Recent changes to bugs

Handle probe_id header in upload file

Ben Dalziel — Wed, 14 May 2008 19:07:12 -0000

Currently we vomit

It thinks the probe_id header is a sample, and looks for data there...

File system doesn't handle period characters

Ben Dalziel — Wed, 14 May 2008 18:54:50 -0000

Currently looks for first . and decides if the next token is "csv", for example.

Should look for the last .

TCP/IP connections should be enabled in MySQL setup

Ben Dalziel — Thu, 08 May 2008 16:16:08 -0000

I the tutorial I said that tcp/ip connections should be disabled. I was being a little overzealous, and this blocked XMAS from communicating with the DB.

Persist user defined file location between sessions

Ben Dalziel — Fri, 02 May 2008 20:32:49 -0000

This is currently hard coded which makes life easy, but will be a big deployment problem.

Should allow the user to define file locations

Redesign file system

Ben Dalziel — Fri, 02 May 2008 01:21:45 -0000

Dependency on local system i.e. C://xmas_files...

Should be moved to web server folder - tackling this now as it's a big road block for deployment. Should be associated with a similar bug to update the tutorial

here it is - update tutorial also!

Trajectory file generation slow

Ben Dalziel — Fri, 02 May 2008 01:10:17 -0000

Only a little slow - i've done some work to improve this, but should be bettered by bulking the SQL queries.

Rendering slow due to node cardinality lookup

Ben Dalziel — Fri, 02 May 2008 01:09:15 -0000

Description tells the story. For each node, a SQL query is made = slow.

This query is particularly slow when filtered. Should be addressed.

Re-enable interestingness measures through the system

Ben Dalziel — Fri, 02 May 2008 00:59:14 -0000

These were lost through the pathway integration re-design. The good news is that there is a huge improvement in the versatility of the data entering these measures. The down side is that the subsystem has to be rewritten.

De-couple presentation and annotation data layer

Ben Dalziel — Mon, 28 Apr 2008 20:02:55 -0000

Currently the renderers know what values to expect for a probe - should generalize this to accommodate any data about a probe...

Make process of annotation integration simpler

Ben Dalziel — Mon, 28 Apr 2008 20:00:43 -0000

This is a big draw back right now - the process to integrate annotations is poor. Should be more guided and customizable. Should expose the process further to make the process more transparent.

Once annotations are integrated their should be a similar DB insight tool suite like we have for primary data