Difference between revisions of "User talk:Sven"

From Organic Design wiki
(Selenium tests)
(Selenium tests: answers)
Line 153: Line 153:
 
::I have got the idea now, just going through your examples now where you made accounts based on a GUID first logging out. I will add some content tonight. The issues I see are;
 
::I have got the idea now, just going through your examples now where you made accounts based on a GUID first logging out. I will add some content tonight. The issues I see are;
 
*Do we need to lock test scripts down so that passwords cannot be grabbed from articles?
 
*Do we need to lock test scripts down so that passwords cannot be grabbed from articles?
 +
:Some test articles may need to be locked down (eg ones which test sysop security), but I don't think there should be any issue with doing that.
 
*Using a Namespace, either Test:[Extension].php, or Selenium:Extension.php
 
*Using a Namespace, either Test:[Extension].php, or Selenium:Extension.php
 +
:As described in [[Extension talk:Selenium.php]], "spec" may be more appropriate since they should contain a english description of each item of functionality along with the script to test that feature.
 
*Correctly validating an edit.  
 
*Correctly validating an edit.  
 +
:What's the issue here? just create a test for that
 
--[[User:Sven|Sven]] 14:37, 5 December 2007 (NZDT)
 
--[[User:Sven|Sven]] 14:37, 5 December 2007 (NZDT)

Revision as of 01:48, 5 December 2007

You were right, our exchange is the first to get unbundled ;-) --Nad 19:50, 10 August 2007 (NZST)

Ah so that the reason for the outage. --Sven 09:16, 14 August 2007 (NZST)

See Sizes of things for notes regarding stuff at different scales and an animation too... --Nad 19:14, 30 July 2007 (NZST)

Cool, still sick but on the mend now, see you tomorrow. --Sven 15:09, 2 August 2007 (NZST)
Be prepared for a bashing though, cos Dad caught your sickness, and he's real pissed about it! --Nad 15:20, 2 August 2007 (NZST)
Your joking. I definately had a lung infection, been festering for a while... --Sven 15:23, 2 August 2007 (NZST)

There's no point trying to fix QuickCat yet - a bug has cropped up in SimpleForms which is stopping the Ajax from working properly - see the Ajax links - example number 3 doesn't work. Will fix soon. --Nad 20:14, 13 July 2007 (NZST)


I deleted All news it was causing an infinite loop in the main page. I take it was so that you could see old news items if you needed to? I've added next/prev links to homepage news instead --Nad 12:16, 7 July 2007 (NZST)

How was it causing an infinite loop, was the main page trying to also reference it or something? --Sven 13:38, 7 July 2007 (NZST)
All news was in the list of items rendered by Main Page and visa versa - it could be fixed by excluding those articles from the results with nottitlematch or similar, but I thought next/prev links would be cool anyway - also next/prev is a cool generic thing which could be used elsewhere. --Nad 14:26, 7 July 2007 (NZST)

sven- saw your link to findmyforte, and thought you might find this useful: http://www.assessment.com/ i paid for their career assessment, and it was pretty much right-on.

also, if you check out http://www.wikiexpert.com/Category:Domain_names, you will see the wiki stuff i have done under the auspice of "wikiexpert" i just landed another wikia install (which im going to ask aran to do!). let me know if you want to see some numbers for hosting, installation, etc. that I am charging. cheers! --Phalseid 03:26, 26 June 2007 (NZST)

Sounds interesting, got a fair bit out of findmyforte, I am certainly interested in what others have to say. Any details on cost etc would be good to stick somewhere in Sven/Biz. --Sven 08:29, 26 June 2007 (NZST)


I have no idea why Template:R isn't categorising, bloody weird!? --Nad 10:30, 12 June 2007 (NZST)

Yeah it beats me, Ive tested {{php}} is working, must be something in extension code

--Sven 10:34, 12 June 2007 (NZST)

Comparing with C is best since it's also a one-letter one which works - it's weird, if you put "uses=template:c" into search it works, but "uses=template:r" gives no results even though it's being used...? --Nad 10:56, 12 June 2007 (NZST)
Is case an issue? --Sven 11:03, 12 June 2007 (NZST)
No >< --Sven 11:04, 12 June 2007 (NZST)

You were asking about the MW DB layout - here's a big diagram of the tables. All the conceptual info regarding history and links etc is in the bottom right yellow section. --Nad 20:32, 2 June 2007 (NZST)


The BFO (Basic Formal Ontology) has an excellent manual which is worth reading (at least the first few chapters anyway, it's about 130pages in total), it's an excellent overview of the current state of semantic technology etc. It comes from your neck of the woods too - bioinfomatics - you heading over tomorrow? --Nad 18:25, 31 May 2007 (NZST)

