[pgrouting-dev] Boston Code Sprint

[Stephen Woodbridge]

Hi Daniel and Mario,

There is going to be a Boston code sprint in February or March 2013 that 
I am planning on attending. (see email below). I would like to work on 
pgRouting and want to open up the discussions about integrating it with 
postGIS again, at least from a build and release point of view if not in 
other ways that might be acceptable to all.

I know Mario is working on stuff now. It would be good to try and coral 
the various random work being done into some kind of mini freeze before 
this event which is TDB.

To this end, I would like to get a list of what Mario is working on and 
any timeline if he has one. Additional input from Sanak, Steve Horn, Max 
Weninger, and Jay Mahadeokar, Razequl, Jinfu and anyone else that might 
have a fork and changes that might need to get integrated. If you have 
generated any RFC's or have ticket numbers that would be handy if you 
supplied links to them.

I would like to also get a list of priority items that I should try to 
focus on for the sprint. High on my list of issues is putting in place 
an automated test framework that can be run via "make test" so that 
developers can check that changes do not break things. And to establish 
a pattern for future additions. My thoughts on this are that we need two 
levels of testing:

1. high level via sql commands
2. low level command line tool the reads a text file, sets up structs 
and then calls the algorithm code all outside of postgresql.

This allows for simple low-level algorithm verification and debugging 
via gdb. While the higher level test can be used to validate wrapper and 
database integration separate from the algorithm. Both of these can use 
the same test data using a simple tool to extract the sql tables into a 
text file for the low-level tests. TRSP already has a commandline tool 
like this. and I believe Razequl and Jinfu have developed a tool like 
this for there own test efforts.

Code sprinters, I would value you input as well if you have input on any 
of these issues.

Thanks to all,
   -Steve

On 7/14/2012 2:13 PM, Paragon Corporation wrote:
> Steven,
>
> It would be great to have you.  I've cc'd Mark, Mat, Paul, Olivier, and strk
> since you mentioned pgRouting.  I know Olivier is interesting in pgRouting
> as well.
>
> One of the issues I see is that pgRouting has a CMake build structure.  Mat
> and Mark have been working hard to change PostGIS to have a CMake build
> system as well.  I think this will make pgRouting easier to integrate with
> PostGIS.
>
> One of our hopes is that once CMake build is in full swing we can start
> packaging pgRouting for windows alongside our PostGIS windows builds.  This
> is one issue some windows folks have raised about not wanting to use the
> newer PostGIS/PostgreSQL versions because they rely on pgRouting and
> binaries (at least for windows) are hard to come by.
>
> We're also thinking about having a similar campaign for packaging pgRouting
> (at least for windows) that we did for the window 64-bit.  We've been
> hesitant just because we haven't surmised the effort involved to do so.
> This could help with the cost of migrating to a CMake build system and
> testing the CMake build system.
>
> Thanks,
> Regina
>
>> -----Original Message-----
>> From: Stephen Woodbridge [mailto:woodbri at swoodbridge.com]
>> Sent: Friday, July 13, 2012 8:41 PM
>> To: Paragon Corporation
>> Subject: Re: Boston Code Sprint
>>
>> On 7/10/2012 8:25 PM, Paragon Corporation wrote:
>>> Steve,
>>> You're near Boston right and are heavily involved in MapServer,
>>> PostGIS, etc?  You think you could partake in the Boston
>> Code sprint
>>> coming up that we are hosting?
>>> Would be nice to chat about geocoding and other things.
>>> http://wiki.osgeo.org/wiki/Boston_Code_Sprint_2013
>>> Thanks,
>>> Regina and Leo
>>
>> I would like to come and meet people.
>> I might be able to do some hacking also. :)
>>
>> I would like to talk about pgRouting also. I know there have
>> been some discussions in the past about combining this with
>> postGIS but I was not part of that.
>>
>> Thanks,
>>     -Steve
>>
>
>