London Bike Theft Database

+1 on the RSS feed. Good idea.
+1 on the trackday checks. Enduro events would also be a good place to check - but in both cases, more than likely you wont find a stolen VIN. Just the parts…

Thanks Driesie,

We have an excess of resources, so hosting isn’t an issue, we have a relatively unused dedicated server with unlimited bandwidth we can use for this.

There’s quite a lot of scope with this, so I’m thinking it would be an open initiative, so developers could join the team (upon application) and contribute to it. LB would host it as a good-will gesture. Scope wise, it could expand to cover all vehicles around the world, with the localisation features people have suggested here to make it relevant to specific communities.

We would use an LB SVN server for source-control and then the following technologies would be used: ASP.NET 3.5, c#3, Linq-to-SQL, SQL Server 2005. Hosting and deployment administration would be managed by us.

I could probably get a prototype running quite quickly once we have a good definition of what people want, and then we could open it up to developers. This is quite good timing as I’m due to do an LB project shortly and once that’s done, we should have a good enough idea of what we want for this.

We could get this running in association with other popular bike sites.

Is it worth posting on to MCN, Visordown etc asking members on there for any suggestions?

Could we have a non registration section opened up on here for suggestions from people, that want to contibute but not necessarily become an LB member?

Sounds good, Jay.
Not a great Linq to SQL fan, more of an nHibernate person, but it won’t bother me too much :wink:
I think it would be a good idea to run it as an open project. Not so much open source code, but let the data be open as well. It’d be great if all the data would be exposed as well structured web services for example. That way, if people want to write apps to consume the data, they just can (for example flash banners, custom feeds, mobile apps …).

Afro, as for teaming up with other bike sites, I think it’s a good idea, but I think if there is a “working prototype”, then it’ll be easier to “sell” I think.

You’re right…The collaboration could come later.

Two suggestions from a tech perspective.

Make the search work well. Its got to be easy to find stuff.
Google-style search. One textbox (search on VIN, rego, make , whatever you can think of)
Dont use the database for that - have a look at (if .net is what youre going to use).

Leave the RSS/Atom stuff till last. Get a simple web ui up and running and get feedback.

Er, and thats it from the peanut gallery… :smiley:

I’d be hesitent to open this up to a wider audience, as that’ll slow things down and bloat out the requirements. We can open it up once it’s established.

Launch soon, iterate quickly, that’s the key.

And yes Drisie, it wouldn’t be open-source, there’s no benefit there as this is a single-instance application. Not to mention it might raise security concerns.

We need someone to research into the data-protection aspect and how much we can make it open by. There are going to be legal considerations. There’s also going to be common-sense issues, we can’t make this a tool for thieves.

London map, plotting where all the bikes where stolen, might see hot spots to avoid or see if there’s any pattern accruing!

Yes, nice idea Woo Woo.

Why would you not use a database server for searching? That’s what they’re specialists at. SQL Server 2005 FTI is exceptionally good.

Lucene looks good, but from what I can see, it’s a text-file indexing and query system, not a database query abstraction system.

Dont want to hijack this thread - happy to discuss further offline - but look at google. They dont use a db.
No-one who does search well uses a db. They use a search engine.
There are many reasons why a search engine is much better than any of the database offerings (ranking, boosting, speed, etc, etc). (and databases are not specialists at searching…)
My advice to take or leave - Ive built various systems the past few years leaning heavily on Lucene - and had this discussion numerous times & I dont mind having it again over a beer. :slight_smile:

I think the discussion above illustrates my point. What I meant with the system being “open” (I agree, there’s not much point in releasing an open source version of something like this), is that everybody who has an opinion can just go ahead and hack their own app together. I know doing something like giving access to the data using web services might mean they scope creap etc etc, but once you have that, you open a lot of possibilities. If somebody wants to do a google type search engine, go ahead, consume the data and index it. If somebody wants to do a google maps mashup, again, just consume the data and go ahead.
From a personal point of view, I’d be really interested in building a OLAP cube with this data and try to see what you information you can get through trend analysis. I think that would be very interesting, but I appreciated that doing something like this would be distracting from the main immediate goal. However, if I would have access to the data, I’d give it a go, and who knows you could get some really interesting information out of it.

I think the DP issue is a very valid point. I think you probably could get around that by making the web services “subscription only”, meaning somebody would need to create an account and confirm they agree to the usage conditions.

Interesting points. Well, there’s time for this to come together before any architectural decisions need to be made. I’m glad though we’ve got folks willing to help who know what they’re talking about :slight_smile:

May be worth talking to the Met I’m sure they would be glad to give advice and probably use some of the info themselves.

I know we have a few guys on here that are job who may be able to point you in the right direction or I can ask my other half to dig out a contact for you if your interested

Not just the Met, either… Kent, Hertfordshire and the rest might be able to do summat with it, too.

Any further news on this project? i think it is a cracking idea and would love to see it take off.


Just to say I recently tried to vet an engine I was buying off someone and neither the Police, nor the DVLA could tell me whether it was stolen. The DVLA apparently didnt have the facility and the Police couldnt because of the Data Protection Act.

Most annoyed, I was. If that engine is stolen, I will be prosecuted, but I cannot determine whether this is the case. I have logged my complaint with the Police in an effort to document the above and cover my back.

Ridiculous, frankly.

Hallelluya… Amen… Having lost 3 bikes in 2 years 1 datatagged I think its a brilliant idea, I’d like to see something modelled on the 2 ideas from Rush hour racer & Button Monkey. The landrover type screen as a searchable database. Having had quite a few dealings with various insurance companies I think they may well be interested in some way, wouldnt hurt to ask for some feedback from them now would it?

It’s still on the cards, we’re just waiting to close off some other projects first.