From The Urban Dead Wiki

Jump to: navigation, search


Danger to Users?

Q:Can you imagine, being a survivor, and your every move being logged and available to everyone? (EX- PKers) Every. Move. You make.

A: Good point, and one that I took to heart from the very conception of the idea. Trust me, I am more privacy / info security conscious than the typical person, by far. If you doubt this, ask any of the folks in TNT07...

Which is why I/Witness is very different in its basic conception and function and purpose from DEMON and similar tools.

My impression is that DEMON effectively records a "file shot" of every UD game action you take, puts this info in a collective database, and makes the database info available to all other DEMON users. I may be wrong, but that's close enough for the "wiki trial", and good enough for me to know its not something I want to be a part of. And indeed, making such a thing publicly available would be an awful idea.

I/Witness would only record a "file shot" when you requested it to do so, by clicking a "take file shot" button. The info in those "fileshots" would only be available to you (or, more precisely, your I/Witness account) and to individuals (accounts) you gave permission to view the fileshot. You could also set individual file shots as "public access" or "friends only", and so on. In effect, I/Witness would be a secure, verifiable blog of "screenshots" (actually file shots) for UD characters, and not a(nother) DEMON clone.

There would be the potential for the adminstrators of the I/Witness website to extract DEMON style data from the collected file shots, or just browse file shots at will, but again- I would not want this done without the permission of the people (accounts) that provided those file shots, for the exact reasons you give above. I'm actually going so far as to look into encryption schemes that would make it impossible for even the site administrators to access file shot and other account data without permission from the originating account. If such extracts could be done in such a way as to preserve the anominity and safety of the information source- such as by providing map updates with a 48 hour lag, or as bulk statistical overviews of entire suburbs- then they might be a standard feature, though I'd still prefer to have an "opt out" program (though opting out may limit your access to other I/Witness features).

Mobile Phone Masts

Hey there,

I just discovered this page on the Wiki, and I'm quite interested in your tool (Or something like it). Would it be possible to build a tool that would automatically update a website (The Mobile Phone Mast page, for example) whenever someone used their mobile phones? This way, it would be much easier for us from MalTel to gather information on which phone masts are up and which are down!

Let me know,


Depends on who that "someone" is. If that someone is willing to open a secondary web page while playing or download a firefox extension, it could be done. If not, no. Either way, it would be a pretty major undertaking.
One possible use of the basic I/Witness structure would be to make saved views searchable by the location (as well as by the poster, by time, by people in the view, etc). Such a feature would make updating pages like yours easy, because you could just pull all then data from views taken at the appropriate location. A concern here is that obviously you'd want to protect privacy a bit by not letting everybody view any and all info the search turned up; each user would have to give permission to who could view their entries, or some such.
Anyhow, as the creator of the NT Status Map, such location status tracking is certainly one of my greater concerns. MalTel has 100 locations to deal with; the NT Status Map covers 211. So trust me, Any way I can figure to reduce the work, I will do!
--Seb_Wiers VeM 20:44, 2 May 2007 (BST)

Development History Log

17 May 2007

  • Decided to use PHP / MySQL as main tools.
  • Purchased "PHP and MySQL Bible" - 1000+ pages, $50. Yikes. But amazingly informative, without being over the head of somebody like me who can sometimes handle (and sometimes bungle or entirely miss) the more complex stuff on the wiki.
  • Contacted the Coalition for Fair Tactics on thier talk page to see if we can head off any accidental implementation problems. Waiting for response.
  • Started this log.
--Seb_Wiers VeM 02:58, 17 May 2007 (BST)

26 May 2007

Asked Kevan about the project, no issues on that front. There's still some features I;d like to be able to implement that I can't figure out a basis for, but that is more a "wishlist" concern than something that will stop initial progress. At this point, it's all about pounding the keys and checking the code.

--Seb_Wiers VeM 14:58, 26 May 2007 (BST)

28 May 2007

Got apache with PHP module and mysql installed on my cranky old p3 machine, so I can fire it up and test my php when needed, work on the database offline, and so on. So work now begins in ernest- I'm actually coding at this point.
I hope to have SOMETHING up on the web within a month. It will likely be very crude. My design philosophy is to be as lazy as possible and build in small (but publicly usable) chunks. I can't do bells and whistles with my lousy skills, but I figure if I keep it simple and focus one one small step at a time, it will get done much faster.
The initial "pre-alpha" version I'm shooting for at this point will have the following (lack of) features:

  • One click uploading of game view HTMLs as files
  • Message that gives you a link to your uploaded file when uploaded
  • No log in / session tracking for users
  • No (user) permission control over who can view files
  • Google based searching of uploaded files

--Seb_Wiers VeM 14:50, 28 May 2007 (BST)

