[DotNet-OSGeo] some random thoughts

Tue Jun 1 14:06:11 EDT 2010

So this is a great conversation thread that you guys are starting.  I have a
lot of experience with working more towards the pure windows forms
application side of things, and less so as far as setting up web stuff, so
we can use ideas about how we need to tweak things for optimal compatibility
with as many folks as possible.  I agree with the principals outlined in the
previous posts and much of this is simply good coding convention.  I have
been using resharper in order to set up some formatting rules, but I suspect
that people won't necessarily want to pay for it, so if you know of
something else to help with that, I'm interested.  I like trying to set
things up so that people can use C# express in order to work with the
libraries, however, I am getting reports that people writing plug-ins might
not be able to have their plug-in launch an external program for debugging
from the express version.  Also, I'm not sure about the implications with
trying to use system.Reflection for run-time referencing in other
platforms.  Is it ok to use, for instance, a plug-in architecture for
referencing OGR/GDAL, or should we be thinking in terms of porting content
to C#?

So we currently have source code that is good for accessing shapefile data
in C#, but it doesn't offer anything close to what OGR or other vector
providers could do.  It also relies heavily on the unmanaged GDAL libraries
in order to support various image/raster data formats, via a plug-in
architecture.  Unfortunately, according to your comments, it also makes use
of DataTables.  (It can either load all the data at once for small dbf
files, or can set up a page query to take advantage of the virtual data
tables in .Net, but it does not take advantage of the new IQueriable
structures or linq)  I like linq, but what I have done with it tends to work
extremely well for scenarios where the developer knows the schema ahead of
time, and I'm not sure how well it works in situations where you were
dealing with the unknown schema.  However, I'm sure there is a way to make
all that work, I just haven't done much of it yet, but would not mind
tailoring our data access libraries to move more in that direction.

Once Dan sets up an IRC line, I will be happy to get on it.  I'm not a
current IRC user, (we used Windows Live Messenger).  However, I have no
problem switching over to that.

I'm really looking forward to getting our libraries out there in a way that
is more useable and more versatile for everyone.  I also appreciate the time
that you guys are putting in in order to take this idea seriously.

Ted Dunsford (Shade1974)

On Tue, Jun 1, 2010 at 10:12 AM, John Diss <john.diss at newgrove.com> wrote:

