have any known time conflicts during the<br>
        official coding<br>
                 period?<br>
                 (June 17 to Sept. 27)*<br>
<br>
<br>
<br>
                 No .<br>
<br>
<br>
<br>
<br>
<br>
                 On Thu, Apr 25, 2013 at 12:47 AM, Stephen Woodbridge<br>
                 <<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.com</a><br>
        <mailto:<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.<u></u>com</a>><br>
        <mailto:<a href="mailto:woodbri@swoodbridge." target="_blank">woodbri@swoodbridge.</a>__<u></u>com <mailto:<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.<u></u>com</a>>><br>
</div></div>
                 <mailto:<a href="mailto:woodbri@swoodbridge" target="_blank">woodbri@swoodbridge</a>.<br>
        <mailto:<a href="mailto:woodbri@swoodbridge" target="_blank">woodbri@swoodbridge</a>.>_<u></u>___com<div class="im"><br>
<br>
                 <mailto:<a href="mailto:woodbri@swoodbridge." target="_blank">woodbri@swoodbridge.</a>__<u></u>com<br>
        <mailto:<a href="mailto:woodbri@swoodbridge.com" target="_blank">woodbri@swoodbridge.<u></u>com</a>>>>> wrote:<br>
<br>
                      On 4/24/2013 2:34 PM, Mukul priya wrote:<br>
<br>
                          Hi Steve and Daniel ,<br>
<br>
                          I have posted my proposal on Melange Gsoc<br>
        system. Do<br>
                 have a look<br>
                          at it.<br>
                          Waiting for feedback and suggestions.<br>
<br>
<br>
                      Can you post a link to your proposal? or the text<br>
        of it<br>
                 here also. I<br>
                      can never find anything in Melange or I may not have<br>
                 visibility to<br>
                      it yet as I just applied as a mentor.<br>
<br>
                      Thanks,<br>
                         -Steve<br></div>
                      ______________________________<u></u>_______________________<div class="im"><br>
<br>
                      pgrouting-dev mailing list<br>
        <a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.osgeo.org</a> <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>><br>

                 <mailto:<a href="mailto:pgrouting-dev@lists." target="_blank">pgrouting-dev@lists.</a>__<a href="http://osgeo.org" target="_blank"><u></u>osgeo.org</a><br>
        <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>>><br></div>
                 <mailto:<a href="mailto:pgrouting-dev@lists" target="_blank">pgrouting-dev@lists</a>.<br>
        <mailto:<a href="mailto:pgrouting-dev@lists" target="_blank">pgrouting-dev@lists</a>.>_<u></u>___<a href="http://osgeo.org" target="_blank">osgeo.org</a> <<a href="http://osgeo.org" target="_blank">http://osgeo.org</a>><br>

                 <mailto:<a href="mailto:pgrouting-dev@lists." target="_blank">pgrouting-dev@lists.</a>__<a href="http://osgeo.org" target="_blank"><u></u>osgeo.org</a><br>
        <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>>>><br>
        <a href="http://lists.osgeo.org/______mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/______<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/____mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/____<u></u>mailman/listinfo/pgrouting-dev</a><u></u>><br>
<br>
        <<a href="http://lists.osgeo.org/____mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/____<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><u></u>>__><div><div class="h5"><br>
<br>
          <<a href="http://lists.osgeo.org/____mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/____<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><u></u>><br>
<br>
        <<a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/<u></u>mailman/listinfo/pgrouting-dev</a><u></u>>__>__><br>
<br>
<br>
<br>
<br>
<br>
<br>
                 ______________________________<u></u>_____________________<br>
                 pgrouting-dev mailing list<br>
        <a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.osgeo.org</a><br>
        <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>><br>
        <mailto:<a href="mailto:pgrouting-dev@lists." target="_blank">pgrouting-dev@lists.</a>__<a href="http://osgeo.org" target="_blank"><u></u>osgeo.org</a><br>
        <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>>><br>
        <a href="http://lists.osgeo.org/____mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/____<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><u></u>><br>
<br>
        <<a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/<u></u>mailman/listinfo/pgrouting-dev</a><u></u>>__><br>
<br>
<br>
             ______________________________<u></u>_____________________<br>
             pgrouting-dev mailing list<br>
        <a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.osgeo.org</a><br>
        <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>><br>
        <mailto:<a href="mailto:pgrouting-dev@lists." target="_blank">pgrouting-dev@lists.</a>__<a href="http://osgeo.org" target="_blank"><u></u>osgeo.org</a><br>
        <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>>><br>
        <a href="http://lists.osgeo.org/____mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/____<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><u></u>><br>
             <<a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/<u></u>mailman/listinfo/pgrouting-dev</a><u></u>>__><br>
<br>
<br>
<br>
<br>
        ______________________________<u></u>___________________<br>
        pgrouting-dev mailing list<br>
        <a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.osgeo.org</a> <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>><br>

        <a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><br>
        <<a href="http://lists.osgeo.org/mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/<u></u>mailman/listinfo/pgrouting-dev</a><u></u>><br>
<br>
<br>
    ______________________________<u></u>___________________<br>
    pgrouting-dev mailing list<br>
    <a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.osgeo.org</a> <mailto:<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.<u></u>osgeo.org</a>><br>

    <a href="http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/__<u></u>mailman/listinfo/pgrouting-dev</a><br>
    <<a href="http://lists.osgeo.org/mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/<u></u>mailman/listinfo/pgrouting-dev</a><u></u>><br>
<br>
<br>
<br>
<br>
______________________________<u></u>_________________<br>
pgrouting-dev mailing list<br>
<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.osgeo.org</a><br>
<a href="http://lists.osgeo.org/mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/<u></u>mailman/listinfo/pgrouting-dev</a><br>
<br>
</div></div></blockquote><div class="HOEnZb"><div class="h5">
<br>
______________________________<u></u>_________________<br>
pgrouting-dev mailing list<br>
<a href="mailto:pgrouting-dev@lists.osgeo.org" target="_blank">pgrouting-dev@lists.osgeo.org</a><br>
<a href="http://lists.osgeo.org/mailman/listinfo/pgrouting-dev" target="_blank">http://lists.osgeo.org/<u></u>mailman/listinfo/pgrouting-dev</a><br>
</div></div></blockquote></div><br></div>