Whereis Mobile: Not quite the Diet GPS

Platform 18 and 19 at Central station.Image via Wikipedia

Being based in Sydney for a two day stop over gave me another opportunity to test drive the Whereis Mobile capacity on my Samsung handset.

It was hysterically funny, but that’s because I laugh in the face of Fail Whales.

For a starting point, when I fired up the device just outside of Central Station, it cheerfully reported that I was standing in OXford Street.  Not quite.  So I tried to input where I was standing (Elizabeth Street), only to be told that no such location exists (404: NEO NOT FOUND).  Then I tried querying Goulburn Street (my destination) and that didn’t exist either. Okay, so the Matrix was having a bad day.

Then I asked for the name of the Hotel (The Vibe Hotel) only to be repeatedly told that the connection to the server was lost.  A procession of the same black cat walked across the doorframe for a minute or two. I gave up after the phone showed signs of despair and the hysterical laughter attracted way too much attention.

Later, for the fun of it, I asked the SENSIS search about the local Nandos.  To be fair, I was standing about a building away from one, and didn’t realise it was where it was (I was in QVB or QVG. There’s one in the food court connected to the QVA2Z).  Nearest Nandos that was suggested was 8km away. So I asked for directions (walking).

The connection was reset by the server.

The Matrix rebooted and dumped me out of the network somewhere around Darling Harbour, and I realised the time had come to stop mocking Telstra’s Mobile Whereis, and make use of the 3G modem on the laptop back at the hotel room to get real answers from the intertubes.  Amusingly, it took less time to walk back to the hotel, google find the information and walk back to the stores than I’d spent trying to beat useful data out of the Sensis systems.

I know Sensis Whereis is not a replacement for GPS equipped phones, and I know that it draws location data from the mobile phone tower etc. I’m conversant with limits to the crude tech.  I’m just hysterically amused that I was thrown off the network four times in the process of asking “Where is…?”.

Once, admittedly, was for asking where the hotel was when I was in the lobby.  That one I probably deserved though.

Reblog this post [with Zemanta]

Disengagement in the New Media

Foyer of the w:en:Australian Parliament

Image via Wikipedia

I spent the morning at a seminar on politics and technology where noticable by the absence was the technology, and present by the ministerial load was the politics.

Three observations of the session

1. Senator Joe Hockey plays a character in Parliament.  Given he registered his own namespace domain name in 1996, he’s got geek blood.  He showed his colours for a few seconds, remembered that journalists and political types were present, and reverted to his party endorsed yobbo bloke persona.

2. Senator Kate Lundy is awesome.  She had five minutes and said more useful stuff in that time than most people did for half hour speeches.  Still, if you have someone who has to leave 20 minutes after the start of the session – give them the first question, not the second one.  Poor form moderator, poor form.

3. The event was so horribly managed it was frustrating. I walked out of the last session frustrated at the inept management – if you’re going to do a panel of Web2.0 and political campaigns, and open the floor to discussion, you might want to let the audience engage in the debate.  Moreover, if this was a showcase event of politics and technology, bring the wireless networks and have the technology.

This panel concept was an old media dinosaur town hall smashed into 55 minute windows with brief intermissions for commercials.  Two five person panels sitting around talking about politics and technology should have been a full day session.  Hell, the first guest could have used 90 minutes easily.

All up, I’m probably not going to attend future sessions of this nature if I have deadlines, due dates or a better offer from a SPSS analysis output (unless Senator Lundy is speaking for longer on technology. Then I’m there)

Zemanta Pixie

What can save Twitter?: Downtime, User shedding and paid accounts

Twitter 6x6

Image by Steve Woolf via Flickr

Another “Twitter is doomed” post because the system dropped below some magic marker line in the last week and probably has an uptime of 95%. (Royal pingdom scores it at 96%) Clearly it’s all over at 95% uptime.

Does anybody who writes about Twitter play World of Warcraft? They take the servers down deliberately for several hours a week. Rolling reboots just happen with a 15 minute warning. Periodically you just can’t get into your server. Clearly, World of Warcraft’s success is killing it. Hence, World of Warcraft is over people. Move to a new game.

In my professional marketing opinion, there are three things Twitter could do right now to improve server performance – charge for follows, charge for API, and demarket the service. None of these suggestions mean the service is dead. It’s well alive, and here’s some tune up exercises to make it a little more robust.

1. Pricing

All accounts can post unlimited tweets for free. Twitter needs content to create value for the user base (hello Livejournal, remember that? Free accounts generate the content that make the paid accounts valuable).

Free and subscriber accounts get a base of 200 follows. Users can purchase an additional 200 Follow credits for a flat fee of $5/month for 200, or $50/200 for 12 month package. This could dramatically cut the auto-follow spam accounts out of the system (or compensate twitter for their drain on service). A power user like Robert Scoble would be up for $105 a month. If Scoble is getting $105 or more value from following 21,145, then it’s a cheap deal. If not, then it’s good time for Robert to cutdown on the number of people he follows.