Yeah round @ 10am --Sven 19:10, 31 May 2007 (NZST)

I don't know why the R syntax isn't highlighting, I've added that old article etc, but the resulting pre tag should render as <pre class="r"> but its just a naked pre, not sure what I've missed out... maybe there's something funny with that old file? --Nad 19:48, 29 May 2007 (NZST)

I see what you mean, Which old article did you add? --Sven 19:59, 29 May 2007 (NZST)
You said looks like the geshi/R.php had been overwritten and I said Over-written shmover-written, i.e. GeSHi/R.php --Nad 21:02, 29 May 2007 (NZST)

hey sven, peder here. feel free to use wikiexpert as your consulting platform/ one of you marketing tools...id like to work together on this. --Phalseid 06:19, 27 May 2007 (NZST)

ok thanks, I still have to do alot of work to piece together a suitable business model. --Sven 07:34, 27 May 2007 (NZST)

This is how you would parse wikitext completely indpendently of the state of the main parser or page etc <php>$p = new Parser; $o = new ParserOptions; $t = new Title; $html = $p->parse('foo',$t,$o,true,true)->getText(); </php> Usually there'll already be a parser available, like one supplied as a paremeter in a hook function, or by using the page globals, eg: <php>

  1. Notice the last parameter is false now that means do not reset the parsers internal state

$html = $wgParser->parse('foo',$wgParser->mTitle,$wgParser->mOptions,true,false)->getText(); </php> Another useful one to know is how to just expand the templates in a wikitext expression: <php> $expandedWikiText = $wgParser->replaceVariables('{{foo|bar|baz}}'); </php>


Have you checked out w:MediaWiki, it's got a good overview of the versions and their functionality etc --Nad 12:15, 16 May 2007 (NZST)

Cool, I will check it out. Been laid low with a stomach upset all week :( --Sven 13:35, 16 May 2007 (NZST)

Template:Extension allows you to make a generic OO tag and parser-function extension class using transclusion as in the Extension:Example. The new GeSHi transclusion method's not working properly and generating &lt;'s and &gt;'s, but it still serves as an example until that's fixed. --Nad 21:31, 5 May 2007 (NZST) The template is used as follows:

{{Extension
|name    = Example
|author  = [[User:Foo]]
|url     = http://www.foo.bar
|version = 0.0.0/2007-05-05
|type    = parserhook
|magic   = foo
|tag     = bar
}}
You mean like this Sven/Sandbox, I also added the tags in the article --Sven 21:40, 5 May 2007 (NZST)
Fixed with another geshi hack - see GeSHi for details --Nad 21:58, 5 May 2007 (NZST)
Have a look at the sandbox source, that foo and bar are done by the extension which was built from Extension:Example. So you can fill in the Template:Extension the same way and it yields a working extension with a parser function and tag-hook that you can then adjust --Nad 22:25, 5 May 2007 (NZST)

The main calendar is already an example using the query arg, see news on Main Page - you need to ctrl+shift+refresh to reload CSS for red-links in calendar to work again because they're done from their own css rules now. --Nad 14:34, 2 May 2007 (NZST)


You should resize them to a practical size (or the size you'll use them at) before uploading, that's half a meg of images just for icons --Nad 09:34, 1 May 2007 (NZST)

Good point, I need to find some smaller ones --Sven 09:35, 1 May 2007 (NZST)

Yo, here's an example of action=text that I was talking about to allow dynamic queries by supplying wikitext directly in URL,

I've had to use %7b on the first brace so the parser leaves it alone, notice that there's no title specified, only the actual text we want parsed. It only works for action=render currently, but later will be able to be used with all actions. --Nad 20:02, 27 April 2007 (NZST)

Nice --Sven 20:05, 27 April 2007 (NZST)

I changed the MW:Extension:CategoryHook to add a hook called CategoryHook which you can add you rules to instead of adding them directly to the code which was messy. It makes a good simple example of how to add a new hook, rather than just adding new functions to existing hooks. --Nad 16:29, 27 April 2007 (NZST)

k, will have a look, the author dpl stuff is not working as expected, the template is getting parsed aswell see template:author --Sven 16:30, 27 April 2007 (NZST)

You have to get the case right with security directives, groups are always all lowercase, and user must have the first letter uppercase. --Nad 14:39, 26 April 2007 (NZST)


Remember that the "message" css class applies to all templates using the message template such as warning, stub, legacy etc, not just to the site message. --Nad 20:29, 23 April 2007 (NZST)

Yes, the template:message you created was similar to what I wanted to do, having a transcluded format used with templates. --Sven 20:32, 23 April 2007 (NZST)

The fella I was gonna see has cancelled, so whatever time you wanna head over's fine, just lemme know... --Nad 09:27, 23 Mar 2007 (NZST) --- Did you know there's an IETF RFC for foo? RFC3092 --Nad 22:43, 27 Jan 2007 (NZST)

omg --Sven 20:22, 29 Jan 2007 (NZST)

You might be interested in Category:MediaWiki which has all our stuff re hacking normal MediaWiki's (mainly from LocalSettings). Especially check out parsing wikitext and MediaWiki variables. --Nad 21:17, 19 Jan 2007 (NZST)


Hey Sven, I remember our discussion about the science annotation and paper cross referencing. I'm at the MacWorld in SF and I looked at an interesting software product that does some of the same things.

