[Qgis-user] qgis and saga (future) integration hypothesis

[G. Allegri]

Hi Paolo.
SAGA community is apprently small. The dev team is small, but they work
hard, even if it doesn't appear from the SAGA site.
I've just chatted with its major developer (Volker) and he was asking help
to mantain the build chain updated and give a refresh to the site (wiki is a
bit outdated, etc.). The latest version build binary is not available, but
there's a quite new snapshot at Laserdata [1], where some of its developers
work.

I said the community is large, even if its not so noticeable. I have many
collegues that use it (more then GRASS), also through the very good RSAGA
interface for R [2]. Tomislav Hengl has lot of material about this [3]

Why I'm so much interested in SAGA:

I find it has all the features to make lot of (vectorial, grid and imagery)
analysis and automation. It includes functions that GRASS doesn't offer out
of the box, that you can find (partly) in the latter user manual (a litlle
bit outdated) [4].
I don't have time in this moment to list all the differences between it anf
GRASS, and please don't misunderstand me: I love GRASS, and will continue
using it too... but:

 1 - the GRASS Windows version situation is a major problem, while SAGA
works smooth both on Windows and Linux. I don't know which version you used,
I haven't tried the linux version yet, but I don't see bad feedbacks in the
resent ML posts.

2 - It doesn't need to be installed: I use it from my pen drive :)

3 - It doesn't work with topology. Yes, it can be seen as a missing,
important, feature for someone and for some uses, but the end users often
prefer to work directly on shapefiles, without the overhead of clean/build
issues during GRASS imports... And that's one of the points that make QGIS,
Gvsig, etc. more usable.

4 - The latest version (that I'm going to buid), links well with the the
latest libraries (GDAL, PROJ) and Python 2.5.

5 - While the GRASS dev team is putting much effort in the GUI development,
SAGA GUI is a bit secondary. It would gain great benefit from an integration
with QGIS...

Ok, this is just my personal, subjective, overview. But I repeat I don't
wnat to discourage GRASS usage, but to highlight this good alternative...

Giovanni

[1] http://www.laserdata.at/saga_packages.html
[2] http://cran.r-project.org/web/packages/RSAGA/index.html
[3] http://spatial-analyst.net/wiki/index.php?search=rsaga&go=Go
[4]
http://garr.dl.sourceforge.net/sourceforge/saga-gis/SAGA2_UserGuide_Cimmery_20070401.pdf

2009/7/11 Paolo Cavallini <cavallini at faunalia.it>

> G. Allegri ha scritto:
> > SAGA algorithms were the basis on which Sextentegis was developed. Then
> > this was integrated into Gvsig... Why don't make happen the same for
> > SAGA? :)
>
> Hi Giovanni.
> I have tested saga a few years ago, and I was rather disappointed by its
> (lack of) stability on Linux. Also, the apparently very small community
> was a bit worrisome to me. More generally, I do not understand well what
> is the advantage of using saga rather than grass: have you checked their
> different strengths and weak points?
> Thanks a lot.
> --
> Paolo Cavallini: http://www.faunalia.it/pc
> _______________________________________________
> Qgis-user mailing list
> Qgis-user at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/qgis-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20090711/c340fbca/attachment.html>