>  Hi Dan please see comments inline:
>
>
>
> Also you may be interested in Mono.Addins and MEF for your plugin
> architecture consideration.. alternatively spin up new App Domains to host
> your plugins.
>
>
>
>
>
>
> All external apis where possible should be parallelizable. E.g collections
> should be synchronized.
>
>  Cool, but not sure how feasible. Thoughts?
>
>
>
> I am thinking along the lines of the new System.Collections.Concurrent
> namespace. Computers aren’t getting any faster - we need to make sure we can
> use as many cores as possible to get the speed gains we need.
>
> Parallelization is also very hard to retrofit so if we can get a plan in
> early we will benefit.
>
>
>
> Data access should be async.
>
>  Especially for web based data access. Maybe not for file based?
>
>
>
> IMO all IO bound access should follow the same pattern : web, db or file.
> There is no point hitting datasources sequentially when you can do so in
> parallel.
>
>
>
> There should be a caching block developed with multiple backends. Uses
> ranging from caching data/ partial screen renders, wms/wfs output etc.
>
>  Not sure exactly what the implications are... but interesting idea.
>
>
>
> Basically have a generic caching block with rules for scavenging and a
> variable backing store which can be plugged into any part of the processing
> /rendering pipeline. It could be equally at home as an in-memory cache of
> features on a layer or as an image cache speeding up identical wms requests.
> – There is some work on this in the asp.net bits of SharpMap2 but it could
> be made more generic..
>
>
> Supported frameworks should be considered up front - my personal preference
> would be to start with a clean slate at .Net 4 or at a minimum 3.5.
>
>  We were targeting 3.5 because it would be more likely to be mono
> compliant... (does MONO really support all of .NET 4 now?)
>
>
>
> Miguel certainly seemed to think so in a vid I saw recently..
> http://mono-project.com/Compatibility tells a slightly different story -
> but being an OSS project the documentation is out of date and probably wrong
> J I will try and locate the vid.
>
>
>
> If we hope to support MONO (which I think we should) we must ensure that
> any rendering api's have a potential replacement in Mono. Mono is currently
> compatible with .net 4.
>
>   Any ideas on how much of .NET 4 is implemented in MONO now?
>
>
>
> See above – I think they are also working on making the GC better.
>
>
>
> Any class expecting to retrieve (vector/attribute) data from a datasource
> should demand IQueriable<I/TFeature> rather than a FeautureProvider or
> similar. DataTable / Dataset should be avoided. The query syntax should be
> either linq against TFeature and/or based on expressions taken from the
> OSGeo Filter spec.
>
>  Not sure I understand... can you explain further?
>
>
>
> It was partly a spelling mistake on my part: should have been
> IQueryable<IFeature> which is basically the underpinning of Linq.
> The OsGeo filter spec details a dialect for describing a query which is
> similar to the System.Expressions namespace used in defining Linq queries.
> Both would also provide an easy mechanism for serializing/persisting
> queries.
> Datasets and DataTables are very heavy/ verbose and don’t exist on certain
> platforms. Pretty much all the trends as far as I can see, is to kill them
> off in favour of domain entities / ORM. DataTables / Datasets are also
> frowned upon in WCF which is another set of technology that could be very
> useful in this field. We have an expression model in SharpMap 2 but it is a
> bit clunky at the moment and disconnected from linq.
>
>
>
> Silverlight/Moonlight has massive potential for browser based gis / mapping
> / processing / grid computing. Supporting it should definitely be planned
> for but it imposes its own forced async style which may have implications
> elsewhere. Also win phone 7 will be Silverlight based. Novell have MonoTouch
> and MonoDroid in progress for IPhone and Android respectively and
> www.KoushikDutta.com  is working on another free Android/Mono platform.
>
>  What are all the implications of supporting Silverlight and Moonlight? Is
> it possible to have a standard .NET DLL that can be used in a Silverlight
> application?
>
>
>
> The binaries are compatible at .net 4 not so previously afaik. Silverlight
> is both a super and sub set of  WPF so there are differences. System.Data is
> one key difference. IMO WPF and Silverlight will merge into one over the
> next couple of releases.
>
>  MS C#/dotspatial group style guidelines should be followed -  everyone’s
> code should look the same. Perhaps a developer R#/stylecop profile should be
> created and enforced by a CI server.
>
>  Agreed. Is it possible to do this kind of enforcement using Codeplex?
>
>
>
> I don’t think codeplex allows this. I have heard good things about
> Jetbrains’ TeamCity  and lots of people use CruiseControl. BTW jetbrains
> give away R# licenses.
>
>
>
> Source control should be backed with a CI / Test server.
>
>  Can this be done on Codeplex?
>
>
>
> Not afaik.
>
>
>
>  Perhaps the logs should be taken and made available somewhere.
>
>  Yeah, any ideas on how to do this other than just through the codeplex
> log browser?
>
>
>
>
>
> I actually meant the IRC logs.
>
>
> Documentation should be easy to find and of high quality. Perhaps the group
> needs a sister website separated from codeplex where it is easier to manage
> the wiki/ public contribution etc. The codeplex wiki is very limited in this
> respect.
>
>  Yes agreed. This is the number one complaint I believe of any/every open
> source GIS project that there isn't adequate high quality documentation.
> We've registered "dotSpatial.org" so maybe that should point to a
> non-Codeplex server where we can maintain documentation and project
> description/etc. Then have pointers to the Codeplex site for source code,
> issues and forums. What do you think of this approach? Would you (John or
> others) be interested/willing to help decide what would go such a separate
> site? layout, etc?
>
>
>
> I am happy to help out wherever I can..
>
> _______________________________________________
> DotNet mailing list
> DotNet at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/dotnet
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/dotnet/attachments/20100601/a25d35c2/attachment.html