Tracking feedback

feedback/andy-pavlo-cmu.txt

@@ -0,0 +1,25 @@
+On Monday, November 23, 2015 12:57:25 PM Oliver Kennedy wrote:
+I don't think we've ever been properly introduced, but we've run into each other at some conferences here and there --- I worked with Yanif on DBToaster.
+
+Oh word. DBToaster is (was?) dope. Thanks for getting in touch.
+
+After seeing your Databaseology post, I was hoping to get your input on an NSF Community Infrastructure proposal that some folks at UB are working on. 
+
+The embedded DBMS stuff was just something that I was interested in, and not a new research area that I am pursuing.
+
+Android smartphones run about 2 SQL queries per second on average, all handled (poorly) by SQLite.  We're putting together real-world studies to measure how smartphone apps use embedded databases, and developing a benchmarking toolkit for embedded databases based on the needs of smartphone apps.
+
+- Would you benefit from having access to real-world traces of smartphone embedded database usage in the wild, or statistics derived from these traces (and if so, how?)
+- Would you be able to make use of a benchmarking toolkit simulating the query behaviors of smartphone apps?
+- Do you have any suggestions for features you'd like to see in either?
+
+This is a good idea. It's close to what we're doing, but we are focusing on OLTP web apps. We have built a framework that finds DB applications on GitHub/Bitbucket and then automatically installs/runs them inside of a VM. We then capture the query traces from MySQL and analyze them. We were able get about 5000 apps to work this way (although some of them are trivial). We have a tool that you can run to install and run any web app that we find inside of a VM so that people can run them locally and recreate the query workloads.
+
+Our website will go live next month, but we have some basic info here. I've also attached screenshots:
+
+http://db.cs.cmu.edu/projects/dbac/
+
+So my two suggestions for your project are to make sure that you have software that can be easily re-used by other people and maybe look into a programmatic way of running and interacting with the phone apps that would allow you to collect the data automatically.
+-- 
+Andy Pavlo
+pavlo@cs.cmu.edu

feedback/ashok-joshi-bdb.txt

@@ -0,0 +1,13 @@
+Hello Oliver,
+ 
+I got some feedback from one of my colleagues on this topic.  Yes, the real-world traces of embedded data usage would be useful; so would the benchmarking toolkit.
+ 
+On a closely related topic, I think synchronizing device data with server data is a very common occurrence in this space.  As a simple example, you should be able to synchronize your ‘contacts’ database on your cell phone with a server repository.  Recently, Mike Brey, Raghu Nambiar and I proposed  a “strawman”  IoT benchmark – I think extending your work to include large-scale data synchronization would be worth considering. Please see http://link.springer.com/chapter/10.1007%2F978-3-319-20233-4_4.  
+ 
+I’ve bcc-d Dave, Mark, Mike and Michael.
+ 
+Thanks and warm regards.
+ 
+Ashok Joshi
+Senior Director, Oracle NoSQL Database, Berkeley DB, Database Mobile Server
+508-329-3361