User talk:PanSola

Welcome to the talk page of a devout follower of Lyssa!

= Archives =
 * Past mistakes are moved into the /MistakeArchive
 * Other closed issues are moved to /Archive, /Archive2, /Archive3, /Archive4, /Archive5, /Archive6, /Archive7

= Messages =

Sig
Heya Pan,

GW:SIGN is policy now, and it expressly forbids using template transclusion to sign talk pages. Could you change your sig, please?

&mdash;Tanaric 01:20, 4 April 2007 (CDT)


 * I've changed my sig in my preferences. As for the old sigs, I don't have a bot to deal with them.  Let me know if I am expected to manually get rid of them, I'll find time to do it. -talk 18:44, 4 April 2007 (CDT)


 * There's a heck of a lot of them Pan, over 500! Does that mean you'd find time to write a bot or go through them all! Anyway nice to see ya around again :) --Xasxas256 18:50, 4 April 2007 (CDT)


 * I used to run crazier crusades manually q-: -talk 18:51, 4 April 2007 (CDT)


 * Heh heh that's true, see you're not around enough, I forget these things. Also *cough* GW:SIGN *cough* says no sub or sup tags! --Xasxas256 18:54, 4 April 2007 (CDT)


 * And your user name should be made clear with your sig. (It all reads in the policy ;) ) --[[Image:Gem-icon-sm.png]] (gem / talk) 18:57, 4 April 2007 (CDT)


 * Gah. Testing. -PanSola 19:04, 4 April 2007 (CDT)


 * Sorry, not my fault. *puppy eyes* --[[Image:Gem-icon-sm.png]] (gem / talk) 19:06, 4 April 2007 (CDT)


 * Oooo, eyes! *gauges them out* -PanSola 19:11, 4 April 2007 (CDT)


 * Oooh the new Pan is crazy, good one! Gem, you could have at least made some kind of funky sig image, you know while you still had eye! How about [[Image:PanandSola.jpg|100px]], it's nice and literal! I don't know how well it scales though, where's LordBiro when you need him! --Xasxas256 19:14, 4 April 2007 (CDT)


 * My eyes! My eyes! Oh my god, save me! (Where is Boo anyway? I thought he would be the one who gets me) --[[Image:Gem-icon-sm.png]] (gem / talk) 19:16, 4 April 2007 (CDT)
 * Your eyes have been sacrificed to our new goddess!!! (though I remain a devout follower of Lyssa). -PanSola 23:27, 4 April 2007 (CDT)

I'll change them for you later, Pan. Remind me if I forget. --Fyren 19:23, 4 April 2007 (CDT)
 * Thanks. Somebody else remind me if I forget to remind Fr... Fyren. q-: -PanSola 23:27, 4 April 2007 (CDT)
 * Cool, thanks for quick compliance. That whole policy exists because of you, so it's only fitting we finally nailed you with it. Welcome back, bro. &mdash;Tanaric 00:52, 5 April 2007 (CDT)
 * Ooo, there's an entire policy for me!!! Mwahahahahahahahaha -PanSola 06:25, 5 April 2007 (CDT)


 * That's right, no laughing either, your reign of terror is over, Balthazar rejoice! --Xasxas256 06:52, 5 April 2007 (CDT)


 * Bal the Brute isn't even fit to kiss Lyssa's toenail! -PanSola 07:31, 5 April 2007 (CDT)


 * Bottage will be delayed while I figure out how I broke my bot. --Fyren 02:23, 10 April 2007 (CDT)
 * Ok. Well, as long as I don't get banned for it, take your time q-: -PanSola 02:25, 10 April 2007 (CDT)

Sortable QR
that is by far the coolest thing i have seen all week. --Honorable Sarah 00:54, 12 April 2007 (CDT)


 * And I'm surprised nobody has done it on GuildWiki yet, considering we got the Javascript to do it ages ago (before I came back). There's a minor issue with the current version of the script though, making the spacing around the sort button ugly )-:  -PanSola 00:56, 12 April 2007 (CDT)


 * The whole thing also looks better in Chinese, because characters naturally go in the vertical direction. http://guildwars.wikia.com/wiki/MesmerSkillDescriptionChart  Sadly, most of the data is out of date there, as I don't have the energy to keep it updated. -PanSola 00:58, 12 April 2007 (CDT)

Is there anyway we could get a stable sorting algorithm? With such an algo, it would be possible to do multiple level sorts by just doing them one after the other (for example: sorting by attribute and name -> sort by name, then attribute). From the tests I did, this seems to not work right now, which would mean that it is using an unstable sorting algorithm. I couldn't find the piece of javascript implementation we use, but if you can point me to it, I could modify it into a stable sort. --Theeth (talk)   12:55, 14 April 2007 (CDT)
 * Yes it's possible, it's on my to-do list, but it's much slower so I'm thinking of making it an "advanced" option that is not enabled by default. The javascript I am using right now (which isn't the one you guys see) is User:PanSola/sortable_mod.js.  -PanSola 16:06, 14 April 2007 (CDT)

