Hi Massimo,
I acknowledge and appreciate your passion and effort in pushing to get
notebooks ready for the osgeolive 9.5 release.
I also hear your concerns that you would have liked to seen feedback
earlier and we can discuss that separately. However, I still don't think
notebooks should be easily visible in OSGeo-Live 9.5, for the reasons
I've already mentioned, and we should be targeting OSGeo-Live 10.0.
I'm open to being out-voted on this point, and would appreciate to hear
opinions from others in the osgeolive community.
Warm regards, Cameron
On 10/03/2016 11:39 am, massimo di stefano wrote:
On Mar 9, 2016, at 5:59 PM, Cameron Shorter
<cameron.shor...@gmail.com <mailto:cameron.shor...@gmail.com>> wrote:
-1 from me for including all python notebooks on the upcoming
OSGeo-Live 9.5 release in a few days. I don't think we can make it
ready in time.
This is based on my preliminary review of the gsoc iso:
http://osprey.ucdavis.edu/downloads/osgeo/gisvm/gisvm/9.5nightly/osgeo-live-nightly-gsoc-amd64-c83a060.iso
the are are several commit already after that iso
I haven't done a comprehensive review, but have reviewed enough to
feel the initial user experience when looking at notebooks is close,
but is yet to meet our OSGeo-Live presentation standards.
My comments so far:
1. Quickstart review comments haven't been addressed yet, although
hopefully Massimo should have that complete soon.
the quickstart and overview with your comments addressed:
https://github.com/epifanio/OSGeoLive-doc/blob/master/en/quickstart/jupyter_quickstart.rst
https://github.com/epifanio/OSGeoLive-doc/blob/master/en/overview/jupyter_overview.rst
2. Structure of notebook docs directory should be by project, rather
than singled out as a GSoC directory, then by project. Long term we
want GSoC content to be part of the core OSGeo-Live offering.
That’s a personal thought, you had 10 months to think about this.
3. We are missing an obvious Table of Contents.
it is here:
https://github.com/epifanio/OSGeoLive-Notebooks/blob/osgeolive/index.ipynb
4. I've randomly picked iris_basics.ipynb, running step 2, I get a
warning about a library being deprecated. (It still works, but this
is the sort of thing that should be cleaned up before presenting to a
new user).
i’m not contributing to this, but it is just a warning.
5. There should be an introductory notebook which starts with "Press
the RUN CELL button to see the output of [In] cells”.
it is here:
https://github.com/epifanio/OSGeoLive-Notebooks/blob/bf2fcb56480738796cf8a0e0a1a1fc5195f96e92/introduction-to-jupyter-notebook.ipynb
6.1 I've found a number of spelling and grammar errors in the first
GSoC notebook I reviewed. Nothing major, but they should be
identified and cleaned up all the same.
For example, this is the first paragraph of the GDAL/OGR Quickstart
(which is the first Docbook I looked at):
/"The first Notebook is dedicared [SPELLING] to the use of the
Geospatial Data Abstraction Library GDAL [USE BRACES AROUND FIRST USE
OF CAPITALS] from the bash command line. GDAL is a powerful
translator library for raster and vector geospatial data formats
[INSERT FULL STOP] it presents a single raster abstract data model
and vector abstract data model to the calling application for all
supported formats.”/
6.2 In the same notebook, I found:
/"#unfinished - need to add ogr2ogr features yet, I'm considering to
split this notebook in 2 parts (one for gdal and one for ogr2ogr)
i'll study the gdal tutorial released recently"/
Massimo notes that he is fixing this, but my point is that I'm
assuming that if reviewed, similar issues would be found in other
notebooks.
they are fixed:
https://github.com/epifanio/OSGeoLive-Notebooks/blob/osgeolive/GSoC-2015/Access%20to%20Geospatial%20data/GDAL-OGR%20Quickstart.ipynb
7. The term "IPython" and "Jupyter" is used interchangeably, which
would be confusing for a new user, and I'd be interested to
understand if is technically incorrect?
where?
if you refer to the the code, that’s normal ipython is part of jupyter.
