×

Linking to a spotify search with both track and artist

Published februar 16th, 2017 by

It seems that Spotify has made some changes to how one should format links when performing more complex searches, causing some problems with one of the services I work on. This post describes my solution.
(mer…)

Posted in Code, Webdev

New mixtape feature on Urørt.no

Published mai 29th, 2008 by

We are doing a soft launch of a new feature on Norwegian music demo site Urørt.no (nrk.no/urort for those lacking ÆØÅ), namely the option to create you own mixtapes. Each mixtape can contain 20 songs, selected from the over 57000 unsigned mp3s uploaded by the 22000 bands on the site, and you can share it with friends by sending them to Urørt, or by embedding the player in your own website, like this:


Gå til Urørt for å lage din egen miksteip

A while back we also took the time to contact facebook to have them add our domain nrk.no to their Share Partner program, which was necessary to be able to automatically embed flash in their site, when you use the facebook share feature.

Oh, and every mixtape also comes in the form of a podcast rss so you don’t have to manually download each track if you want to keep all the songs on your harddrive or you ipod.

So if you are like new music and like showing it off, then head on over and create an account.

ur�iksteip screenshot

Posted in NRK, Webdesign, Webdev

How to present a music festival-guide?

Published mai 22nd, 2008 by

My colleagues over at Lydverket published a great festival guide today, and I have been thinking of ways to enhance it.

We recently made a page that fetches artsts info from several websites, like imeem, youtube, last.fm, flickr and wikipedia, for use together with NRKs radio playlists, and I wanted to miks Lydverkets guide with this aristinfo service.

The easiest way would be to simply link each artist in their festival guide to the corresponding arist page, but the first problem is that the layouts are quite different. To solve this we changed the artist info page to take a folder id as a paramter, so that we can change its layout dynamically. And is so happens that we have a folder that has no layout at all, which means I can do something like this:
http://www2.nrk.no/spillelister/LaatInfo.aspx?artist=Santogold&ppfolder=p3

This gave me the sudden urge to just insert an iframe dynamically under each artist name in the festival guide, so that you can simply click on an artist, and get a ton of information, music, videos, etc, but the problem with this is that there is simply TOO MUCH information.

Luckily, the artist info page is build up of several smaller, freestanding pages, placed in the artist info page using iframes (mostly). This means that I can break it all up, and perhaps create a tabbed interface for each artist in the festival guide, which I will work some more on tomorrow.

But until then I want to test a few other things:

A few minutes ago I went through the list of the 17 bands that are scheduled to play at the Hove festival (www.hovefestivalen.no), seached them all up on youtube, and placed them in a standard youtube playlist. I then created a custom player, that looks like this:

Hove – Day 1:

Hove – Day 2

Perhaps it could be as easy as this? I you are reading a festival guide, what else do you need than a fast way to check out each artist? Is one youtube video is for me enough to tell me wether or not to see a band live? Perhaps.

I think I will insert this little youtube player in the page and ask around tomorrow if it works or not, and then I will do some tests with the youtube api to see if a can extract my playlist as xml and present is as I wish.

Posted in NRK, Webdesign, Webdev

Four ideas I believe in

Published mai 2nd, 2008 by

PhotonQ-Woman 's Thoughts aKa Complex Memetics
photo credit: PhOtOnQuAnTiQuE

Readwriteweb wrote a post a few days ago about the lack of some New New Ideas for web startups, which I commented on. I wrote down a few old ideas I wish someone would like (and then steal):

photo credit: Whatsername?

1. Anything that uses the web to connect people to each other in a physical location

Bilderesultat for holding hands
(cc)

Any service that makes it as easy to approach people on the street etc as it is on Twitter or Facebook. There are heaps of services that touches on this field, but not many that works or have enough members to make it work. So maybe there is room for a service that connects it all, and shows you on a mobile map where people who are open to «hellos» are. It could be a facebook app (facebook has 1 million members in Norway, of 4.7 million inhabitants). Your mobile phone would simply broadcast itself with a «hello, talk to me» to a google map and perhaps bluetooth. A cross between facebook, twitter and friendfeed, but with the purpose of connecting physical beings in a physical location.

three thumbs down
photo credit: rick

2. A web based service that lets you record video from any television channel.

It makes no sense for everyone to own their own PVR/TIVO/Mediacenter and record to disk at home, when you could do it sentralized and distribute it with bittorrent. The business model is like this: 10 gb storage for free, if you are too lazy to delete, or simply want to keep things online longer, you pay extra. Everything you program to record is automatically downloaded to your harddrive if you want to. The gold: You get access to EVERY channel on the planet. I wanted to create this in 1998, but never got around to it, so you can have that idea for free :)

3. 24/7 live streaming video from your mobile device, with a twist (object identification)

Livestreaming from a cellphone is not new, it exists today, but here is the twist: 1) a small sensor analyses your brain-response to what you see or hear, and anything that creates an «unusual» reponse is indexed and flagged for later. 2) it also has a built in visual «thesaurus» (think wikipedia meets The Terminator), that can look up information on your surroundings in semi-realtime. So if you wanted to know whether or not to eat that delicious-looking red mushroom with the white dots, the system could advise you not to.

Did you know that American kids can identify 1000 products before they start school, but only know the name of 10 local plants (according to the movie «The 11th hour«)?

