One of the defining notions of my professional career for many years now has been that software development is far too difficult, especially considering that we tend to reimplement the same basic functionality over and over again. Stuff is created by the user, generally through some kind of forms-based interface or retrieved from a remote source, it is stored somewhere and (increasingly) send over the wire to other people. My standard AllPeers pitch extends this by explaining that, whatever difficulties traditional client/server development might present, peer-to-peer architectures are an order of magnitude harder. Hence our decision to make AllPeers a platform that we can build our apps on top of, avoiding the need to perform this rocket science over and over again.
I was reminded of this when I saw Ben Goodger’s announcement the other day that Places, a greatly enhanced bookmarking/history module, will not make it into Firefox 2.0. I know a lot of the Mozilla and Firefox developers, and they’re exceptionally talented, so the fact that theoretically straightforward features like those offered by Places take so long to get up to production standard strongly reinforces my point about software development in general.
Now imagine that Places were meant to let you share bookmarks with people, del.icio.us style, as well as managing them locally. A peer-to-peer network topology would be of huge value since it would greatly reduce the reliance on centralized infrastructure. See where I’m going with this? It’s going to be a while before we have an AllPeers API to let developers do this, but I hope that it won’t take that insanely long. And when we do, things like Places-on-social-software-steroids will be a snap.
“Qwerty Maniac”, one of our active Beta-Tester has posted on his blog a review of his first week with AllPeers as well as some screenshots.
Abstract:
AllPeers shocked me at first with a 3.56 MB ; I mean I haven’t installed an extension sized over 1 MB in my Firefox usage history. Whilst installing I prayed Firefox not crash on the restart. And surprisingly it didn’t and neither did the start up time reduce. I registered my BETA test with my activation code and was ready to go vroom. Wait there is more, Firefox though didn’t slow down, it sure took over 50 MB of RAM at startup itself, so the effect of the extension was showing.
Ced’s last post is a good lead-in to what I wanted to say about the current state of the beta. As he points out, we’ve got very ambitious goals for our product, and this means making sure that the foundations are solid before we flesh out the edifice on top.
We’ve had great feedback from the first batch of testers. Nothing unexpected or catastrophic, which is a great achievement in itself. We did unearth a couple of architectural issues we have to resolve. Like everyone I’m excited about the potential of the “Drag-n-Share” technology in our consumer application, but I’m also juiced about the world-class platform that we’re building. By catching these things early we’re ensuring our ability to roll out new features smoothly like comments, tagging, instant messaging, richer profiles and a whole lot more that I’ll keep to myself for now, thank you very much.
Bottom line: we want to clean up some of the more significant glitches that have been identified this week. Nothing that major, and I’m confident that we’ll have another version ready next week. A few of the infamous “known issues” on our website will also be fixed (and some new ones added, no doubt). Then we’ll invite some more people, and we’ll see how the new improved software holds up to their onslaught.
The following is an abstract from a talk I made a few days ago about AllPeers and its future evolution.
“I have been running large online communities for the past 10 years. In 2002, with my previous company, we were given the NewMediaAge award for Best Consumer Website in the UK just in front of the Guardian and the Observer. For that reason, sometimes people ask me “what is it with MySpace, why is it so successul?” The fact of the matter is that there is absolutely nothing special about MySpace and if you are old enough to remember the good old days of the late internet 90’s you will certainly remember Geocities which in 1999 had 35 million members before it was acquired by Yahoo for a share-only deal worth $4.5b. Geocities was a service to allow non-technical people to build a home on the web. Just like MySpace today.
MySpace, just like Geocities at the time, is about sharing. Sharing your identity, sharing your tastes, sharing your pictures, sharing your music, sharing your videos. It’s about existing online. The problem with MySpace is that it is built on a publishing platform called the Web. The Web was never intended for sharing but for publishing. As a result, the user experience is disastrous, the site at peak becomes slow and they have to keep on upgrading their infrastructure to keep up with the growth currently estimated at 250,000 new members a day.
AllPeers has been built from the ground up as a sharing platform. It’s like MySpace but on steroids. In 1999, the web servers used by Geocities were as powerful as today’s consumers’ personal computers. Almost 10 years later, we are using 0.0005% of that local power to browse the web and publish on MySpace. The idea behind AllPeers is to leverage the power of the user’s machine to allow them to maintain and control their online existence without the need to upload files to a remote server. Without the need to wait for pages to refresh. Without the risk to expose your private details to lurkers. With AllPeers, you decide who can see what and with the 20 pages long list of features we have, you will soon realise that it is about time that we rethink how we use the network and give the power back to the users.”
