User:84user/Sandbox

Inspired by this proposal on the Commons:Village pump, I created File:Knot table-blank unknot.svg to see how images can be overlayed on one another.

Using {{Lageplan}}

{{Lageplan}} with half-sized base image:

Using {{Superimpose}}

{{Superimpose}} with half-sized base image:

Note that Lageplan uses percentage positioning which allows the caller to alter the underlying base image width without recalculating the position of the overlaying marker image. However, both templates require the caller to adjust the marker width.

• all
• WPAstronomy
• WPSpace

Timeline:

• 16:27, 18 December 2008 (UTC) (or five minutes before) I sent a repeat email to info (at) isro.gov.in
• Sunday, December 7, 2008 2:43 PM (timestamp on email, so someone's clock is wrong) I received "Undelivered Mail Returned to Sender" in response to email I just sent 14:44, 7 December 2008 (UTC)
• 14:44, 7 December 2008 (UTC) (approximate time) I sent the first email request to info (at) isro.gov.in (the box at the bottom)

Here is the request I emailed on December 7, around 14:45 UTC. 84user (talk) 14:44, 7 December 2008 (UTC)

Examples of #time use:

• {{#time: F j, Y|2008-01-15}} = January 15, 2008
• {{#time: j F Y|2008-01-15}} = 15 January 2008
• {{#time: Y-M-j|2008-01-15}} = 2008-Jan-15
• {{#time: Y-m-j|2008-01-15}} = 2008-01-15

• {{#time: a-b-c-d-e|2008-01-15 19:58:59}} = pm-b-2008-01-15T19:58:59+00:00-15-UTC
• {{#time: f-g-h-i-j 19:58:59|2008-01-15}} = f-7-07-58-15
• {{#time: k-l-m-n-o|2008-01-15}} = k-Tuesday-01-1-2008
• {{#time: p-q-r-s-t|2008-01-15 19:58:59}} = p-q-Tue, 15 Jan 2008 19:58:59 +0000-59-31
• {{#time: p-q-r-s-t|2008-11-24 03:58:59}} = p-q-Mon, 24 Nov 2008 03:58:59 +0000-59-30
• {{#time: u-v-w-x-y|2008-01-15 19:58:59}} = u-v-2--08
• {{#time: z|2008-01-15}} = 14
• {{#time: z|2008-11-24 03:58:59}} = 328
• {{#time: A-B-C-D-E|2008-01-15 19:58:59}} = PM-B-C-Tue-E
• {{#time: F-G-H-I-J|2008-01-15 19:58:59}} = January-19-19-0-J
• {{#time: K-L-M-N-O|2008-01-15}} = K-1-Jan-2-+0000
• {{#time: P-Q-R-S-T|2008-01-15}} = +00:00-Q-R-S-UTC
• {{#time: U-V-W-X-Y|1800-01-01 02:58:59}} = -5364651661-V-01-X-1800
• {{#time: U-V-W-X-Y|1970-01-01 00:00:59}} = 59-V-01-X-1970
• {{#time: U-V-W-X-Y|2008-01-15 19:58:59}} = 1200427139-V-03-X-2008
• {{#time: U-V-W-X-Y|2008-01-15 19:59:02}} = 1200427142-V-03-X-2008
• {{#time: U-V-W-X-Y|2008-11-24 04:58:59}} = 1227502739-V-48-X-2008
• {{#time: U-V-W-X-Y|9999-12-31 23:59:59}} = 253402300799-V-52-X-9999
• {{#time: Z|2008-01-15}} = 0

84user (talk) 12:10, 1 December 2008 (UTC)

Above: Featured articles per section; below: sorted by number in each section.

