User:Lcarsos
The username is lcarsos. pronounced el-cars-oh-ess. The only other acceptable capitalizations are LCARSOS or LCARS OS. The second of those is frowned upon because that should be used for any project attempting to create an LCARS operating system, or overlay/window manager. And lastly, never Lcarsos, despicable.
I'm just a big xkcd and Star Trek fan (props if you got the reference).
Here, I'm just working backwards from the beginning to create missing pages, fix existing pages, updating them to the current header, adding categories, fixing double redirects, etc. I also try to find as much work for the admins as possible (just kidding).
Contents
Personal Achievements
Some of the things I've managed to do here, I'm very proud of. And by golly, I'm going to display them out in public so everyone else can see how vain I am.
Named Characters
- Ponytail - Brought back the name from a dimly remembered past.
- Danish - stumbled across 377 and suggested it to Blaisepascal to be used as her official name
- Miss Lenhart - doesn't really count, I just created all the pages for her. But I'm very proud that I managed to do it all correctly on my own.
Things I Need to Remember
I'm currently at 993: Brand Identity
- Categories I forget
Ruby Importer script
Work on my import-assistant script is progressing. I'm using it to help create pages as I work my way backwards. Currently it is just using information from xkcd's json api and leaving explanation open for you to get.
I'm working on getting the explanation to pull in. But, managing to guess the right url alone was a difficult task, and it still doesn't get it right for some of the more odd ones, but it would be faster to just find the page manually than code for each exception to the rule. I'm going to do a bit more work on that, but I might just give it up and keep manually copying explanations in. Or at least write out a best-guess of what the url should be to get you close, if you aren't walking backward through the blog entries.
I want to be able to pull in the explanation so I can scan through it to get a rough idea of what characters appear to automatically generate Comics featuring ___ categories. But that's a bit of work, including lots of regular expressions (!) but it's good practice for basic data mining and pattern recognition.
I've found that already I'm much more productive simply by not having to type in the comic template info. And the page feels much more finished when it has the transcript.
Now, I'm going to give you a link, but I am requiring you to manually go through and look at what it generates and keep yourself engaged. As I explain in the README, there are things the human brain can do innately that computers need lots of work to get the logic hooked up.
You must go through and make links to appropriate wikipedia and external web pages. I can make links to the links that exist in the html, but you have to be a human and realize that if something has a wikipedia page it probably needs to be linked to.
Please tell me on my User Talk: Lcarsos#Ruby Importer page if you are using it, what features you want, and bugs you find.
Link
https://github.com/lcarsos/explainxkcd-importer/
Usage
- ./importer.rb <number> ...
- ./importer.rb 1000
- ./importer.rb 1000 1001 1002 1003 1004
- ./importer.rb [1..600] (I haven't verified if this works correctly)
It accepts any number of comic numbers as arguments (space delimited, duh) and will iterate through all of them and spit out a text file <number>.txt with Unix line-endings (if you open it up on Windows use Notepad++ or some other real text editor).
If I see you using this tool to blindly create pages without going through and cleaning things up I will kill you.
- I assume that ./importer.rb 100{0,1,2,3,4,5,6,7,8,9} would also work? Blaisepascal (talk) 02:08, 26 August 2012 (UTC)
Known Issues
- Active Development Work
- Explanations are not being culled from the blog page.
- Noted As Feature
- Some transcripts come from the json api with extra newlines. Will not be fixed to encourage humans to look everything through. Good for fingerprinting brain-dead copy/paste usage.
