7:30 – 4:30 SR

• DB backups
• Financial reporting
• JavaScript
  • Working on the WebGL container
    • Add a clock – done
    • Set for 2D/3D
    • background color/image
    • Overlay – done
    • Dprint – done
    • use defaults (ambient/directional light, perspectives, etc)
    • Show/hide stage
    • Add models

8:00 – 5:00 SR

• I was informed that passed my comprehensive exams yesterday!
• DB backups
• Dumped the _reqonciler_processing table for dong.
• Started to build up the threejs/YUI charting package. Step one, make all the pieces work together.
• Behold! 
• That being said, there are some issues in the loading? Need to look into that some more. Still, not a bad day’s work.

8:00 – 5:00 SR

• DB Backups
• We are now secure from Heartbleed, and I know my level of responsibility as tenant.
• Looking into scope and closures a bit more. I am getting close to understanding the Module Pattern, I think.

8:00 – 10:00, 12:00 – 4:00 SR

• DB backups
• webapps backups
• The server seems to be running fine after the restart

Class from 10:00 – 12:00 today

8:00 – 5:00 SR

• DB Backups
• Dong discussed Reqonciler with Lenny
• Spent some time chasing down what to do about heartbleed. Since I admin only the VMs, and it appears that the landlord is doing regular updates, I’m trying to determine if I have to do anything in addition to what I’m already doing. Once more, it would really help if we had a real sysadm on the team.
• Today’s big JavaScript breakthrough.The following have the same results, though the Internet says that bind() has more overhead and takes longer.

  // Typical way to handle callbacks
  var self = this;
  setTimeout(function() {self.callbackFn()}, this.delay);

  // More clear?
  setTimeout(this.callbackFn.bind(this), this.delay);

  Interestingly, you can move the inner function out of the callback and into the “Constrictor” (e.g. combination inheritance) so that the function only gets created once:

  function MyClass(){
  var self = this;
  this.boundFunc = function(){self.callbackFn()};
  }

  This allows for a nicer looking call:
  setTimeout(this.boundFunc, this.delay);

  On the whole, I think I like the last pattern the best.

8:00 – 4:00 SR

• DB Backups
• Added notes to the VizTool main page about the current state of the bugs and fixes
• Linked Lenny’s help document to ReqAssistant.
• JavaScript
• Spent some time digging around in WebGL and found out how to set up the Structure view. Much better.
• Also figured out the code inspector. Also very nice. I’m beginning to see why this editor gets high reviews.
• And don’d forget this: http://learningthreejs.com/

8:00 – 5:00 SR

• Somewhere out there are a bunch of accountants taking a breather.
• DB Backups
• Deployed a new RA. Lenny wasn’t there to check it out, though it seems to be working correctly.
• JavaScript – Back to WebGL. And yes, the code makes much more sense now.
• Worked through the interaction chapter. The first example is actually pretty close to the chart concept,
• Starting on the Combining HTML/WebGL. First example gives tooltip and data options.

8:00 – 5:00 SR

• DB Backups
• Sent Lenny a bunch of links for security workshops and such.
• Put together the querylog script and added in the excerpt. Works fine now. Also took out all the lines with “lock” in them and scheduled it for a once-per-day run.
• Documented what I learned in yesterday’s episode of “Eclipse Hates being a Webserver”.
• JavaScript
  • Started my JS cheat sheet.
  • Ok, now that I think I know more of what’s going on in the example code, back to WebGL

8:00 – 5:00 SR

• Db backups
• Look into truncating queries for querylog script and how it’s keeping tables from being made.
• …and my Eclipse server went kablooey again. Spent the morning rebuilding that. Still no real idea what broke or how it was fixed. Grr.
• The truncated queries work fine here.
• Asked Dong to look into extending the query-log with additional information that can be the basis of some auditing/alerts. The idea is that after one or more queries are run to produce a result such as adding a MIPR or inputting PM_Actuals or even FMPs that additional lines are added to the query_log immediately after to add more directly usable information that we can use for reports.
  • All of these additional ittems use the table name “META%” so they can easily be pulled in a query.
  • The “type” field has an overall name for the activity like “ADD PROJECT”
  • The query contains useful text, rather than a query: i.e. “Added Project Carla Demo”. Short enough to keep the text string under 255 chars.
• JavaScript
  • Working through the Module pattern and having issues with either typeof or my debugger.

8:00 – 5:00 SR

• DB Backups
• Updated FY13 and FY14 slides and slide shows.
• JavaScript
  • A new pattern to get used to:

  • var myGlobalScoped = 0;

    console.log("Hello Block Scope " + myGlobalScoped);
     (function (start, stop) {
         for (var iBlockScoped = start; iBlockScoped < stop; iBlockScoped++) {
             myGlobalScoped += iBlockScoped;
         }
     })(0, 5);
     console.log("Hello Block Scope " + myGlobalScoped);

  • Because JS doesn’t understand block scope.
• Gotta find out who isn’t filling out their forms in RA.

• DB backups
• Deploy new VSS jars
• Deployed new FA and RA
• Webapps backups
• Goals were coming back wrong in the briefing queries. This turned out to be because there needed to be one more item in the ‘group by’ condition.
• Changed the briefing query so that a scratch table is produced first, and then all queries are run against that.
• JavaScript – struggled with the console variable for a while. it looks like Fiddler might have stolen it. removal and restart. My mantra. Still, looking at how the app variable is put together is most interesting….
• The Surprisingly Elegant Javascript Type Model

8:00 – 5:00 SR

• DB backups
• Deploy new VSS jars
• Webapps backups
• In the main RA view, projects appear to be duplicating, or at least showing up on multiple lines. According to Pat, editing one duplicate does not affect the other. I tried it on the demo project, and with a refresh, everything seems to coordinate.
• Probably need to add ‘:;()$#<>?/’ to the regex white list. Check with Dong about any SQL issues. Added in. Also discovered an SVN conflict that had to be fixed, which probably breaks the FA interaction. Committed and built for deployment tomorrow.
• Worked my way through the OO chapter of Pro JS for WebDev again, and it made a lot more sense. The WebGL book is definately using Combination inheritance. And now YUI.extend makes a lot more sense.

8:00 – 5:00 SR

• DB backups
• Add ‘.’ and ‘-‘ to the scripting regex
• 10:55 – on hold
• JavaScript – working my back through the constructor/prototype patterns in Pro JS for Web Developers. I am not getting how the WebGL Sim.js objects hang together.

8:00 – 10:00, 1:00 – 5:00 SR

• DB backups. scripts are still running
• Found a library that supports databinding with javascript: http://knockoutjs.com/index.html
• Found a java rues engine that looks better than drools: http://www.jessrules.com/jess/index.shtml
• It looks like the queries that produce the briefing slide shows are broken. I’ve printed off screenshots, but for example, the goals are up and down and sometimes not there at all.