Gund Done?

Taylor wrote

 formatting – still some minor things here.  Remember, we have 3 types of publications: journal article, book, and book section.  That should be recorded a field in the database we send you.  Here are specifics:
– journal articles: look great.
– books: please don’t underline the Publisher or Place Published fields.  Look at ‘Zia (2013). Post-Kyoto Climate Governance… ‘.  That’s perfect, except nothing should be underlined.  Same with all other books.
– book sections: formatting still a bit off on these.  Formatting should be Author (Year). Title. Editor. Book Title. Publisher. Place Published. Pages.
For example, here is a listing  in the database:
Daily, G. C.; Kareiva, P.; Polasky, S.; Ricketts, T.; Tallis, H.(2011) Mainstreaming Natural Capital into DecisionsOxford University Press, New York. Pages 3-14;
And here’s what we’d like it to look like:
Daily, G. C.; Kareiva, P.; Polasky, S.; Ricketts, T.; Tallis, H.(2011) Mainstreaming Natural Capital into Decisions. in Kareiva, P., Tallis, H., Ricketts. T.H., Daily, G.C. Polasky, S. Natural Capital: Theory and practice of mapping ecosystem servicesOxford University Press, New York. Pages 3-14.
Keywords.  This is really the only new element to the project.  We’d like people to be able to not only search on people or keywords, but also to filter the publications by our 3 main research themes.  We’d like to have 3 simple clickable buttons or words at the top, like the search box, where they can click on one of our three themes and get a search result of papers that fall into that theme.  To make this possible, we are adding to the keywords field for each publication a code corresponding to each of our three research themes.  We are hoping you can add the clickable words or buttons and then have D-base search on those unique codes to produce the filtered publications list.  It’s essentially setting up 3 canned searches, on codewords that we give you and that we’ll put into each record.

I think I got it

http://www.uvm.edu/giee

Posted in Wes | Tagged , | Leave a comment

archaeological project data


On May 14, 2014, at 6:31 PM, Mike Austin wrote:

On Wed, 14 May 2014, Andrew Hendrickson wrote:

This is a long shot, but is anyone else at UVM deploying ARK?  It’s an open source application for recording archaeological project data.  I have a client interested in installing this on an appropriate server, but I thought I’d check to see if such a beast was already in use at UVM.

Not that I know of.  But it looks like this may run on our silk hosting environment pretty well if you are looking for a place to run it.  (and have someone who will be responsible for updating it and signing up for any security-related announcements for the product)


http://ark.lparchaeology.com

 

I told Andrew this sounded like a CTL project. He hooked me up with Parker VanValkenburgh, Ph.D. After small struggle, managed to get it installed on http://ctl.w3.uvm.edu/ark . But it is kinda sucky, difficult to use, several PHP coding errors. Parker also looking at one other system, entitled FAIMS, which is still in the latter stages of Alpha, I believe. Here is the link:

https://www.fedarch.org/wordpress/development/information-for-developers/

 

the wiki for the fairs server.

https://wiki.fedarch.org:8443/display/FAIMS/Mobile+Platform+Home

 

This wants Ubuntu, and installation on a portable ‘server’ than can be carried into the field, not necessarily with internet axis.

Implementing this lead to discovery of VirtualBox, a fee open source hyperviser virtual machine environment. Installed on my Mac, installed Ubuntu, and imported mt Parallels Windows 8 VM. See separate post for that.

Now waiting for FAIMS servers to come back online before proceeding

Posted in --Project Summary, Wes | Tagged , , | Leave a comment

Hybrid Case Studies

Ines asked,

I met with J and Henrie this week because they want me to write some case studies on the hybrid course development/teaching process.
Eventually there will be 30+ but it’ll be a smaller number to start.  They want these to be presented as part of the CTL website.
They want faceted search capability. So people could, go to the “case studies page” and do a search for all courses in CALS that went through the hybrid program…. OR all CALS courses that are over 100 students…OR all undergraduate courses from any college.
There would therefore be several different search variables. Here are just a few of them :
College e.g. CAS, CALS, CEMS… etc  
Level e.g undergrad, graduate;  
Class size e.g. 0-25, 25-50, etc;
Hybrid model of teaching  e.g. Split, Alternative, etc.
Here’s what I want to know…
is MySQL the way to do it?  Can it just be done with with PHP and tagging?
Also, I should mention that the queries that are generated would produce a list with any number of “finds” (courses) listed on it, so if someone searches for 
College: BSAD  +  ClassSize: 25-50  + HybridMode: alternating
they may only come up with one course that meets that criteria (or none), but if they search, say, only one variable like “College: CAS” and choose no other parameters, they may have 20 results.

I delivered


http://www.uvm.edu/ctl/apps/hybrids/

Posted in --Project Summary, Wes | Tagged , , | Leave a comment

IT-Align survey

I was tasked to extract select statistics an summarize for the committee .

