User:Sdowne/summer 2010 worklog june
Tuesday June 1st
- Worked with notmasteryet thorugh irc on new lightweight parser . I merged the current parser work into a working copy of 0.9.3. This would include the word done on 0.9.1, and anything that was staged in 0.9.3 up to that point. I had to do a few tweaks to make everything work, and I had to fix one small bug with array parameters defined inside a function definition, using the [] after the name, instead of the type caused an error.
Wednesday June 2nd
- Fixed an issue with spaces before the int in a for loop --> for( int i;m in the new parser.
- Set the super review process for the new parser.
- I really did spend a lot of time just trying to read, and understand the new parser.
- Went through what is left for 0.9.3 and picked up a couple tickets 611 and 368
Thursday June 3rd
- Did an initial attempt on 611.
- Did and created a test for 368.
- Continued work with notmaster to get the light parser finished by helping to track down trace issues, and issues with brightness2 example in chrome not user getters and setters as fast as ff.
Wednesday June 4th
- Found an issue with opera not working on linux, as version 10.10 does not yet exist for Linux, and only 10.10 has support for string.trim. Had to add string.trim. The fix was easy... getting there was not.
- Coded a fix for the light parser to use static variables inside classes that have yet to be created into an object. This fixes the neighborhood test. The way it was working before was the static variable would not exist until atleast one object was created from the class, but it should allow use of static variables before that.
- and fixed a few of the smaller issues in the light parser's review.
Monday June 7th
- I did review of tickets 400, 706, 392 and 55. These were all in the same branch, so they were tested at once.
- I reviewed anna's xml ticket.
- Reviewed 596.
- Created a demo for tagging people in html5 videos for tomorrow.
Tuesday June 8th
- Worked with Brett and Dave at Mozilla to do requirements gather for the video.js project. This project is going to be quite small when it comes to code, but working on getting the details of the library just right seems to be something I get hung up on. Not that the finer details should not be talked about, I just have a hard time with it. We did get a lot done, and I did get some good direction about the details of the code, I just feel the large picture is too much for me to grasp, and I need to break it down and focus on one issue at a time. Just how I work best, and it's what I bring to the table. Either way, we won't get this right the first time, and will have to work with Brett and the community as it grows.
Wednesday June 9th
- At Mozilla working with Brett. I spent most my time figuring out how we are going to do the basics of this project. We have three things: 1. an xml parser of some sort. 2. Loading external xml files instead of putting them in the video tag. 3. Wrap everything inside a closure. 4. Still making using this easy form the html perspective. Using this should be simple, and it should be a matter of linking to the video.js library, writing an external xml file, and linking to that file in the video tag as some sort of source. The rest should be handled in the library, while still doing the above step I worry about, like closure, parsing, etc.
- I also did some time talking to Brett, getting to know him, and where he's coming from for this project.
Thursday June 10th
- Today I went over a lot of what we talked about yesterday, and added the ability to grab text from an external file and start parsing it. I lot of what i did was research as I went through 3-4 different methods for parsing, and still not sure which is best.
- Removed the dependency for JQuery.
- Put everything inside a closure.
- Still need to put the actually work, of displaying tags and subtitles back into the new closure, without jquery.
Friday June 11th
- Started work on the new design and here. The new design uses a command pattern.
Sat-Sun June 12th-13th
- This weekend I worked on new structure and got that said ticket up for review on Sunday.
Monday June 14th
- I had a meeting with Dave and Dan about our javascript video library popcorn.js. It went well and I got a better understanding for what I am responsible, and this allows me to have a greater degree of focus for my work. I work best this way.
- I also have my new command pattern finished and working, but after showing it to Dave, I need a few small updates to make it better. Started on those fixes.
- Merged the new work into the old work.
Tuesday June 15th
- Set command pattern for peer review.
- I cleaned up my popcorn branch. The problem was I started over again, and had my work on a new tree, but I didn't consider the previous history for the project, so I cleaned it up, pushed it into the old tree, and updated everything.
- I added some new functionality to the popcorn parser. The parser now does what it did before, but it contains properties passed down by parent elements, along with not using a single regular expressions. Instead, I am using the xml dom object tree, and I transverse through that using a library, much easier.
Wednesday June 16th
- I did instances preload and tag people.
- Did some google map api research.
Thursday June 17th
- I finished geolocation
- I fixed up map implementation in popcorn.js.
- I did some initial research on twitter feeds, and communicated with Brett to get a better understanding of what he wants. This is an on going process.
Friday June 18th
- Today, I added twitter feeds to popcorn.js. here is it.
Monday June 21st
- Today I had a couple meetings with the popcorn team. They were very helpful.
- I reviewed the linting ticket, also made a few changes, and all fixed linting errors.
- I finished time friendly units
Tuesday June 22nd
- This morning there were a few features that needed to be added to time friendly units. I added them and it's back up for review.
- I spent a lot of time researching twitter widgets and the map api. There are a lot of details to work out still in the way we display and clear these addons from their target divs.
Wednesday June 23rd
- I was able to get twitter widgets working quite nicely. Still a few nitpicks I expect in review, and count on, and other features that will go with twitter, like preloading and web workers that need to be implemented, but I got it working and it's feature rich.
- I spent the rest of the day on doing what I did with twitter widgets to the google map api, but it's proving to not like being moved in and out of a div. I figure there is an absolute positioned div above the map to use the map controls, that is not being moved when I move the div the map is contained in... still researching for tomorrow morning, then on to the manifest parser.
- Almost forgot, I also got default in and out working on the time attributes in the xml. I had to give the commands access to their parent video manager to accomplish this, but it makes a lot of sense.
Thursday June 24th
Friday June 25th
- Reassessed the requirements for 0.1 release in this ticket. Then started on one of the tickets I assessed for flickr feeds.
- Added a flickr feed tag to popcorn.js here
Sat-Sun June 26th-27th
- I spent a few hours on the weekend merging and debugging everything we have up to this point, so come Monday we have a solid platform to work from.
Monday June 28th
- Spent the afternoon at Mozilla, working with Brett again. I worked on popcorn image oveerlay. It's not quite complete, but it's very very close. I spent a large portion of the day failing, until the very end, when it came together. I will update the ticket tomorrow after going over it one final time.
Tuesday June 29th
- Finished yesterday's popcorn image oveerlay then started working google news feed, which went good, and is finished.
Wednesday June 30th
- Worked on Popcorn.js.
- Thanks to Dave, I was able to finish twitter title.
- Updated overlay, made it so the overlay is relative to the div the video is contained in, and framework to do this to subtitles relative to video's position.
- Did lots of merging.
- Closed collapse empty divs because divs that are collapsed are the same as empty divs.
