the scrum (2/19)

[Jon Kiparsky]

Been working on the directional problem tonight  - some interesting problems are coming up, I'm going to sleep on it and get something in tomorrow evening. I hope this will allow me to check in the third leg of the grouping algorithms as well.

[Andrew.S...@bbh.com]

Guys, this probably isn't the best news, but I think there may be a bug with the Applet.  I was doing some work on my convex hull stuff and I kept seeing this error message popping up in the console when I would start the AdminApplication:

 

Error loading tree: Craine, Sasha Linn

 

So I went in to investigate and here is what her tree string looks like:

O:0:Bat:791:108:true:():():#
O:1:Beetle:810:531:true:():():#
O:2:Bird:795:236:true:():():#
O:3:Butterfly:809:496:true:():():#
O:4:Crocodile:801:276:true:():():#
O:5:Fish:801:437:true:(79,):():#
O:6:Horseshoe Crab:639:553:true:():(70,):#
O:7:Human:808:33:true:():(28,):#
O:8:Jellyfish:341:278:true:():():#
O:9:Leech:697:326:true:():(69,):#
O:10:Lizard:801:358:true:():(69,):#
O:11:Millipede:809:603:true:():():#
O:12:Rat:803:145:true:():():#
O:13:Shark:792:394:true:():(79,):#
O:14:Snail:617:417:true:():(70,):#
O:15:Spider:809:567:true:():():#
O:16:Squirrel:805:184:true:():():#
O:17:Starfish:618:459:true:():(70,):#
O:18:Turtle:797:318:true:():():#
O:19:Whale:801:69:true:():(28,):#
E:20::37:556:false:():():#
E:21::707:141:true:():():#
C:22:21:0#
C:23:21:12#
C:24:21:16#
E:28::557:84:true:(7,19,):():#
C:29:28:21#
E:36::726:253:true:():():#
C:37:36:2#
C:38:36:4#
E:40::739:556:true:():():#
C:41:40:3#
C:42:40:1#
C:43:40:15#
C:44:40:11#
E:46::412:281:true:(69,):(70,):#
C:47:8:46#
E:48::674:279:true:():():#
C:49:48:18#
C:50:48:36#
C:51:46:28#
C:56:48:46#
E:69::584:359:true:(9,76,10,):(46,):#
E:70::514:460:true:(6,17,14,46,):():#
C:70:46:69#
C:71:69:9#
C:72:69:76#
C:73:69:10#
C:74:28:7#
C:75:28:19#
E:79::746:417:true:(13,):(5,):#
C:80:6:40#
C:81:70:6#
C:82:70:17#
C:83:70:14#
C:84:70:46#
C:85:79:13#
C:86:5:79#

 

The line that is causing the error message is "C:72:69:76#".  This is causing a problem because this is supposed to be a connection pointing to the elements 69 & 76, but element 76 does not exist in the tree.  After looking at that I also saw that there were two element #70's in her tree.  This is not good, she must have figured out some way probably using undo to create this bug, but we good enough error control that it doesn't crash the applet or the admin applet.

 

One thing that I noticed is that the display of what she has comes across a lot better in the AdminApplication appliction you guys wrote, because the way trees are setup in the admin applet some of her connections are lost in the parsing.

 

I don't know what we can do about this.  Once again if we had some way to get logging in this applet we could probably fix this, but otherwise we have no idea what the order of operations was to get to this point in her tree.

 

Just wanted to let you guys know about this.

Andrew Schonfeld
Brown Brothers Harriman & Co.
Infomediary Systems
617-772-6738

*************************** IMPORTANT NOTE ***************************** The opinions expressed in this message and/or any attachments are those of the author and not necessarily those of Brown Brothers Harriman & Co., its subsidiaries and affiliates ("BBH"). There is no guarantee that this message is either private or confidential, and it may have been altered by unauthorized sources without your or our knowledge. Nothing in the message is capable or intended to create any legally binding obligations on either party and it is not intended to provide legal advice. BBH accepts no responsibility for loss or damage from its use, including damage from virus. ************************************************************************

[Jon Kiparsky]

Of course there's a bug in the applet. There are probably several. This is interesting, not fatal.
Post it as a bug, include the tree file with this commentary, and we'll figure out how she managed to do this. Be sure that version info is included, so we can work from the right version.

The important thing is to not dive in and start stirring the pot in the applet - this is a weird and a fragile bug, and probably reveals something more serious about our handling of  something. (what? your guess is undo, that's reasonable- but maybe something else) If we change stuff in response to this before we find the problem, we won't know what we've fixed, if anyhting. So let's carry on with what we're working on and add identifying and fixing this bug to the next sprint.
Sound good?

--
You received this message because you are subscribed to the Google Groups "tbs-group" group.
To post to this group, send email to tbs-...@googlegroups.com.
To unsubscribe from this group, send email to tbs-group+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/tbs-group?hl=en.

[Andrew.S...@bbh.com]

I like it.  Thats what I was planning on doing, cuz after assessment of the bug this would require a lot of work & like you said, its not a showstopper.  They still recieved their 15 points and the AdminApplication is correctly displaying the graph & analyzing it.  The only thing I may look at is the display of the tree in the admin applet so it looks like what is being displayed in the AdminApplication.  I'm not going to change anything just analyze how it could possibly be done.  I'll post this bug.