In full disclosure, my firm is an investor in AllPeers. I have accepted Matt’s invitation to share my thoughts and to take the opportunity to encourage innovation and risk, both of which are dear to me as a venture capitalist.
Taking a step back from the stress of the current product launch, it has occurred to me how truly global Allpeers already is. With French and American founders, living and operating in Prague, the company epitomizes global entrepreneurship, where skills, drive and ambition are the key ingredients and not geographic location. As the Internet continues its spectacular growth, companies with a DNA which allows them to openly consider global strategies will prevail. Being US centric will no longer suffice.
Interest for the product has also come from around the globe as more than 50.000 people have signed up for the beta – a few years ago this would have sounded like madness! This is indeed further proof that the early adopter community has become borderless and is slowly become less American, something which is probably in the long term interest of innovation and global competitiveness.
That great ideas can be born in Europe is undeniable, implementation and marketing have long been its Achilles heel. In the current frenzy surrounding consumer startups so visible in the US and in China, I can only encourage people to take a serious look at what Europe has to offer. Good things are happening.
Of course, users will be the ultimate judge, so judge on…
As I am writing these lines we are starting to send invitations to the first 100 people who registered with us. The emails will go out slowly as we monitor the process. Before you start complaining that you did not receive the invitation please note that today we are only inviting 100 new people out of the 10,000 who agreed to participate in this first phase.
Of course it is highly possible that there will be problems. This is why we are doing a beta test phase: to test the software and our network. Rest assured we are not sitting back now. We already have a long list of bugs, issues and new features we are working on and we are waiting for you to help us track more bugs so we can kill them efficiently before sending the software to the general public.
Please use the ‘Support’ forums you will find in the AllPeers toolbar to send us your feedback. All messages go directly to our development team.
Congratulations on the first 100 and please help us improve AllPeers and make it even more simple to use.
Last week was so busy that I haven’t had time to talk about the week before last, which was eventful in that Prague was hosting the latest LinuxExpo. This was significant first of all because I gave my first ever presentation in Czech, entitled “AllPeers - vývoj komplexní aplikace pro Mozillu”. (On the off chance that someone reading this doesn’t speak Czech, their not-very-accurate translation is “AllPeers - Development of a Comprehensive Application for Mozilla”.) I was slightly nervous that I would be groping for words, but actually it went very smoothly. You can find my slides on their website, although they aren’t particularly edifying without the accompanying commentary. I am planning to make a proper screencast on this topic when I have time.
The other highlight was the visit of Tristan Nitot, President of Mozilla Europe. The Czech Mozilla community, CZilla, put together an evening of dinner and drinks to hang out with Tristan, and kindly invited us. It was great to meet more of the local community and have a chance to chat with Tristan in a relaxed atmosphere. Though we’re all open source proponents, we have different takes on its evolution, particularly as regards future business models. After a few glasses of red wine, Cedric cornered Tristan and made him promise to do a podcast with me, so I hope we’ll get a chance to make this happen.
So the good news is that we have a very nice AllPeers version that is working well internally, and we’re happily consuming our tasty dogfood. The slightly less good news is that we still haven’t finished setting up our new registration server and superhub at the data center. We’re running the new Microsoft compiler (Visual Studio 2005), which we discovered requires us to use the trunk version of Mozilla (i.e. the future Firefox 3.0) instead of the branch (i.e. Firefox 1.5.0.x) that we use for the normal AllPeers client. This has required more work adapting our code than I expected, but we’re nearly done now. We’re now targeting Monday to send out the first invites.
My apologies for the delay. In the interest of expectation management, I should also point out that we still have a lot of work to do to make this version as awesome as we want it to be. But we’re getting there. More news very soon.
I’m interrupting our regular programs to inform eager readers of this blog that the next batch of beta-testers will be invited next week just after the long Easter week-end. A few hundred lucky people will receive an invitation on Tuesday or Wednesday once we have moved our SuperHub to its new data center.
Last night, we integrated our new icons designed by YellowIcons who did a great job. Here’s what they look like at an 128×128 pixels size (they are actually smaller in the software):
|
|
|
|
Home
|
Group of Buddies
|
|
|
|
|
Buddy
|
Folder
|
|
|
|
|
Video File
|
Audio File
|
|
|
|
|
Picture File
|
Share
|
I’m sure some people in the comments will seize this opportunity to order us to stop teasing and to release the software. This is what we are actually doing. Step by step. It will take some time before everybody gets to test a version. Meanwhile I wanted to share the work from YellowIcons.