Business rationale: Based on a remark from the series of posts on scaling microblogs series at Hueniverse,

Going through a timeline request, the server first looks up the list of people the user is following. Then for each person, checks if their timeline is public or private, and if private, if the requesting user has the rights to view it. If the user has rights, the last few messages are retrieved, and the same is repeated for each person being followed.

Lower the number of follows per user to a level that the user self moderates as valuable, and you’ll lower the number of calls to the timeline. Given free accounts can have at least 45000+ follows (that’s the highest score on a spambot account I’ve personally), that’s a lot of database work. Charge for the use of follows and the abuse of the system should decline rapidly.

2. API Access

Separate desktop client API accesss with a registered protocol for the clients (since they’re an asset in spreading the Twitter experience) from the services that use the Twitter API (eg Harvest or even Remember the Milk‘s twitter system). In short, if your device helps people access Twitter, you get the free API. If your device piggybacks on Twitter, you can pay your way.

Set basic API developer access at Free, with enough features to be useful for encouraging open source developers to give it a look. But at the same time, free should buy you what you pay for – if you want mission critical applications to use the Twitter infrastructure, you should be investing in the system that supports you. Paying for access to tiered levels of the API is a business model option that is effectively a licence arrangement between two businesses. If your business model relies on a free account service never charging you for the use of their systems, change the model.

Business rationale: Free API to get development started, and cheap rent/low risk investment levels will continue to encourage the use of the API to make new products, services and keep development happening outside of the big chequebook businesses. Pricing also means that the businesses draining the service with API calls and server loads are chipping in cover their costs, and a “cost per API call” billing structure would bring about dramatic rewards for API efficiency at the programmer/client side.

3. Demarketing

Back in 1971, Kotler and Levy put forward an ideas that marketing was fully capable of reducing the number of users of a service in order to increase business performance through demarketing. Twitter is having scaling problems due to being too popular, too soon, and the system’s not ready for it. So, Twitter should look to alternate ways and means of dealing with the demand through shutting down periodically. Make Monday “TweetFree Day” (and set the Monday to run from the dawn on New Zealand’s Monday if they want 30 hours down time). Shut the service off for a full week – with the exception of the delete account button. If you can’t go seven twitter free days, you can delete your account and start up at FriendsFeed (who are going to have to suffer a Twitteresque server melt soon)

Business rationale: Deal with the scaling issues by scaling down Twitter use and users. Decrease the followers per account at the extreme end of the bell curve. Twitter should shed heavy load people onto other services. Sack clients. Block APIs if the APIs are killing the system.Twitter is a free service, with an open API, and not supported by subscription, sponsorship or advertising and it’s been insanely useful. Given we’re not paying for it through any visible means right now, maybe we could expect a freeware web service to have a reliability of freeware.

2008 Federal Budget Speech

Second analysis of the budget using tag clouds, after key phrases have been coded so that “1.8 per cent per year per child per family” style phrases didn’t make ‘per’ the most common word.


created at TagCrowd.com

The main story out of round 2 of the analysis is simply that government talks about itself a lot, and for a proclaimed long term focused budget, there’s a whole array of four to five year statements in the mix of policy and promise implementation. It’s amusing to see a Labor Party budget say “Five year” quite so frequently – Marxian or Freudian slip?

Team Fortress II: Where my spare time went…

getaround1Image by Bryan Sutter via Flickr

Sunday Night, I found myself loitering around the TF servers, and in a match where on the Red Team were between four and five members of the (GCS) clan, who are very good at what they do in TF2.  Sufficiently good at it to completely school my side for a few rounds.

Then came the last round of the night, which consists of four levels.  We were comprehensively monstered over the first levels, and really beaten back in the last level with the Blue team cutting through our line in what felt like record time.  At seven minutes and forty seconds, the cart was within a heart beat of Blue winning.

At 4.03, I had enough presence of mind to snap a screenshot as I came off a respawn.  The cart had been in that same spot for several minutes.

It\'s a game of inches...

In fact, at the 7 minute 40 mark, I thought we were gone.  For seven minutes, 30 seconds, the cart would go no further than about four cart lengths from the end, and get within that same nail biting proximity until Blue swamped the point with 10 seconds left.

For seven minutes, I was playing one of the most intense video sporting matches of my amateur career, against a squad of good (and possibly even pro) players.  Not only that, but playing such a hair raising defensive game that my usual level of video game cursing was way off the charts for emotion, intensity, and…well… me completely losing the plot because I screwed up a defensive move and thought I’d cost the game (it didn’t…but wow, the intensity).

What was especially sweet was playing a glorious defence game… as an offense orientated soldier.  Not my usual support role as medic, or defensive sweeper in the Pyro, but a straight out aggressive Soldier spot.  Not bad, not bad at all.