|
Post by louisxiv on May 29, 2013 17:49:00 GMT
I spent some time yesterday working on the script to make it an all-inclusive single pass. Tonight I'll try to get it working...
So I'll update the page when either i) the new script works, or ii) I give up and do it manually.
|
|
|
Post by snipertom on May 29, 2013 17:56:14 GMT
Re: script- yay!
I found this while trawling through the archives by the way:
"Tom: Both Zimmy and Gamma will be back, and maybe we will learn more about them. Such as, how did Zimmy know Reynardine was more than a toy? Why don't Zimmy and Gamma speak a word to each other throughout the whole chapter? Why were they still fully dressed in the middle of the night? All this and more... sometime in the future.
Also, Annie is a harsh mistress."
!
|
|
|
Post by louisxiv on May 29, 2013 20:46:41 GMT
Script is looking good: I'm testing on the Misfile index, because it is a bigger set of data, and I'm at the point I'm squashing more bugs in the data than the bare bones processing.
(Note to self: add input data sanity checking sometime...)
One of the fun things about indexing is spotting things on the requisite careful re-read. Mostly makes up for the tedium, I find.
"Also, Annie is a harsh mistress."
Does that mean she's going to Moon someone???
|
|
|
Post by louisxiv on May 30, 2013 21:00:46 GMT
Character index updated with snipergirl's latest, to Chapter 5's bonus page. Chunks of the page are generated automagically by gun2.py, though I'm still having to tweak some of the output - there's an irritating blank entry I haven't hunted down yet... Still, far better than doing bits by hand. Onwards and upwards!
|
|
|
Post by snipertom on May 30, 2013 21:50:50 GMT
woohoo! Will transcribe more today/tonight Seeing as Tom's comments are often interesting/insightful, what do you think of using the "@" and putting tom's comments after them for each page? I'm also thinking of maybe dropping one of the "#<hr>"s on the bonus pages or highlighting their bonus nature in some other way? Also, I'm (finally) learning jQuery, which gives me an idea for automating the css-swapping stuff with the keywords; basically you could alter each keyword/mention's <span> element so that its class is "tag" and "[the keyword]"; then whenever you click that keyword, you make it so that a "highlight" css class is attached to each instance of the keyword, and any lines without that keyword are hidden. In addition it means that you could select *multiple* tags with some work! i'll work on an early prototype EDIT: chapter 6 is up on github and part of chapter 7, as is the newest cover page and updated instructions. 136/1204 done which makes it 11.3%!
|
|
|
Post by louisxiv on May 31, 2013 16:30:32 GMT
I've suspected that if I knew any js at all (and understood the DOM, etc) cleverer things could be done with the tagging, but getting as far as I have was "monkey see, monkey do" in the finest traditions of cut'n'paste programming And I'm failing entirely to work GitHub - it can't possibly be this difficult...
|
|
|
Post by snipertom on May 31, 2013 16:46:18 GMT
You can download a program for win/mac/linux, then you run it, click on your name below "github" on the far left of the program, click on the repository in the middle that you want and click 'clone' and that will give you a copy on your HDD. Then you can edit the files yourself, and when you are ready to submit your changes, you save the file, go into your github program, which will show that you have unsaved changes in whichever file(s), you put in a commit message about what you have done in the right hand area, then click "sync" and it should upload. hope that helps! Here is a github tutorial- one i sorely need to take myself :/ try.github.io/levels/1/challenges/1basically the way you designed the original site makes it really quite easy to snap in the jquery stuff, which is the awesomes . i was going down the road of thinking that it was going to require php & mysql stuff which was going to be hella slow. this is much better! until about a week ago i had *no* idea how to use javascript to manipulate complex webpages. it seems it is really not very good at that, and using a library is way better!
|
|
|
Post by snipertom on May 31, 2013 18:48:09 GMT
|
|
|
Post by louisxiv on May 31, 2013 19:46:32 GMT
GitHub: it was working fine a few days back when I first set it up, as you suggested, and forked your repository, but I couldn't get your current state over the last couple of days. For a while I just thought I was misunderstanding something. Eventually I started working through the various examples using the command line so I knew exactly what I was asking git to do, and blow me, it started reporting errors in accessing my and your repositories. So I've deleted my repository and will start from scratch as soon as I can kill my local copy. RTFM time - the GIT Book looks cool. Meanwhile I'll do a text grab of your current state and update the page.
|
|
|
Post by louisxiv on May 31, 2013 20:03:59 GMT
W00t indeed. So doing it that way you don't have to dick about with zillions of alternate stylesheets I see. That's good, cause I think that is what really hits the performance on my iDevices. Mind if I steal that handler.js and try it on Misfile? Is handler.js (and the googleapis link) that all I need?
|
|
|
Post by snipertom on May 31, 2013 20:11:35 GMT
you'll need to call on jquery.js (as linked in the head), use my version of event.css and also have a dig through the changes I made through the html- which is mostly getting rid of things, but also changing the links into spans and some other such things! I think (hopefully) it should lead to your python scripts also being shorter and easier. and definitely go for it and use it! Re: github, what I would do is delete your repos. When you re-run the program it will be all "wtf your repo isn't there", and then you just re-clone it/find a new target for it. I'm not sure what badness the fork/merge thing caused but it did it to my repo too and in the end I just deleted and redownloaded and it was fine
|
|
|
Post by snipertom on May 31, 2013 20:12:19 GMT
LAUGHING ON LINE I am tagging the symbolism in treatise 1 right now...