Try this instead of shaker sort. It's a swapless stable quicksort with non-random pivot. That should fix the speed issue of the sorting. (no guaranty on the errorlessness of the code, it's adapted from some C code I have lying around, my JS skills might be a bit rusty). --Theeth (talk)   16:43, 14 April 2007 (CDT)

static void qsort_data(list, comp_func, head, tail) { pivot = list[head]; ihead = head; itail = tail;

while (head < tail) {		while (comp_func(list[tail], pivot) >= 0 && head < tail) tail--;

if (head != tail) {			list[head] = list[tail] head++; }

while (comp_func(list[head], pivot) <= 0 && head < tail) head++;

if (head != tail) {			list[tail] = list[head] tail--; }	}

list[head] = pivot; if (ihead < head) { qsort_data(t, ihead, head-1); }	if (itail > head) { qsort_data(t, head+1, itail); } }

function qsort(list, comp_func) { qsort_data(list, comp_func, 0, list.length - 1); }


 * I thought quicksort with non-random pivot, by definition, has terrible worse-case performances. O(n^2). -PanSola 17:48, 14 April 2007 (CDT)
 * With presorted list, yes, it's O(n^2) operations and O(n) storage (recursion) whereas shaker sort's worst case is O(n^2) operations and O(1) storage, but average cases are much better with quicksort than shaker. Moreover, the non-random pivot is rather easy to fix, I just didn't want to hunt around for JS randomize functions. --Theeth (talk)   18:33, 14 April 2007 (CDT)

I've been working with some sortable tables in my User space when I came across the work you've been doing. One thing I have been trying unsuccessfully to implement is marking specific columns as non-sortable. I saw that you have a template/function call { {weapon qr top}} that implements this but I can not for the life of me find the article to look at the implementation. All attempts to find weapon qr top return no results. Where can I find this article? I also read in one of the posts that you have a work-around for the current javascript stripping out template/function calls from header rows (my second problem). Would you also let me know where to find information on that fix as well? Thanks! --  Glamtre   (Talk) 13:55, 16 April 2007 (CDT)


 * Template:weapon qr top sets certain header cells to be class="unsortable", which then gets processed by the javascript MediaWiki:sortable_mod.js. It's not exactly documented, but hopefully the code isn't too hard to understand.  Let me know if you have specific questions.  As for the tag-stripping, it's a very tiny modification used in sortable_mod.js also.  The original behavior is in, but my sortable_mod.js completely overrides it. -PanSola 02:07, 17 April 2007 (CDT)


 * Thanks for the info. I totally forgot about the Template namespace and I REALLY hate wiki search!  Are there any plans for your updated code to replace the installed class="sortable" code?  I am happy to implement your code in my namespace for testing, but if it works one way for me and other for someone else I consider that a bad situation. -- [[Image:Axe-icon-left.png]]  Glamtre  [[Image:Axe-icon-right.png]] (Talk) 12:41, 20 April 2007 (CDT)

ERROR
THIS IS AN ERROR!!! = ASODFINADOFSINASDF --Dirigible 05:59, 12 April 2007 (CDT)

JS overriding
Were you able to override the function you wanted? --Fyren 07:49, 12 April 2007 (CDT)
 * Indirectly. I had to override a function higher up the call chain.  I don't entirely understand why, but at least it works right now... -PanSola 07:55, 12 April 2007 (CDT)

CSS changes
You seem to have been making a lot of CSS changes recently. What are you trying to do? I'm not really sure what your goals are, but most of your edits seem to be experiments. This can't be good when you're modifying a global stylesheet :P  &lt;LordBiro&gt;/&lt;Talk&gt; 13:02, 13 April 2007 (CDT)
 * The major changes are non-experimental. The experimental ones are trivial changes (changing order of some lines just to see if MSIE's behavior will change).  After some googling, I've figured out that I'm stuck with a CSS bug in IE7 that I'll have to use javascript to fix. -PanSola 13:39, 13 April 2007 (CDT)


 * Perhaps my understanding is flawed (maybe Fyren could make some suggestions) but I was under the impression that most browsers use a conditional get, so that they only get the contents of a file if it's been modified since a certain time. If this is the case then, trivial or not, the whole common.css file will be downloaded regardless of the change. This was my only concern.


 * And I was curious :) Unfortunately I don't have IE7 but if you're willing to discuss the bug I'd be happy to give my input? Of course, it's unlikely I will have anything to add that you can't find on google ;)  &lt;LordBiro&gt;/&lt;Talk&gt; 16:19, 13 April 2007 (CDT)


 * IE7 doesn't not implement "display:table-cell;", so my clever ith-column-hiding trick gets messed up with IE7 users. The issue doesn't show up with IE6 because IE6 didn't even implement Adjacent sibling selector, kinda like if you don't even have access to a car, you can't get caught drunk driving (though you are supposed to be driving, not drunk). -PanSola 16:42, 13 April 2007 (CDT)


 * That's the worst analogy ever :P And that's saying something with the amount of analogies the build wipe debate has produced! But yes, I know what you mean. That's a pain!


 * There's a separate stylesheet that monobook uses for IE7 hacks, according to the source it's "/skins/monobook/IE70Fixes.css?42a". I realise it's not common.css, but would this be useful for monobook? I imagine editing it would involve a trip to see Fyren in his dungeon over at his really long talk page.  &lt;LordBiro&gt;/&lt;Talk&gt; 18:07, 13 April 2007 (CDT)
 * Those in particular don't help. I'm trying to conjure up a javascript fix, but can't really think of how either...  The only other alternative is to implement the correct CSS behavior via javascript on all browsers without using CSS, but that's kinda annoying cuz it's IE's fault and not the other browser's.  I don't want to use javascript to do something I'm supposed to be able to do with CSS. -PanSola 20:52, 13 April 2007 (CDT)