8. Quantity of material to review and short time period. We have days
left before the our final release candidate is due, and multiple
notebooks (say 20+ ?) which I suggest need review to get to same
level of review as Project Overviews and Quoickstarts.
the GSoC notebooks have been reviewed,
you’re finding typos ok, my apology … as a non native speaker
but again, you had 10 months to think about this
if you find more, i’m open to fix them ASAP
9. Also on quantity of material, we are introducing a maintenance
load for future OSGeo-Live releases, and we are as yet to build up a
community of volunteers willing to maintain these notebooks. Massimo
and Brian have been doing a great job creating these notebooks, but
are already under time pressure to update for this release, let alone
future releases. I think we need to reach out to all projects with
the offer of:
"A notebook has been created for your project. If you are prepared to
review it now, and commit to verifying it is still valid before each
OSGeo-Live release, then we will include the notebook on OSGeo-Live.”
As per the GSoC notebooks, I think I’m demonstrating that you should
not worry about their maintenance.
I addressed your comment, I can’t speak for the notebooks under the
projects directory.
My recommendation is:
we can rename the GSoC-2015 directory .. perhaps with the name of the
idea itself which is:
Open Source Geospatial Notebooks.
then if you have concerns about the “projects" directory, we can
comment it out from the table of contents and leave it on disc.
note the quick start and overview they both uses the GSoC material, so
they will not be affected.
--
So my recommendation is to either:
A. Include Notebooks on next OSGeo-Live 9.5, but remove reference to
it from main docs. (My preference at this point)
B. Include ONE Jupyter notebook example which can be found from main
docs, which we make sure is well reviewed. All the remainder of
Notebooks to be hidden. (This will likely be a messy solution, as
we'd probably need to move directories around so they are hard to
find for a new user).
On 9/03/2016 11:44 pm, massimo di stefano wrote:
To All,
during the last 2 years I worked in having the python software
ecosystem included in the OSGeo-Live.
The PR proposed below [1] is a first step towards the official
inclusion of this work into the project.
As you may have noticed from previous e-email [2] there isn’t a
clear agreement on the decision to take.
We had an IRC discussion (logs [3]) with an initial motion to merge
the notebook PR.
I would like to have a voting started here on the mailing.
this is mine:
+1
Thanks,
Massimo.
===
[1] https://github.com/OSGeo/OSGeoLive-Notebooks/pull/5
[2] http://lists.osgeo.org/pipermail/live-demo/2016-March/011010.html
[3] http://irclogs.geoapt.com/osgeolive/%23osgeolive.2016-03-07.log
<http://irclogs.geoapt.com/osgeolive/#osgeolive.2016-03-07.log>
_______________________________________________
Live-demo mailing list
Live-demo@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/live-demo
http://live.osgeo.org
http://wiki.osgeo.org/wiki/Live_GIS_Disc
--
Cameron Shorter,
Software and Data Solutions Manager
LISAsoft
Suite 112, Jones Bay Wharf,
26 - 32 Pirrama Rd, Pyrmont NSW 2009
P +61 2 9009 5000, Wwww.lisasoft.com, F +61 2 9009 5099
_______________________________________________
Live-demo mailing list
Live-demo@lists.osgeo.org <mailto:Live-demo@lists.osgeo.org>
http://lists.osgeo.org/mailman/listinfo/live-demo
http://live.osgeo.org
http://wiki.osgeo.org/wiki/Live_GIS_Disc
--
Cameron Shorter,
Software and Data Solutions Manager
LISAsoft
Suite 112, Jones Bay Wharf,
26 - 32 Pirrama Rd, Pyrmont NSW 2009
P +61 2 9009 5000, W www.lisasoft.com, F +61 2 9009 5099
_______________________________________________
Live-demo mailing list
Live-demo@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/live-demo
http://live.osgeo.org
http://wiki.osgeo.org/wiki/Live_GIS_Disc
