Development Tips according to Flickr (FOWA Miami 2008)

After a disastrous trip I finally got here (albeit tired after a 4am NYC connection to Fort L followed by drive to Miami). Thus far some OK stuff .. my picks:

Cal Henderson (Flickr) is talking about process in development. He’s a funny presenter. Thumbs up from me.

Here are some development pointers, in his opinion:

• 10% spent on developing tools to help develop
• Version control – Emails on commits notes – send them out and have them filter into an email
• Version control – Invalid syntax hooks on commit – refuses if invalid
• Amend your own interface for your bug tracker – Bug trackers are often simply too complex
• Continuously test when integrating .. i.e. write tests
• Use a Tinderbox and combine it with alert
• Incremental releasing – Permanent manufacturer but with:
• Dev > Q&A > Stage > Production
• Flags not Branches – use Feature flags; this maintains consistency against the head for testing purposes
• Single button site deployment – however complex it gets; it should still be one button to deploy
• Deploy log – simple, but effective – highly visual error messages are important
• Deploy process – Automate common tasks

Yahoo Open Search Platform brings us a step closer to Web 3.0

I’m surprised the announcement of Yahoo’s Open search API isnt bigger news. I cant help feel it would be were it Google.

It’s great news for us as we can provide access to Rummbles external content and integrate Yahoo Open Search to embelish the search results with extra information. This is a pre-cursor to the obvious next step which would be for us to identify who is doing the search and providing personalised content back to the Yahoo Search results, for that individual. This would mean they have to be logged in to Yahoo … We will have to check if we can get access to that persons information (perhaps their OpenID?) to then tie it to our trust network.

This would be much easier – and will be! – once the world embraces a standard or platform along the lines of Googles Social Graph and/or equivilent (think XFN etc which Social Graph is based on…)

You can see where I’m going with this … we can then personalise Yahoo’s search results for location based content in the same way we do for existing users with-in our network, but without them stepping foot inside the Rummble website. Cool, huh? … who said Web 3.0 doesn’t exist.

VC Funding the Loic LeMeur way

Full marks to Loic LeMeur for his detailed and open post on his Seesmic funding round. Admittedly he’s got plenty to be happy and shout about – but its good all the same to see a post with an open description of the process and who he got on board and when.

I generally agree with his comments regards the sometimes-negativity of Europeans (including Brits) when it comes to start up and funding. His comments about shipping out to the U.S. ring home for me in many ways;  we may well follow suite.

Best of luck to him and Seesmic.

Facebook statistics stumble … is it a blip or is the madness cooling?

OK so its been a while since I posted. In truth, there is a ton of stuff I want to write about, but things are crazy here. There is so much to do – bootstrapping doesn’t seem a suitable term – I think we’re whole-body strapping; but all in all its busy because things are good. We have got a lot of interest and even the journalists are taking an interest in Rummble.

So, whats this about Facebook? Well, its being widely reported that Facebook stats are levelling or atleast slowing. Really, no one should be suprised about this. Facebook is not worth $15billion and while it is a fantastic product, I believe it is also going to go through a tough time in the coming months, as it becomes a victim of its own success. There are BIG question marks over the Application Platform. Technical it works well, from a business point of view it has justified much of the hype and valuation of the business. Many people though question their interest in continuing vombie bites. I have 160+ notifications which I have ignored (and will continue to) as I really dont want a vombie bite or hot or not or to find out who has a crush on me. The vast majority of applications are rubbish (with the exception of Rummbles of course which is a shame as there ARE good and useful applications out there, but they are being overshadowed. I’ll talk more about application platforms in a future post, but for now we wait and see what will happen to the Facebook growth.

Ive spoken to more than a few people who find the torrent of dumb notifications and invites, pointless at best…

It could never continue at such a pace and I suspect while the growth will continue to be great in new territories as the Facebook craze takes hold, Facebooks job over the coming 24 months for existing markets will be:

• to retain their existing users
• to retain the user traffic (rate of visits and eyeball time)
• to continue to improve the filtering of an ever increasing amount of information that hits your Facebook profile
• to work out how to get users to actually click on Facebook adds, without setting off a revolt (see Beacon).

For more statistics info, see the trusty Techcrunch Facebook statistics article here.

Lift 08 – Geneva – The First Day

Unfortunately, as my flight arrived early morning, despite only grabbing one hours sleep after working until 430am, I didnt get to lift until lunchtime. There was a fantastic list of afternoon sessions – so much so it made it hard to choose. The sessions were actually workshops; in the evening there was a Lift Venture Night with a bunch of companies giving a short pitch and a panel (Pierre Chappax, Robert Scoble, Ouriel Ohayon, Jasper Bouwsma, Fred Destin and Faris Sabeti).

This was followed by a drinks and cheese-sticks mingle outside. As one would expect I met some fun people and made some interesting contacts. After seeing Robert Scoble mobbed everywhere he went (well, almost at Le Web 3, I finally introduced myself this evening at the relative calm of Lift. I met then went in search of food and joined a mixed group from Zurich, London, Hamburg and Helsinki at a nice restaurant called Cafe du Parc des Bastions. We found this place thanks to some friends of Bjorn Ognibeni and Sarik Weber, who I met over a glass of wine, who were already there. This was after a couple of false starts, including to go here but they had apparently (at 9:30pm!) already stopped serving. We’re rather lacking Rummbles for Switzerland so I have of course immediately Rummbled both! Once again it just proves the desperate need for Rummble, as I tried to Google for a restaurant on my Blackberry, then cut and paste the address to the GPS maps, then …… <sigh> It really should be easier. With Rummble, it will be!