3 June 2007

Much cussing and gnashing of teeth.
After some expereimenting, I got the basic elements of some code in place that would demonstrate a "proof of concept".
And it works. The code works, that is. It functions just great on my own machine, looking at local files, and when running on my nice little pocket apache system looking at stuff on the localhost.
The concept, however, fails when I take it "out on the road" because apparently what I'm doing is called "cross domain scripting"- IE, using javascript served up by one domain (I/witness) to look at a document served up by another domain (urbandead). And apparently that's not something most browsers allow, because its a security risk - it allws some pretty nasty phishing attacks, for example. OK, I can see that - even I was smart enugh to realise that I/Witness had some potentially nasty privacy issues to deal with. But apparently most browsers (and this includes Firefox) do not even give the user the OPTION to allow cross-domain scripting. Pheh, thanks for protecting us form ourselves, guys! You'd think blocking it as a default setting and putting in a big security warning would be enough, but NOOOOoooo...
So for now, the project is dead in the water, as the entire basis for its function seems to be anathema to the major browser makers. And if I did ever get it floating, I'd be discovering what apparently is considered a major security hole, and so could expect them to close whatever loophole I found.
More sub-vocal grumbling.

4 June 2007

Ok, apparently both user scripts (implemented via greasemoneky) and bookmarklets (implemented as plain code) will do what I need without much change to my concepts or code, or forcing me to learn any new esoteric techno-lingos. Both are just special case uses of javascript. So I guess I'm gonna be able to roll on pretty easily. I'm actually surprised how well bookmarklets will serve this purpose; I'd think if they do my job so easily, they would be as much a security risk as just allowing cross domain scripting would be. But I'm not gonna complain- I'm just happy they will work. --Seb_Wiers Imagine 05:10, 5 June 2007 (BST)

9 June 2007

Well, I seem to have gotten most of the "front end" javascript coding needed to submit the screenshots completed. Which means I can soon start on actual PHP and database coding, using actual "screenshots" as working data.
Ironicly, the entire front end consists of one Bookmarklet- currently less than 400 characters long, ideally less than 255. I may be able to trim some fat, but modern browsers can handle it. Still- that's a damn lot of work I did for 400 characters of code. Then again, that short bit of code does a lot of work. --Seb_Wiers Imagine 21:36, 9 June 2007 (BST)

13 June 2007

I have been watching your project with great anticipation and it seems to be steaming along nicely. I am concerned that you may be about to drive off the road. Have you read up on the concept of Ajax? This is accepted by all prominent browsers and allows you to be browsing any website and contact another server in the background. It is quick and easy as its already in place in explorer and firefox. A simple ajax solution is also available to Opera with Java enabled and a few security tweaks. Check it out. --Max Grivas JG / M.F.T. 20:53, 13 June 2007 (BST)

I'm already working on the server-side code, and I don't see any problems developing- everything I have planned, PHP and MySQL can handle. My current implementation should allow everything I have planned to be developed modularly over time, with minimal coding complexity / learning. Which given that I had zero programing knowledge when I started this, is much more important than using the most up to date technology.
Ajax is a client side technology, and I really don't need the client to do anything except upload some document data. I've already written a bookmarklet that handles that task, and if I ever want to get fancier, I can develop that script into a firefox extension. I honestly can't see what benefit ajax could give, and have no idea how to implement it. Maybe once I get the pre-alpha version online, you can give me some more specific tips, or even help with development?
Truth told, when I started this project, I had no programing knowledge beyond a little javascript and basic concepts like flow charts... 21:38, 13 June 2007 (BST)
Sounds great! I am looking forward to seeing a working model. Keep plugging. --Max Grivas JG / M.F.T. 22:09, 13 June 2007 (BST)
Thanks. Once I have the data-input and database construction aspects in place (that would bet me up to alpha) I eventually want to build some fancy data display tools. That's where something like ajax might come in handy- your malton explorer is pretty nice, for example. The tricky part is going to be making sure people only have access to data if the person who originally provided that data wants them to- which I why I'm putting that stuff off until the beta version, and a slick interface off for the "final" version. 22:57, 13 June 2007 (BST)

15 June 2007

PHP / server side coding is proceeding at a surprisingly quick pace, and I should have a crude but functioning prototype ready to put online by the end of the month. With that in mind, I went ahead and registered the domain name, and am using for my hosting service. The plan I got is actually one of the smaller ones I came across (in terms of storage and bandwidth) but its absurdly affordable and has almost no restrictions on numbers of Databases, sub domains, and other configurable features. And the host has great reviews, unlike some I was looking at the provide more space but also have more restrictive terms. 17:02, 15 June 2007 (BST)

22 Jun 2007

Although it is not function in terms of recrding files or creating a database, the basic interface for taking Iwitness "screenshots" is now complete. You can check it out at and get an idea of the look, feal, and basic capabilities.

Below is an example of this interface in use. The highlighted red boodmark is a bookmarklet created for Iwitness. When clicked while viewing the UD game window (hidden in this screenshot, but the tab is visible and highlighted red), the bookmarklet submitted data to http:/, which in return created the window you see.


24 Jun 2007 - Iwitness is now in live pre-alpha opperation. It captures data and makes files. It looks pretty, and works fairly well, although there are still a few minor bugs, and features I'd like to add / get working before I'd actually call it an "alpha launch". ... 23:11, 24 June 2007 (BST)

Please leave comments in the discussion area above, and create a new heading if appropriate.

Personal tools