The above timeline should be visible but was blank, but now works after the bugfix. 84user (talk) 19:03, 16 December 2008 (UTC)

The below timeline generates html with an img tag with src attribute URL that does not exist on wikimedia: http://upload.wikimedia.org/wikipedia/en/timeline/1455f41de5ae0fa00af5fe18dc65879b.png

Above URL shows the chart missing the text, when the bug was active. After the bugfix, the img tag's source URL is http://upload.wikimedia.org/wikipedia/en/timeline/067dc5326eb942425c40e1cc5bc8cd27.png, which now included the text. 84user (talk) 19:03, 16 December 2008 (UTC)

Does the timeline mechanism no longer work on wikipedia?It works now. 84user (talk) 19:03, 16 December 2008 (UTC)

The above should be a bar chart, but is blank containing a non-found URL. and was broken but now works. 84user (talk) 19:03, 16 December 2008 (UTC)

Below is a third try at a timeline. Which was blank, but now shows "London" and "Paris" in large font (truncated).84user (talk) 19:03, 16 December 2008 (UTC)

Another :

Sometimes the timelines partially render, other times they do not render at all (completely blank images or no wikimedia image file at all). Now they work. 84user (talk) 19:03, 16 December 2008 (UTC)

In 2008 Bootle explained he was a fundamental economist and that he had detected and announced the overvaluation resulting in the Financial crisis of 2007-2008 ten years earlier. He now believed he was too early in his announcement and felt he should have waited a few years, when the signs would be clearer to others. He said now in mid-crisis that he was more worried about the coming reaction by people to what they are seeing, and that there is "a lot of anger but not a lot of understanding".^[4]

Some wikipedia article web pages take a long time to load. It seems the reason is some combination of: the total kilobytes in the page's html, css, javascript and images; the total number of HTTP requests made by the browser when loading.

The actual number of bytes in the page's wikicode is not a good predictor for the total kilobytes experienced by a user that clicks on a particular wikipedia link for the first time. (See Wikipedia talk:Article size for long discussion)

To find out whether this lengthy page loading time is only caused by a minority of articles, I'll start a table to record total kilobytes and time takes for Wikipedia:Featured articles (FAs).

I chose the first FA from each of the first few sections in that list and then tried Phoenix (spacecraft) followed by two DNA articles.

Kilobytes is the total kilobytes downloaded (HTML, CSS, JS, XHR, Images). Load time is in seconds with a fast ADSL. Load times were generally not affected by the downloading computer, ie. neither CPU was ever taxed beyond 25% and the network use never exceeded 40% (except for the DNA articles). All non-DNA articles peaked around 15%. The Computer was run with over 1000 megabytes of spare memory (which did not get used, but I set this to avoid any chance of page swapping). I noticed a variation of around plus or minus 20 % in load time when I repeated the fresh reload.

I was surprised to see the Citizendium's DNA article fully loaded in 4.33 seconds, despite the massive 3 megabyte animated image. Of course that giant image could make both DNA articles useless to a dial-up user.

How I made these measurements: I used Firebug 1.2.0b7 on Firefox 3.0.1 on Windows Vista on an Athlon 64X2(W) 5000+ 2.6GHz with a 2.4Mbps ADSL, starting at 21:50 UTC on August 5, 2008. I loaded each page using control-shift R to ensure each element is loaded afresh in order to duplicate the experience a reader would get the first time a wikipedia link is clicked. I stayed logged in which probably made the load times somehwat longer than for an IP user. -84user (talk) 23:08, 5 August 2008 (UTC)

What prompted me to do this test was this version of the Phoenix (spacecraft) article was taking a long time to load. As it was over 2300 kilobytes and took over 17 seconds to load I reduced the images so that it is now "only" 732 Kilobytes and takes "only" 9 seconds or so.

After looking around, I now think my expressed view (about social fixes) is a bit pointless.

There are higher level problems, a big one is that editors (myself included) easily forget that most readers want and can manage simple, intuitive mechanisms.

Back to categorisation, yes there is probably too much pointless diffusion occuring. But indexing all Men, or all Women is best done by an automated mechanism with access to a database. It seems the German wiki makes more use of Wikipedia:Persondata than the English.

In conclusion here are some main categories found in biographical articles (unusual in bold and note the interwiki links are not logical).

