jbond posted a photo:

Mmm. Claret for Christmas Lunch.

[from: Flikr Photos]

Needs no further explanation. Use it, don't abuse it. [from: del.icio.us]

[from: del.icio.us]

A post by Dave, led me to an article by John Battelle, from there John's blog, to reworking thoughts by Ross based on Marc Canter's Openlistings. Who says it's not an echo chamber! And note that MIT provided no contact info for John on his article and so no way to write back. I had to Google for his blog and then search for the corresponding post.

Anyway. First some immediate thoughts that ended up in John's comments.

1) Bloggers don't have time to pick ads. It's ridiculous to think of a blogger wading through Ads looking for the ones to show. But what I do want is to be a bit more involved in the placement process. I should be able to nudge along the Ad Agency saying I want "These types of ads related to these keywords but not those". Perhaps a one click solution on the Ad saying "Love, Skip, Ban" that only the blog owner could see.

2) "Internet-based ad is already a little piece of software". Really? it just looks like html to me... I think what you mean is that Ad systems should have an API so you can query details about a particular Ad. Good idea.

3) AdSense sucks for bloggers. I'm tired of my blog getting nothing but adverts for blog systems. Just because the reader is looking at a blog doesn't mean they want one too. Especially as they've almost certainly already got a blog of their own. And Google doesn't index the blog post fast enough and so know what Ad to show. A lot of blog reading is done within minutes of the post via RSS. Getting the right Ad a day later is too late.

4) See 2) I want to see Ad Systems (like AdSense) have an API so that Listing systems (like Craigslist) can automatically post Listings (for a premium price) as Ads. Let's have a decentralised eBay based on Ads that appear on Blogs. I think this is a big idea. The problem is that it drags Google further down the long tail and they may not be ready to do it. And if not them, then who?

Taking this further let's stand back a moment and think about the overall system.
- Advertisers placing Ads with lots of metadata. Competing against each other to maximise click throughs by adjusting the text, keywords, impression charges and clickthrough charges
- The Ad agency providing an API to let advertisers place ads automatically
- Listing Services (craigslist et al) providing moderation of listings and reselling placement on the Ad Agency. Automatically generating metadata, tags and keywords from the categorization of the listings on their system. They get revenue from the reselling but also get implied advertising because every click through comes back to their site where the listing is read.
- The Blogger providing loads of metadata about their blog both explicitly by keywords. Passing judgement on particular ads by marking them as good or bad. And implicitly by their page rank, content and links to and from them.
- The Reader, by generating impressions, racking up hits on tracking cookies, and clicking through Ads.

This is a rich, rich system with loads of explicit and implicit activity and content. The Ad agency should be able to build lots of feedback mechanisms to maximise the return for both the Advertiser and for the Blogger. I get the impression that current traditional advertising is all very manual. Traditional Adverts were bought from a small number of corporates as campaigns sold onto specific sites. Google (and Overture) has opened this up by providing a sell side system aimed at SME advertisers. But their focus appears to be getting their advertisers to do the work to maximise effectiveness.

The big idea from Ross is to now open up the demand side and build and provide systems that maximise the return for the Inventory sites that are showing the ads. For the Ad agency, this looks at first glance to be counter productive since they represent a cost. But in fact, maximising their return is actually making the ads more effective. Which should mean that they get more money from the advertisers. But we really don't have time to look through all the ads and choose which ones should be displayed. And as Ads have relatively short lives we'd have to keep doing it. We're already providing lots of information about Ad effectiveness on our sites in terms of clickthrough rates. The AdSense tools for Publishers let you block specific Ad URLs, but there's no other way for us to feed our views back into the system. This is too blunt an instrument for us. I think what we need is the sort of simple controls we have on the Last.FM radio. Just three buttons that only we see against each ad for "Good, Bad and Ban". "Good" says this Ad fits, give me more like it. "Bad" says This doesn't fit, reduce the likelihood of this ad showing and ones like it. And Ban says never post this on my blog again. It should be possible to provide an Amazon recommendation style page for the publisher where they can work down a long list of potential Ads setting each one to one of the three values (plus "no opinion").

The key here is that Publishers will do some extra work if it means they make more money. They just won't do all the work.

The Ad Agency API fed by listing sites, represents a new source of Ad Content. But it also adds value because they take some of the work on moderation and they can add metadata which provides additional information to maximise the rest of the system. Most of this article is about the demand side. But with my Ecademy hat on I'm very interested in the sell side as well.