|
|
|
Post by louisxiv on May 31, 2013 22:23:57 GMT
Updated - into Chapter 8, thanx to snipergirl
|
|
|
Post by snipertom on Jun 1, 2013 11:12:41 GMT
louisxiv, just wondering, when you cloned the repo, did you clone the snipergirl/gunnerkrigg or the louisxiv/gunnerkrigg ? also wondering how you are uploading at the moment (as it's showing up as pull requests at the moment) also- yay thx for the most recent city face & treatise For now, Morpheus is taking odd numbers forwards, I'm taking even numbers forwards. Off to do chapter 10!
|
|
|
Post by louisxiv on Jun 1, 2013 12:06:18 GMT
I'm editing direct on GitHub presently. Haven't had the necessary resolution to face anything more complex yet
I'll work backwards for now, though that's probably me done for a few hours - I'll update and upload my page. I haven't switched to your .js yet; I'll do that later today when I've grabbed your event.css changes.
And: Welcome Morpheus... blue pill or red pill?
|
|
|
Post by louisxiv on Jun 1, 2013 14:00:39 GMT
Character index updated: 188 pages covered, out of 1204. Slight layout tweak to better accommodate the 'tags per page' load.
|
|
|
Post by snipertom on Jun 1, 2013 20:01:19 GMT
at some stage we're going to have to find a way to organise the tags aren't we... I'll find a way to jquery a breakdown by starting letter for now?
|
|
|
Post by louisxiv on Jun 1, 2013 20:52:47 GMT
I was wondering if switching the css to bootstrap, with all its bells, whistles and particularly show/hide features might address that, but
1) It is a bit big, to the extent of overkill
2) Personally I don't mind scrolling down long lists (portrait monitors for the win!) and find opening pre-collapsed list sections mildly irritating.
|
|
|
Post by louisxiv on Jun 1, 2013 20:54:37 GMT
Or we could class the tags: different colours for people, objects, locations?
|
|
|
Post by Per on Jun 1, 2013 21:09:53 GMT
The "terrible" tag should be grey.
|
|
|
Post by louisxiv on Jun 1, 2013 21:14:14 GMT
woohoo! Will transcribe more today/tonight Seeing as Tom's comments are often interesting/insightful, what do you think of using the "@" and putting tom's comments after them for each page? I'm also thinking of maybe dropping one of the "#<hr>"s on the bonus pages or highlighting their bonus nature in some other way? The <hr>s work nicely for chapters - an entirely fortuitous discovery as I didn't realise they would not be hidden - I think it is because they are block elements and terminate the 'containing' paragraph. Really should throw up a validation error, imho - but over use just makes the page listing look fragmented - I prefer your use of bolding for flagging the special pages. Tom's comments: do we need a special solution? If a particular comment is worth including I'd just add it on to the end of the comment: # Annie loses her temper with ... rhubarb, rhubarb, et cetera<br>Tom: Annie get your gun.
|
|
|
Post by louisxiv on Jun 1, 2013 21:15:40 GMT
The "terrible" tag should be grey. There's only one thing to be tagged with it...
|
|
|
Post by snipertom on Jun 1, 2013 22:19:15 GMT
The alphabet searching thingy is now live on my demo site snipergirl.com/gunnerkrigg - basically it will by default show all tags but if you want to filter something out quickly you can just click on each letter, and click "show all" to show everything again (which also shows all comic pages). things that start with a number don't work yet but don't need to. I've also put "demo.htm" in the github repo along with event.css and handler.js so it's more obvious what i'm changing and why! I totally agree with dropping the hr's from the bonus pages! I think the bold looks *much* better! Re: tom's comments, that makes sense. Once we have all the other pages up and running, it should be "easy" enough to go back and add his comments in. I'm thinking about tagging the chapter pages as well, and finding a way of also putting up what its icon on the archives page is, somehow? Also did you want me to have a look at your gun2.py and help you come up with a gun3.py for the jqueryised version or are you content to be THE PYTHON MASTER (i am not a python master, only an acolyte)
|
|
|
Post by louisxiv on Jun 1, 2013 23:06:20 GMT
WTF: new interface just ate my reply (with a bit of help from me *sigh*)
|
|
|
Post by louisxiv on Jun 1, 2013 23:23:03 GMT
Also did you want me to have a look at your gun2.py and help you come up with a gun3.py for the jqueryised version or are you content to be THE PYTHON MASTER :D (i am not a python master, only an acolyte) Nothing much to do to gunN.py - just change a few strings. There is no substantive difference in the processing AFAICT. I may take the opportunity adapt it to your index / tags split, as the script is actually set up for a combined file which gunN.py separates appropriately. This setup is cool with my editor's autocomplete feature and allows me to define new tags on the fly when they don't autocomplete - i.e. are new. Nor am I any sort of snake charmer - this is just a small practical exercise in a language new to me. I'll put gunN.py up on GitHub when I'm confident of what I'm doing there.
|
|
|
Post by snipertom on Jun 1, 2013 23:34:09 GMT
Cool cool! *whew*
|
|
|
Post by louisxiv on Jun 2, 2013 0:21:19 GMT
|
|
|
Post by snipertom on Jun 2, 2013 2:29:49 GMT
It's looking great! And yay for the new pages
|
|
|
Post by spritznar on Jun 2, 2013 4:49:00 GMT
apparently i'm attempting pg 844 onward? assuming i'm doing this right... (well if i'm not doing it correctly now i'm sure i will be eventually)
|
|
|
Post by snipertom on Jun 2, 2013 8:58:26 GMT
Sure! Sounds great
|
|