Andrew Schonfeld
Brown Brothers Harriman & Co.
Infomediary Systems
617-772-6738

-----tbs-...@googlegroups.com wrote: -----

To: tbs-...@googlegroups.com
From: Jon Kiparsky <jon.ki...@gmail.com>
Sent by: tbs-...@googlegroups.com
Date: 02/20/2010 12:17PM
Subject: Re: [tbs-group] Possible Bug in Applet

[Andrew.S...@bbh.com]

Today: Spent most of today reading works from Raymond and Spolsky. For tomorrow's class.
Tomorrow: Jon, I forgot to ask you what Professor White said when you met with him last thursday. Was he around?  What did he say?  Also, I need to finish up my work on the additions to the analysis prompt in the admin applet.  I also want to discuss possible solutions for maintaining the integrity of a tree, after the discovery of bug #84.

Roadblocks: none

Andrew Schonfeld
Brown Brothers Harriman & Co.
Infomediary Systems

617-772-6738 *************************** IMPORTANT NOTE*****************************-- The opinions expressed in this message and/or any attachments are those of the author and not necessarily those of Brown Brothers Harriman & Co., its subsidiaries and affiliates ("BBH"). There is no guarantee that this message is either private or confidential, and it may have been altered by unauthorized sources without your or our knowledge. Nothing in the message is capable or intended to create any legally binding obligations on either party and it is not intended to provide legal advice. BBH accepts no responsibility for loss or damage from its use, including damage from virus. ********************************************************************************

[Jon Kiparsky]

No work on the applet today, but I did have a serendiptitious encounter with a trio of White's intro bio students. I asked them about the applet, and got some feedback.
Main thing they mentioned that I'll pass on here: they wanted selectable text in the question prompts.
Not to say we have to do it, but it was one of the things they mentioned first off, and came back to.
I'll put it in as an enhancement on the issues list, we can decide later if we want to get to it.

Tonight/tomorrow: don't know, depends on progress of compilers homework.

--

[Glenn Thelen]

Today: Working on the readings/essay for tomorrow's class
Tomorrow: Won't know until after class
Roadblocks: None

---

From: "Andrew.S...@bbh.com" <Andrew.S...@bbh.com>
To: tbs-...@googlegroups.com
Sent: Mon, February 22, 2010 11:32:22 PM
Subject: [tbs-group] the scrum (2/22)

--

[Andrew.S...@bbh.com]

Ahhh yes, selectable text!  This idea definitely came up while I was putting enhancements on Glenn's original cursor code for text editing.  Its an avenue to explore in the future possibly.

Andrew Schonfeld
Brown Brothers Harriman & Co.
Infomediary Systems
617-772-6738

From:	Jon Kiparsky <jon.ki...@gmail.com>
To:	tbs-...@googlegroups.com
Date:	02/22/2010 11:42 PM
Subject:	Re: [tbs-group] the scrum (2/22)
Sent by:	tbs-...@googlegroups.com

---

[Andrew.S...@bbh.com]

Today:  I have added the 225 student data rows from production to Professor White's test server so if you log into the http://cluster.bio.whe.umb.edu/cgi-bin/Test/TBSTestSurvey.pl Administrator you will see them.  I have also deployed the latest version of the applet with the new Organism group hull toggling feature.  Check it out if you can.

Tomorrow:  Working on the bug #84 from google code and also adding more of the graphing application functionality to the admin applet's analysis prompt.  Maybe add database connection functionality to the adminApplication for the test server.

[Glenn Thelen]

Today: Spent some time looking at the student trees.  I noticed quite a few sets of trees that show some good understanding, but will fail all of our current tests.  I put having human scorers enter data about tree structure under issues as an enhancement.

Tomorrow: Do some work on scoring.  I would like to make some educated guesses about some of the trees, such as if there are no links but organisms are placed right next to each other, to assume the student meant to link them.

Roadblocks: None

---

From: "Andrew.S...@bbh.com" <Andrew.S...@bbh.com>
To: tbs-...@googlegroups.com

Sent: Tue, February 23, 2010 9:59:59 PM
Subject: [tbs-group] the scrum (2/23)

--

[Andrew.S...@bbh.com]

Today: I tried setting up database connectivity to Professor White's test server for the AdminApplication, but unfortunately he does not have a port open for outside connections.  I will continue to write the code, but it will only work from the computer that MySQL is running on.

 

Tomorrow: Continuing to work on the bug #84 and updating the analysis Prompt.

 Roadblocks: Lack of outside database connectivity on test server.

Andrew Schonfeld
Brown Brothers Harriman & Co.
Infomediary Systems
617-772-6738

[Andrew.S...@bbh.com]

Today: finished database connectivity code, started tree integrity debugging code. 

Tomorrow: Continuing debugging and working on any display issues with AdminApplication.  I need to test the current version in depth, last time I looked there were some small things I know I could take care of.

 

Roadblocks: Got stuck at work late today.

Andrew Schonfeld
Brown Brothers Harriman & Co.
Infomediary Systems
617-772-6738

[Glenn Thelen]

Today: Researching opening up another window for the student data table.

Tomorrow: More UI stuff and/or scoring

Roadblocks: None

---

From: "Andrew.S...@bbh.com" <Andrew.S...@bbh.com>
To: tbs-...@googlegroups.com

Sent: Thu, February 25, 2010 11:50:49 PM
Subject: [tbs-group] the scrum (2/25)

--

[Jon Kiparsky]

Today: working on including current version info in the build process. Learning some perl tricks.
Tomorrow: ?

[Glenn Thelen (Yahoo)]

Today: Doing some research on UI stuff / scoring
Tomorrow: Don't know
Roadblocks: None.

[Jon Kiparsky]

Today: Finally stopped trying to do it the hard way and figured out that svn has a facility for adding stuff like license info. Issue #5 is done.
Tomorrow: more documentation stuff. Will try to return to the tree direction problem as well.
Roadblocks: own thick skull - RTFM, Jon.