As I've done some research on this area, what I've discovered is that there's a huge industry of support sites and services surrounding AdSense. The Advertisers side of the equation is getting very sophisticated. But equivalent support for the publishers is seriously lacking.

ps. It took a little while to do the search to find Marc's original post about OpenListings. This is not the first time, I've done a Google search for a specific named idea that appeared on a blog and had real trouble finding it. Another example of where blog developers are not terribly good at applying SEO to blog design, and where Google still has plenty of work to do on making search for blog content better.

I'm looking for a simple, easy to use sound recorder. It must:-
- Record Wave out and Microphone mixed
- Be able to output MP3
- Be dead easy to use.
- Preferably with a noise gate and auto record level control.
- Cheap or free

Any ideas?

Note about the sound source. I must have the output of both Wave Out and Microphone. But without playing back the Microphone through the headset. So it'll need to get the sound from DirectSound or the sound card without going via the Windows Mixer. It's easy to record Wave Out Mix and to enable Mic and Wave Out but then you hear your own voice through the headphones which is pretty distracting.

Any ideas? I've tried Audacity but it can only take windows mixer out. And it's too complex and heavy duty. I just want a device with three big buttons for start, stop, playback. [from: JB Ecademy]

14 months to write and publish a book! It'll be out of date by then. [from: del.icio.us]

My good friend Dave W has written a piece pointing out that the RSS 2.0 says one enclosure per item. Ok. I agree. But I'm still left with a problem. Over here on this table I have this one item consisting of a title, description and a permalink. All of them talk about what's on that table over there. Which is a BitTorrent I'd like you to use, and the Mp3 that's inside the torrrent and you can use if you still haven't got a BT client. On a third table is a whole lot of stuff that most of us won't be interested in like a mirror of the MP3, the lyrics, the metadata, AAC and WMA versions in various qualities, a rip of the DVD version of the video, and so on and on and on.

So ignore that third table and just concentrate on the Torrent and MP3. If RSS can't have two enclosures in one item (probably fair enough), then I can either have two items with duplicated title, link and description but then Aggregators will probably try and download both, so that doesn't work. Or I can have two RSS feeds and just try and encourage the users to use the one with the torrents. But then I'm forcing the user to make a choice that the software ought to make.

So the best solution I've seen is that we encourage a convention in aggregators. We put the Torrent pointer in the enclosure in the RSS. If the BT part can't find a seed it looks for the contained file in the same place at the same URL. Download it and then start seeding it for everyone else. In parallel, we can encourage people to write easier and easier to use BT trackers for the authors. And for people who run an aggregator that can't cope with BitTorrent (shame!), put a pointer to the MP3 as an HTML link in the description.

1) Skype Options, Handset, disable Auto sound setting
2) Open Windows mixer, make sure Microphone is included. Unmute the microphone. Set low. Set Wave Out to the same level
3) Open another Windows Mixer for Recording, select Wave out mix. Set low.
4) Install Audacity with the Lame MP3 encoder
5) Set Audacity to 22Khz, 16 bit mono
6) Set MP3 encoding to 64K or 32 K
7) Start Audacity recording
8) Use Skype to call Echo123 for testing.
9) Stop recording, check it and then Export to MP3
10) Done.

The trick is geting the levels right so that there's no clipping. Audacity has lots of effects too. And it's all free.The downside is that you hear your own voice in the headphones.

What we need is to route the Microphone plus Wave Out to Audacity. The Microphone to Skype. And Wave out to your headphones. You can't do all this just with the Windows mixer controls. It may be possible to do this with Virtual Audio Cables (VAC).

Stuart Henshall and Bill Campbell have a more complex version of this involving two Skype sessions and VAC.

[from: del.icio.us]

I got caught again yesterday by yet another IE6 CSS bug. If you make a TEXTAREA WIDTH:100% it more or less correctly makes it the width of the bounding box. Some times you have to force this by adding a div style="width:100%" round it. Do the same thing with an INPUT TYPE="TEXT" and the input field is always 100% of screen width regardless of the bounding box even if the bound is a TD. Aaaaaarrgh!

Meanwhile, there's now there's an SSL spoofing exploit in a fully patched XP SP2.

Our good friend Scoble has been telling us that Microsoft rebuilt the IE team some time ago. Well Ok, but what are they doing? And why isn't anyone apparently fixing the numerous bugs in the rendering engine?

