(From the old ideas site)

Almost everybody in the IT industry has a horror story about an IT recruiter. This is a proposal for a system to help employers identify potential candidates based on peer approval and selection. This is described here as a software system, but it wouldn't necessarily need an IT system.

An employer identifies a position, and on a website describes the kind of work that is involved, what skills are desired, the kind of company it is, etc.

Potential recruiters who work in similar roles within the industry (but who aren't interested in the role on offer) put themselves forward. The employer can recruit as many or as few of them as they like, based on their experience and ability to understand the kind of candidate they're looking for. Together they identify key criteria that the successful candidate must possess.

Candidates on applying for the job put forward a CV and forwarding letter that are anonymised (no age, gender, race, nationality, names, or any other identifiable traits are revealed to recruiters, but are available to employers).

Recruiters then go through the applications to separate those who obviously aren't suited, from those who likely to make a reasonable candidate. Rather than just looking for certain keywords, they apply their own experience of the industry to make an assessment of the individuals.

Once a short-list is prepared, the employer is asked to 'sign off' before the next stage occurs. At this stage they are able to remove people from the short list who they feel would not be able to make an offer to - e.g. a former employee, an employee of a firm they have a no-recruitment agreement with, etc.

Of the remaining candidates, the recruiters can then conduct a standard test. If more than one recruiter is involved, they work together to compose a series of questions that will be asked of all candidates. These questions are then put to candidates within the system, and they have 24 hours in which to respond. At no point are the recruiters' identities known to the candidates or vice versa (thereby stopping friends from suggesting friends).

Recruiters than mark candidates based on their answers against a set of criteria as agreed with the employer at the start. A short-list of candidates suitable for interview at the premises is then given to the employer, who can be sure that the list is not composed of those who have won at "buzzword bingo" but who have demonstrated ability within a peer group.

On hiring a candidate, if the candidate chosen is one the recruiter(s) recommended, a fee of 2% of the first year's salary is paid. This is cheaper than traditional recruitment and provides incentives for the recruiters to find the best candidate.

(From old ideas site)

When's the next bus at your nearest two stops ? How about the next 5 buses ?

Frankly unless its a regular route we don't know, so we don't use it, and we don't care, so we drive.

If it was easier to find out, maybe we'd use it more, and their would be less congestion on the road... err so I can drive about quicker ;)

-Can we filter busesto exclude certain companies (i.e. stagecoach - and will their be any left after ? Maybe we should say 0 buses, but 5 filtered..reveal?)

-Maybe we show the bus stops nearest to you... 5 of them maybe ? Why not make the number configurable in your user profile, where you'd also have your 'regular' routes / stops / times & direction of travel / your regular places (home/office/pub) ?

-Why not overlay this information as a popup layer from the bus stop icon on the map ?

-Maybe to start the whole thing off you just click on two places on a map, instead of all this existing entering street names and

Information / API's involved:

-Bus times : Department of Transport, GMPTE/A, Bus Companies

-Bus Node Db : I have the Manchester one from the DoT (avail for non-commercial usages)

-Maps: Google ? Open Street Map ? Multimap ? Yahoo ?

Reference Web Sites:

-http://nw.aim....
(shh don't mention the A_tos O_rigin developed one!!I'll cry as they're proper nasty peeps!)

Bus Comanies:
-Finglands
-First
-Bullocks

Extensions:
-Bus on time ? ... maybe user updated timetable if a journey is regurarly late / early ? (Though maybe they're just on the wrong instance of that route ?)

Objects:

BusCompany{Name,(Routes)}
BusStop{Lat,Long,(Routes), Direction}
BusRoute{Number,PtAPtB,[[x1,y1,t1],..,[x99,y99,t99]]} t=time, x=lat,y=long

User{name,pwd,email}
Profile{PreferedRoutes, PreferedStops, [TravelTime,Direction], HomePostcode, WorkPostcode }

Routines:

routes[]=Search.findAllRoutes(bustop)
services[]=Search.findNext5Services(routes[],bustop)

Questions:

How do you encapsulate the Route in words ? i.e. "Bus Travels along Palentine Road, Joining Wilmslow Rd, which turns into Oxford Rd,...., reaches Picadilly" - Maybe that should be an extension ?

(From the old ideas site):

Users of mail or news readers are used to sorting content based on source - e.g. from addresses, maybe subject lines, etc. - but this has traditionally been quite clunky and inelegant.

This idea is quite simply to take either Bayesian or LSI filtering currently used to separate spam from legitimate content, and to apply it in a much broader sense.

For example, I might wish to catalogue information about buy-outs or IPOs into one folder, articles about music I might like into another, articles and mails about events into another. By moving articles or mail into folders, the client learns how to automatically filter, much like a junk mail filter learns. That means no matter what the source, information can be categorised non-programatically.

In addition, a reader aware of other sources might be able to identify content from other sources I'm not subscribed to that I am likely to enjoy, because it would score highly in the context of one of my automatic category folders.

(From old ideas site):

http://www.microsoft.com/surface/

It looks a lot of fun, but tied to MS tech like Zune. Now we have the open GL desktops and stuff on Linux, it would be a shame not to use some of the ideas here and see if there isn't a way to make photos bigger, email mom pictures of your dog and find places to take your kids that isn't proprietary :).

I also think their vision is limited, data goggles and so on overlaying the entire external environment would be more interesting. I think a tool like this would be a great way of walking through data warehouses and looking at output from DSS software.