User:Sdowne/summer 2010 worklog july

Thursday July 1st

• did frame rate fix lowethird and google news div fix

Friday July 2nd

Saturday July 3rd

Sunday July 4th

Monday July 5th

• Added the user's name to the flickr div here this was a simple matter of finding the right variable from the flickr stream json
• Added a colour flash to the div when a popcorn event fires here and grayed it out when it's not active.
• Looked into jquery's animate method, need something to grow and shrink the overlay image when they appear and disapear, but I may just use fade in and out.

Tuesday July 6th

• fixed some html errors in the popcorn demo. Fixed a JavaScript error in the popcorn library here
• Decided fade in and out were easier than grow and shrink, changing the size would do weird things, and display them as a block, and not inline. Also, growing htem from the middle wans't work, and just from bottm right or top left, decided it wasn't worth it and that fade in and out look just as good.
• Added a fade in and out to popcorn overlays and hard coded a default image link here, the default can be overwritten with overlay="link/to/new/png"
• Started finishing the credit's data. Finished here
• Started debugging popcorn with jslint, jquery $ is a problem, will look into it more tomorrow.

Wednesday July 7th

• Blogged about popcorn.js
• CDOT meeting, demo'd Popcorn.js
• Worked on Processing.js again, and did three small tickets remove instance arrayList set and arrayList add
• My two arraylist tickets I did today were put under review needs work, but I think I may disagree for the reasons, being that my patch supports everything p5 does, but doesn't support what it doesn't do, meaning, I don't support its errors, just its functionality. I can add the errors easily, but I don't think it's the right thing to do and need to get a second opinion. I spent some time today talking to Pomax about this on irc, and did some testing and thinking on it, and I still think we should allow the quirks of javascript to shine through processing.js. An example would be in p5 when you use ArrayList.set(-1, "item"); an error is thrown on the -1 not being an index, but in processing.js, the -1 becomes the second last index, which, I know is not exactly how p5 works, but why bother adding it's flaws?

Thursday July 8th

• Blogged about google translate, made a fun little google translate demo and followed Mozilla summit
• Short day because I went to a funeral in the afternoon

Friday July 9th

• Discussed more on arraylist.set and arraylist.add.
• fixed linting errors in popcorn, and started to asses jquery and direct references to html elements using a "string". I have a work-in-progress patch. fixing the linting errors was harder than I thought, especially with jquery syntax.
• did research on getting a json feed, without using the getJSON jquery function, which is the main reason I added jquery. I actually don't think jquery is much of a problem, but we don't want to have to tell our users to directly add in a <script> other than popcorn, so we have to load the script from inside the popcorn library, which causes a problem because the script has to fully load, and we only want to load it if that script is going to be needed in this xml, for example, if google map is not defined in the xml, don't load the google map script, but if we do load it dynamically, it'll stall the code because we should wait for it all to be loaded before we start running the js. Some sort of scan of the xml, then checking to see which commands are used, load all the scripts, then start the parser... but that doesn't seem ideal.

Saturday July 10th

• Started a demo using processing.js, box2d.js and popcorn.js, where subtitles will fall to the ground, using physics, and crash into a pile.

Sunday July 11th

• Finished my demo demo using Processing.js box2d.js and popcorn.js and wrote a blog about it.

Monday July 12th

• made a blog post on string.split
• Worked on refactoring the popcorn.js code base. The main goal is to continue to support the demo, but allow the library to be more dynamic and not rely on the html elements in the demo.

Tuesday July 13th

• Today I worked more on processing.js.
• Finished 764, 760, 757, 752 and 751.
• started work on 765.
• finished 765

Wednesday July 14th

• started work, and research on createFont

Thursday July 15th

• continued work on createFont

Friday July 16th

• Merged createFont ticket into 771. This made sense for many reasons, the main one being that 771 covers problems in loadFont as well, when can then be wrapped by createFont.
• Continued research and development on createFont and loadFont, and I have it working cept a problem in chrome I have not been able to hammer out.

Monday July 19th

• Today I fixed problems in tickets 752, 760, 764 and 765

Tuesday July 20th

Wednesday July 21st

Thursday July 22nd

Friday July 23rd

Monday July 26th

Tuesday July 27th

Wed July 28th

• I fixed an odd pixel that was slipping in between lines split on a string with newline characters. The extra pixel only existed using TOP, BOTTOM, and CENTER on the y axis. Basically, everything cept BASELINE, which was default... The problem was due to using the text size, instead of using text ascent, which is the same size for smaller fonts, but as the font gets larger, the gap is more noticeable.
• Noticed I totally overlooked the 4th and 5th parameters of text, which are height and width. This also needs to account for text leading using newlines OR spaces that push the text onto the next line. There is also a few pixel padding in the p5 version from the top of the box, to the first bit of text, which is dependent on the size of the font. I am not yet sure where they get this number, but I am looking through p5's source. The last final problem in textLeading() has to do with the number of pixels being padded between newlines and multiple lines.