Scoble has also been bragging about how it only took the MS Desktop Search team 7 months to release. Excuse me? Desktop search is little more than a fancy (and actually rather clunky) UI on their existing Index system that was already built into XP. What's so amazing about 7 months work?

There's something for everyone in The 46 Best-ever Freeware Utilities for Windows. [from: JB Ecademy]

More networking sites than you can shake a stick at to be found at:-

Home of the Social Networking Services Meta List - The Social Software Weblog - socialsoftware.weblogsinc.com [from: JB Ecademy]

Somebody just asked me how to search Ecademy with Google.
Just add site:ecademy to the search terms. eg Tools Importer site:ecademy.com

I'll add this to the site somewhere. [from: JB Ecademy]

Cooler than a very cool thing. mapping Flickr tags to locations. Then mapping the photos onto maps. It's shortcut to proper GPS embedded in EXIF, just tag your photos with the Zip/Postcode where they were taken. [from: del.icio.us]

Comment Spamalot from Ben Hammersley's Dangerous Precedent talks about the issues of MT comment spam on MT owners and hosting companies.

I'm curious about Typekey. I'd like to think but I'm really not convinced that a centralised identity system will help stop spam in the long term. And as I like the idea of decentralised identity systems, I'm even less sure about that. My experience with yahoogroups is that people will sign up to Yahoo with a temporary account, join a group that requires moderated entry, wait for approval and then spam the group. This does not bode well for any system, no matter how many layers of indirection there are, where the entry requirements can potentially be automated. And we haven't yet found a good way of enforcing manual involvement from real people.

As for Typepad, I'd encourage SixApart to work with SXIP as an alternate as soon as possible. As we have more and more identity verification systems, the pain of signing up repeatedly to each one is becoming too much. In this case, I'm not a user of SixApart products so why should I have to sign up to their identity system just so I can comment? I've already got a passport, yahoo account, blogger account, numerous Drupal system accounts, Slashdot, kuro5hin and on and on. Surely one of those would do?

DECAFBAD has got some very interesting thoughts about exploding a PC's functions into myriad computers scattered round the house and wider internet. Two bits really caught my eye.

Miscellaneous Thoughts about Exploded PCs - Archives - Blog - 0xDECAFBAD Blog : Furthermore, I have concerns that ditching the open, general-purpose PC could kill a lot of freedom. This is especially worrisome if appliance vendors lock into proprietary networking and communication standards, or require nasty licenses like for the CSS algorithm used with DVDs. I'd really want to see the fabric between appliances, the circulatory system between my PC's exploded organs, based on open standards and protocols. Right now, the general-purpose PC is virtual and fluid enough to dodge efforts to hold it in place, but firmware and hardware are practically forever.

This is I think critical. Every special purpose machine we're being sold now has some form of lock in. Even the humble iPod (and it's not so humble apologists) has siginifcant lock in to iTunes and Apple.

I’ve already ditched a central desktop computer, and sometimes this makes me feel like I’m living in the future. I used to have a machine that I used for everything— network services, web browsing, email, IM, file storage, movies, games, et al. However, slowly but surely, all the things this one machine used to do for me are being split off into simpler, dedicated devices. It’s like all the internal organs of my old PC are being splattered throughout our apartment.

The Network is the Computer...

I'm seriously considering upgrading my Win XP laptop with a larger hard disk. But I'm seriously worried about the process. Can anyone recommend software and instructions to do this? Preferably cheap or free.

This raises another question. When you've spent several months or even years configuring an XP desktop and you finally decide to upgrade to a new machine, what are you supposed to do to copy all your data, applications and setup across to the new machine?

Hard disks are cheap these days. An 80Gb laptop drive is under £80. Desktop drives can be had for 2-300 Gb for similar money. So it's crazy to be struggling with a laptop that only has 20Gb or a Desktop that has 40Gb which is what was shipping a year ago. But it's not exactly easy to swap drives and there doesn't seem to be much help from the operating system to do it.

So let's hear from the MS experts how you would do it. [from: JB Ecademy]

I'd like to see the new MSN Deskbar search integrated into Firefox as a search plugin. Lazyweb should be redundant here as I'm sure somebody is already working on this. The same goes for the Google Desktop search and the forthcoming Yahoo/X1 desktop search tool.

Linux based bootable CD image [from: del.icio.us]

What is revolutionary here is that Prodigem completely automates the entire process of setting up bit torrent sessions for the distribution of your content. You simply upload your content via the web and with the click of a few buttons, the Prodigem serve [from: del.icio.us]