I’m happy to announce that Mangrove’s Mark Tluszcz (of Pressure Point fame) is joining us here as the newest Peer Pressure blogger. He’s offered to help dial down my workload during this very demanding product release period by adding his voice to mine and Cedric’s. I’m sure anyone who listened our podcast will agree that his perspective will be most valuable. Please join me in welcoming him on board!
The AlwaysOn OnHollywood 2006 conference is where cutting-edge technology innovators from all over the planet converge to mix it up with the entertainment and media establishment. Since we believe our technology can change forever the economics of the media industry, we have secured a slot for Matt to show our product. Matt live on stage in Hollywood. His dream is finally coming true.
So if you want to see the vaporware 
in action or complain to him directly why it took soooooo long for you to receive your beta version you know where to go.
AlwaysOn OnHollywood is slated for May 2–4, 2006 at the ultra-fashionable Roosevelt Hotel in Hollywood, California.
My friend Mike Hogan wrote to me to tell me about his company ZiXXo, an online couponing system that has just launched:
ZiXXo is the only company today offering self-service coupon creation, coupon management, an API for coupon syndication, and many more innovations. Please see our overview screencast and our advertiser screencast (if nothing else they will help with the pronunciation of ZiXXo). For more information, please see ZiXXo’s media center.
I’m a proponent of opt-in advertising, where ads are designed to be informative or entertaining enough that consumers explicitly ask to see them. This contrasts with coercive marketing like TV spots or web popups that try to force people to endure annoying product plugs against their will. Seen in this light, the ZiXXo model is very appealing. Mike can go on for hours about the huge community of coupon clippers who invest incredible energy and ingenuity to gather coupons. Definitely the best kind of opt-in, and done right, this could have a revolutionary effect on online advertising. Local businesses who are trying to find the right approach for marketing on the web would do well to take a look.
If you don’t know what to do from the 15th to the 17th of May, come and see us at Innovate!Europe 2006 in Zaragoza, Spain.
AllPeers has been named ‘Guidewire Group Global Innovator’ and we have been selected to showcase our product at Innovate!Europe 2006 by the very same people who are organising DEMO in the US: Guidewire Group.
Innovate!Europe is a great opportunity to explore the most promising emerging technology trends, business opportunities, and emerging companies across Europe. See you there!
Phil Sim laments the decline of the “tech blogosphere”, as evidenced by the fall in visitors to aggregation sites like Memeorandum and TailRank. (Since I get eviscerated for the slightest divergence from according-to-Hoyle formal written English, I’d be remiss if I didn’t complain about the fact that Phil uses “nadir” to mean it’s exact opposite. Phil, the word you were looking for was probably “zenith”.) One problem with this contention is that the evidence is hardly conclusive. The fact that two popular sites experienced lower traffic over a short time period could easily be a statistical blip, or it could be due to a new competitor siphoning users away from the incumbents.
That said, my subjective impression is similar to Phil’s. There is less vim in the tech blogosphere. My analysis is sharply different, however. I don’t think that we need or want to “start a new wave of smart people blogging about tech.” There’s already far too much quality stuff being written for anyone to have a chance of reading even a tiny fraction of it. What we do need, undoubtedly, are better filtering tools to help us find the gems hidden among the white noise of ubiquitous internet dross. Memeorandum is useless when it comes to finding great stuff written by lesser known authors.
But more to the point, let’s face it: a large component of the blogging phenomenon is a fad. Once upon a time, everyone had a pet rock, and nowadays everyone has a blog. Unlike pet rocks, however, there is intrinsic value to blogs, and they surely won’t vanish completely. But the idea that we should all be blogging is just silly, and it’s hardly surprising that a lot of people are starting to lose interest now that the first flush of novelty has worn off.
No, the real trend to look for is the shift of professionals into the blogging (and podcasting, videocasting, etc.) arena. Why do journalists, radio/TV personalities, musicians and other content producers still have to be tied to the apron strings of a larger organization like a newspaper or TV station? The biggest issue is the lack of sufficiently robust business models, but that nut is going to get cracked eventually. Like every trend driven by semi-obsessive enthusiasts (like myself), the first wave of blogging was bound to lose momentum. The next wave, driven by cold hard cash, will hugely increase the quantity of “smart people blogging about tech”… but only available for a fee.
Three letters: P2P.
Ehm, two letters and a number, actually. Whatever.