Posted in Wes | Tagged | Leave a comment

MS Access Issues

On Mar 31, 2014, at 11:09 AM, Kristen Cameron wrote:

Hi Wes:
I have some MS Access questions again. I’m wondering if you could show me how to show more information in queries. As it is set up know, I only have last names included. I would like to see the full name and department if possible. Do you have some time to help me with that? Not a huge rush or anything. It’s just something that has been annoying me for awhile.
Thanks,
Kristen


Kristen Cameron
Program Assistant
Writing in the Disciplines

Access is a beastly app, but I managed to tame it. It amounts to changing the query that populates the Attendee field in the RSVPs table from

SELECT [Constituents].[ID], [Constituents].[Last Name], [Constituents].[First Name] FROM Constituents ORDER BY [Last Name], [First Name], [ID];

to

SELECT [Constituents].[ID], [Constituents].[Last Name], [Constituents].[First Name], [Constituents].[Department] FROM Constituents ORDER BY [Last Name], [First Name], [ID]; 

Then go to the Create ribbon and create a new data sheet-style form

Copy the Attendee field and paste it bak under the existing Attendee field. Paste and position again. Rename the original Last Name, the next First Name and the last Department. Edit format of the column widths of each to only show 1 column

Posted in Wes | Tagged , | Leave a comment

Sending event followup emails

As discussed at 2014-04-08 staff meeting, Please see, for example

https://www.uvm.edu/ctl/development/ctlcal/default.php?c=admin&m=roster&id=991
Note new menu item at top

Screen Shot 2014-04-08 at 2.20.40 PM

Prompts to send “Thank you” emails to attendees who showed up and “Sorry you missed it” emails to no-shows.
Nina will Let me know what these will really say, and what the survey URLs shall be.

 

Posted in Wes | Tagged , , | Leave a comment

Big Blue Buttton front end completely overhauled, needs testing

Completely rebuilt from ground up. no longer Drupal dependent. Record meeting feature finally available. https://www.uvm.edu/ctl/apps/bbb

Posted in Wes | Tagged , , , , , | Leave a comment

TEO self assessment form

On Apr 2, 2014, at 11:07 AM, Henrietta Marcella Paz-Amor wrote:

Changes to TEO pre-meeting form:
Form changes – add first last name field

At submit message change to: 
“Based on your score you have been signed up for the pre-TEO blackboard workshop.”
 
Message above also triggers email sent to Wendy V-B with name of participant that will attend workshop

Done 12:13 PM, along with some improvements/fixes to http://www.uvm.edu/ctl/?Page=resources-teaching/online-teaching/responses.php

Posted in Wes | Tagged , , , | Leave a comment

MediaManager bug fix leads to enhancements

The application at

https://www.uvm.edu/htpasswd/podcast_manager

was exhibiting strange behavior. Guy from COM had 607 Mb file, wouldn’t upload.

Fixed and improved, can now upload up to 10Gb files (when used with a ‘modern’ browser).
Some background …

It’s just a glorified file uploader, relying on plupload (1.5.8).  There is evidence (/users/c/o/comvideo/public_html/podcast_folder) that as of last October, it was working fine. Now, For a certain file size 343Mb < N < 607MB, upload succeeds in Safari and FireFox on MacOS; fails on Chrome Mac and Win, FireFox Win, and IE. It is setup to use HTML5 runtime, no chunking for all but IE. For IE, Silverlight and chunking.

Chunking? Yes. Upload widget sends file in (now set to) 10Mb 'chunks.' Each chunk is sent to the same php script, along with file name, chunk number, and total chunks. php script appends chunk  to file. This gets us around  the (UVM/ETS/SAA) imposed php max file size limit of 1Gb.

Step forward

I blamed it all on upgraded browsers and old plupload. So I created

https://www.uvm.edu/htpasswd/podcast_manager_dev

with plupload 2.1.1 and chunking all the time. And here is what I discovered:

The podcast_manager lives in /shares/uvmweb/htdocs/htpasswd/ , which has SAFE_MODE turned off, and relies on WebAuth set_uid magic to allow the script to create files in the REMOTE_USER’s home directory

plupload sends the first chunk 0, upload.php  creates filename_part with mode ‘wb’

$out = fopen(“{$filePath}.part”, $chunk == 0 ? “wb” : “ab”);

creates file and writes data to file. plupload sends next chunk 1, upload.php attempts to open file using mode ‘ab’, FAILS with a permission denied error, system crashes and burns. In fact, even trying to create a new file using mode ‘ab’ fails

 

On Mar 28, 2014, at 10:27 AM, Benjamin Coddington wrote:

I just (this morning) fixed a problem where un-authenticated PHP could not open files with O_APPEND.  That shouldn’t have changed the behavior for this case because you are using the WebAuth authentication.

Well, it shouldn’t, but now it does work

Posted in Wes | Tagged , | Leave a comment

Events Calendar Performance

Many complaints about sluggish events calendar performance. After considerable analysis, the bottleneck was www.uvm.edu . Talks with SAA suggested Network File System (NFS) performance as culprit, as was how the CodeIgnitor MVC (model-view-controller) framework used the file system.

 

Case in point: I had a CodeIgnitor “View” (a single php file) that displayed details of one event. The default calendar view displays maybe 70 events, past and present. So, application loops through all the events, invoking the Single Event View 70 times. That incurred 70 file reads from a sluggish NFS file share. SSllllloooowww.

So I recast the 1-event-view as a function call. 1-event view loads function once, invokes function once. N-event-view loads function once, invokes function N times. Fast.

While I was at it, I optimized some of the SQL queries by adding additional indexes to database tables.

Posted in Wes | Tagged , , , , | Leave a comment