I’m looking forward to tomorrow, the day kicks off at 9am with a talk on “Online Environments”. All in all my first impressions are that I’m going to enjoy this conference very much; once again a good mix of people -but perhaps will less of a commercial feel than DLD.

The Problem with e-businesses

The problem with online businesses, is that its a gift to the blank face of corporate might. It enables and encourages, automated dumb impersonal service – the plodding and continued, repetitive email response as the computer systems reply and go through their business processes.

Ebay is no exception. I’ve been explaining why I will not pay the £14.50 sales commission, for something I didnt sell, for nearly 7 months. They have even engaged a debt collection company – sent multiple letters; I’ve now refused in principle as my initial enquiry got no response. Typically of large companies, the only thing they ARE efficient at is sending me court threats, reminders and getting their subcontracted debt collectors to chase me for £14.50 . In 7 months I’ve had only TWO human responses; the situation continues. Having written back twice to the debt collectors, ebay have now again reverted to the initial cycle of automated emails asking me to pay.

Its not only maddening, its pathetic. Online business can be SO MUCH better than this. The cost reductions that this technology enables, the automation for all the transactions that go smoothly and without fault, should mean that personal and attentive, personalized service CAN be provided when needed. Instead, they are even worse than corporates of traditional stock. Born out of an email culture, these new dinosaur’s really do behave like they don’t give a sh*t. I find it so frustrating that companies who come from a beginning as dynamic, exciting, written-in-a-garage type startups, end up worse than the traditional corporate dinosaurs they look to dominate or replace.

I pledge now, that when Rummble has 100 million users, we’ll have a human face; because to become loved by your users, truly loved as a lovemark and not just seen as a brand, you cannot fall into the corporate trap of bottom-line first and stuff the customer service. It might work today, but domineering empires don’t last forever, as we know all too well.

My opinion and emotions towards the eBay brand have hit rock bottom; and without some small miracle, that wont change now, for a very long time indeed.

More Mobile Network Operator Mobile Data Woes

I posted this on MomoLondon today in an effort to understand how any MNO can roll out such a policy and think it is sane / sensible / will work / will grow their data business …

Can anyone explain to me why if I access the internet on my Blackberry one way (e.g. via O2 actives homepage) I get charged, but if I access the mobile internet another way (via “Blackberry bookmarks”) I don’t ?!  Even if accessing the same webpage – e.g. Google.

Another stunning example of wholly opaque pricing and usability from the MNO’s.

Momo Crackberry users, be warned. I have received £600 bill for data for one month, because I apparently accessed data via the wrong APN.

Apparently one pay’s for data via “mobile.o2.co.uk” but not  via “blackberry.net”.  Frankly, this is a joke.  I dont remember being quoted “100 mb of inclusive data provided you use the correct APN”. 

To be fair to O2 they have reimbursed me for the confusion, provided going forward I make sure I use the right APN; and I commend them for making the right decision on this.

There may be a good technical or other reason, why it is like this; but that is not the point.  Don’t give inclusive (albeit capped) data packages, but then apply a string of confusing criteria to the deal. Data is data. Sorry, but I dont care about the complexities of an MNO’s network or international roaming deals.  I am paying for an inclusive, international 100mb of data for email and browsing. That is what I expect to receive, without strings attached.

Lift 08

I’m looking forward to Lift Conference 2008 next week, in Geneva. I’ve not been before so anyone who is going please say hello or feel free to fire any tips! It looks like it should be a good mix of people; something which I felt benefited DLD 08 in Munich so much. Tech conferences are great, but having a variety of views with people coming from different directions always mixes things up and gives better panels, better questions and better partying!

Understand Social Media and Web 2.0+

One easy way, is just to understand the breadth of services out there and how they fit together. Here is a fantastic, and fun, starting point by way of AI’s 2008 Web Trend Map laid out like the Tokyo metro. There is an interactive version here.

Why not print one out and stick it on the office wall – you might find your staff take some interest!

I think I preferred the layout of their 2007 map, but who cares what I think…

Twitter continues to be unreliable..

..one has to ask, what the hell are they doing? Was their software platform really SO flakey to start with that they are having this many problems ongoing? They only have a couple hundred thousand users – and I bet they are not all active(!). The answer is yes it probably was that flakey.

For those who don’t know, Twitter is a microblogging service – a way to distribute and share very short chunks of text (although some video micro-blogging services have started up) with friends or strangers. I question its important; many say its the unimportance of the service which is its strength. It has slowly gathered some pace after copious and unwavering support from the West Coast tech and social media community; in my opinion unreasonably so, and completely out of proportion to its accomplishments, especially alongside services like Jaiku (recently bought by Google) who started first and offered a far better service. They were hampered by being in Europe. We simply dont support our “own” as well as the Americans support their “own”; let alone champion them.

One of my biggest personal failings, I believe, is not to have grasped the concept of building something ugly, rough and ready (from the point of view of coding, back-end infrastructure, scalability and stability) in the Web 2.0+ world. Rummble was built properly from the beginning; I’m hoping one day soon I can sit smug in the knowledge that this turned out to be the wise tortoise-raced-the-hare decision. I guess we’ll see…