4. A google earth for the body

Read more about «A google earth for the body» here

Posted in Ideas, Web2.0, Webdev

Code: Gameinfo + Gametrailers + Youtube Mashup

Published mars 8th, 2008 by

Norsk? Du finner en norsk versjon av denne artikkelen på NrkBeta.no

I recently made a Gameinfo service for spiller.no. It pulls in licenced gameinfo from our partner GameXplore though XML, and combines it with a search for related videoes from Youtube before it is presented on the page.

Example: FIFA STREET 3 Gameinfo – NRK Spiller

Youtube + Google Api
I made the Youtube part as a standalone page, and I include it with iframe. The reason I used Iframe was to make it easily reusable on other projects where it would fit to do something similar. I can for instance include it in my blog like this:

The page takes any search term as a GET, in this example it simply queries the youtube google api with the title of the game, with a few extra parameters (limit on the number of returned movies + a filter to remove «adult content»). I spent some time to create the XSLT to convert the xml from youtube/google to html, which is basically the same format for many of Googles services. I spent hours trying to find out if someone had created this already, but since I couldn’t find anything I had to create it myself. If you want to learn more about the youtube api I suggest starting at the Youtube Api group on Google Groups.

Open source

I have gotten some emails asking for the xslt for the youtube search, so I wanted to share it here:

This is my input, an example Youtube search

This is my example output:
http://pappmaskin.no/opensource/youtube/youtube.php

And here is the xslt I used to produce the above result:
http://pappmaskin.no/opensource/youtube/youtube3.xsl

If you have any questions, feel free to ask in the comments.

Spam? Porn? Nope!
I was of course a little worried about what the youtube search would return, so when testing I went through quite a few games to see if the search returned movies that were unrelated to the gametitle, or contained nudity, but so far it looks really good, the youtube videos are actually the most interesting part of the service, so I’m really happy with it!

Blip.tv gametrailers
Yesterday I also added gametrailers to the mix, check out the gamepage for Assassins Creed for an example (click on the huge image at the top of the page) to make it play. I have bought a lisence for the wonderful Jeroen FLV Media Player which I use to present the gametrailers.

New gametrailers page, based on blip.tv
After that I spent a few minutes to create a new page that lists all the latest gametrailers published through GameXplore. I use the Jeroen Flv Media Player for this page as well. I created this yesterday, so I have to debug it a little bit before we start to promote it, and I also have to consider to increase the video buffer time, it seems like blip.tv har pretty slow servers. Right now there is a five second buffer, which should be more than enough, but even with broadband connection the video stops and buffers every five second. I don’t want people to wait forever for the video to buffer, but it is super irritating when the video stops every 5 seconds to buffer, so I have to decide what to do with it. One (hard) solution could be to download all the .flv to my server and play them from there… I am a big fan of blip.tv, they have the best player around, and their cross posting abilities are great (at least in theory, I still haven’t gotten it to work with this blog or myspace), so I am a little dissapointed to discover that the blip.tv servers (or lines?) are so slow to deliver video. (to Norway at least).

Thickbox
I love the Jquery javascript library, and I love Thickbox, so I decided to use it on the Screenshots on the site, popping them up in their own little box. I modidied the thickbox script and css slightly, replacing the animated loading gif and moved things around a little bit, and I am really happy about the outcome.

Add the spiller.no Gameinfo search to your browser

Spiller.no Gameinfo

Sjekk ut info, screenshots, trailers og anmeldelser på nye og kommende spill:

So there it is, hope you like it :)

To embed the search in your site:

<div class=»placeholder»><h3>Spiller.no Gameinfo</h3>
<span style=»color: #fff; width: 160px; font-size: 0.8em;»>Sjekk ut info, screenshots, trailers og anmeldelser på nye og kommende spill:</span>
<form id=»gamesearchform» name=»gamexploresearch» enctype=»application/x-www-form-urlencoded» method=»get» action=»http://www.nrk.no/spiller/gameinfo/?»>
<fieldset>
<input type=»hidden» value=»143″ name=»task»/>
<input style=»padding: 1px 0 2px 0; width:100px; border:1px solid #F7F7F7; border-width: 0pt 1px 1px 0pt;» type=»text» value=»» maxlength=»500″ name=»finn»/>
<button type=»submit»><span>Søk</span></button>
</fieldset>
</form></div>

Posted in Code, NRK, Open Source, Portfolio, Webdesign, Webdev

Grafiskformgivning.no

Published mars 4th, 2007 by

Customer: Oh! Ohren Grafisk Formgivning
Delivered: Advice and production of a flashbased minisite. Webdev, Flash (AS2), design and development of an XML-driven gallery, contact form with PHP backend.
Designs delivered by customer as PSD.
Release: March 2007

Posted in Portfolio, Webdesign, Webdev

Userscripts.org: Amazon Sambok and BIBSYS ISBN Search

Published oktober 1st, 2005 by

La caverne aux livres
photo credit: gadl

Hey, my first Greasemonkey userscript!

Userscripts.org: Amazon Sambok ISBN Search:

This is just a fast remake of the Hight Library Linky script. Its my first script, just wanted to check out how hard it would be to rework. Luckily for me it was super easy. It inserts links to two norwegian bibliographical databases.

Bibsys, containing material from universities and colleges.
And Sambok, containing material from most Public Libraries in Norway.

Pretty useful, at least in theory, since ISBN search is quite limiting. But it will have to do for now.

Posted in Code, Ideas, Portfolio, Webdev