--Rob 21:47, 9 Jan 2007 (NZST)

Thanks will check it out --Sven 21:04, 15 Jan 2007 (NZST)

Is your sister still interested in extra design work? azzuro want their external site reskinned - just need HTML/CSS templates for some standard layouts, probably $1000-$2000 arena --Nad 11:02, 6 Dec 2006 (NZDT)

I am sure she would be interested, I will let her know --Sven 13:52, 7 Dec 2006 (NZDT)

Hi there, I have recently met a guy that runs a bioinformatics company - Biomatters. They make a software that you may be interested in called "Geneious" http://www.geneious.com/. I recall you were developing ideas around setting up bioinformatics software for corporate clients and these guys are looking for ways to market their software. It sounds like an excellent product and it seems like they are looking for further contacts in the bioinformatics industry. Maybe you have some ideas for them regarding applications or distribution channels for their software? There might be some consultancy work in there or whatever. Anyway, I can put you in touch with their CEO if you're interested.--Milan 15:16, 25 Oct 2006 (NZDT)

I have already checked out the software and spoken to Alexi. --Sven 10:06, 26 Oct 2006 (NZDT)
Sweet! Are you guys going to be working together? I bump into Daniel Batten a fair bit, but I haven't met Alexi yet.--Milan 12:07, 26 Oct 2006 (NZDT)
I havent approached alexi to be involved, my area is in microarrays which I think is slightly different from what their software is about, although its all part of bioinformatics --Sven 14:10, 26 Oct 2006 (NZDT)

SWISS-2DPAGE

SWISS-2DPAGE contains data on proteins identified on various 2-D PAGE and SDS-PAGE reference maps. You can locate these proteins on the 2-D PAGE maps or display the region of a 2-D PAGE map where one might expect to find a protein from UniProtKB/Swiss-Prot

Links

List & Node spaces

I found those code snippets and added them to the discussion of nodal concepts, Category talk:Nodal Concepts --Nad 14:59, 5 October 2007 (NZST)

Great thanks. --Sven 16:02, 5 October 2007 (NZST)

MediaWiki script history

Check out ajax.js history, it shows the changes to the script throughout the major mediawiki versions (the missing ones are just when no change happened) --Nad 18:10, 5 October 2007 (NZST)

Portals help

You may be able to give ol' Duesentrieb some help, he's wants a page explaining how to make nice portals at MW:Project:Support desk#How to make portals --Nad 09:44, 29 October 2007 (NZDT)

Selenium tests

I've made a general plan for adding Selenium into the MediaWiki environment, see Extension talk:Selenium.php

Accounts

I also added some info to Accounts which we need to address asap - I think a good way to get this happening would be to set up a bank account for the biz and get all the current OD income and expenses moving through it. --Nad 16:22, 3 December 2007 (NZDT)

Ok will have a look --Sven 16:30, 3 December 2007 (NZDT)

You've got mail

I sent a cupla emails... --Nad 11:17, 4 December 2007 (NZDT)

Selenium tests

Once you've got the general idea of Selenium installation and usage and refined the Selenium article to reflect that, then the next stage is to move on to the Extension talk:Selenium.php article which is about how we'd go about using Selenium in our wiki environment. I've added quite a few notes and ideas and also added the new Template:Collab to it which is an attempt to get some of these collaborative efforts moving forward. --Nad 14:14, 5 December 2007 (NZDT)

I have got the idea now, just going through your examples now where you made accounts based on a GUID first logging out. I will add some content tonight. The issues I see are;
  • Do we need to lock test scripts down so that passwords cannot be grabbed from articles?
Some test articles may need to be locked down (eg ones which test sysop security), but I don't think there should be any issue with doing that.
  • Using a Namespace, either Test:[Extension].php, or Selenium:Extension.php
As described in Extension talk:Selenium.php, "spec" may be more appropriate since they should contain a english description of each item of functionality along with the script to test that feature.
  • Correctly validating an edit.
What's the issue here? just create a test for that

--Sven 14:37, 5 December 2007 (NZDT)