"automation" means I think automation should one day place articles in either a fully-indexed category or in some other kind of indexable mechanism, maybe using Persondata? Also, Category:Man would become Category:Concepts Man

"use CatLargeTOC" means from w:en:Template:LargeCategoryTOC create w:en:Template:LargeCategoryTOC2D to behave like w:de:Vorlage:TOC Große Kategorie - this should not need any social persuasion (change ausklappen to "show", of course)

"leave for concept" means leave the category for its currently defined concept

This is a quick presentation on how an ordinary wikipedia reader might look for something. This compares searching for something on English wikipedia with German wikipedia.

First a little rant to set the scene.

Pretend you only use wikipedia as a reader (the majority), you only click on things, and maybe type a little into boxes. Like the great majority of internet users you are NOT technically proficient, you do NOT want to edit things, you do NOT want to read long lists of instructions. You could be an intelligent layman but with little spare time. You do not like scrolling (yes, otherwise intelligent people are unfamiliar with scrolling).

You like google because it is FAST and has just two buttons. You do not need to scroll to use the search.

You understand how a printed dictionary works, and index books one sees in libraries (those buildings which house books for people to read, maybe to borrow).

So, you see some television programme on a Polish film animator, but maybe he's Russian, you have forgotten. You remember his name started with "Zbi".

Google does not help here. So you think "wikipedia?". You type "Zbi" and search - nothing.

You try Poland and under Culture you see four paragraphs. You find Zbigniew Rybczyński and that's who you were looking for. Good, yes? Yes, but only because ZBig is quite famous, what if it was another Zbigniew?

Ok, you could find "List of Poles" (via "Polish"). There some 18 people named Zbigniew. Are they in one place? No. Is there an index? No. Right at the bottom (people do not like scrolling, remember) you see "Polish people | Lists of Polish people". I try "Lists", impressive, but no list of all Polish people. Is my Zbigniew a painter or a poet, or is he both?

As I am a busy, inexperienced internet user, I give up at this point. Wikipedia is nice if you know what you are looking for, but for searching stuff when you only have some details, it does not help.

But wait, I am bilingual German, so I try the German wikipedia. "Polen" gives me the country but no list of people, but it does have "Kultur in Polen" with a list of people. Also not sorted, (and the one I'm looking for is not there anyway). It does have a "Begriffsklärung" (disambiguation) header though, and that leads to de:Kategorie:Pole ... which has an index by last name.

Here it seems that even the German wikipedia does not provide full people indexes for every nationality. Also it does not sort by FIRST name either.

What the German wikipedia DOES HAVE is:

• Each biographical article has all the categories and does not remove the parent ones
  • Example: "Kategorien: Filmregisseur | Kameramann | Polnischer Künstler | Mann | Geboren 1949 | Pole"

This means that "Polnischer Künstler " has ALL the Polish artists, indexed. The same for "Geboren 1949" - every single German wikipedia article for someone (believed to be) born in 1949 is listed. It also has subcategories, but the article is in BOTH subcategory and the parent one.

This brief exercise has shown me things are not so simple, but also that wikipedia could benefit the user by doing testing the site with real users (not editors) to see how they use it and what works and what is annoying.

A subject for later.

There's a long list of famous Polish people. The one I am looking for is actually there, but I simply do not find it. Why? It is not sorted. There is no A to Z index. It is a long list. My eyes glaze over scrolling the long list. After the entertainment list I keep scrolling. Gosh, there are lots of people listed here. Keep scrolling. Maybe in Others? No.

Now I wonder if this guy is Russian? All I know is it starts "Zbi".

I am bilingual German, so I try the German wikipedia. "Polen" gives me the country but no list of people, but it does have "Kultur in Polen" with a list of people. Also not sorted, (and the one I'm looking for is not there anyway).

Drawbacks:

• results shown in database order, not alphabetical
• results shown in search format and not in category format
• two dimensional table limits searches to combining two categories

Note this user-intuitive browsing only works if editors refrain from diffusing categories!

All the technology is already in wikimedia.