December 18, 2011

Why Do Professional Programmers Use Macs?

Someone just asked this question on Quora. My answer ballooned somewhat :

1) Historically, Macs were the preferred machines of desktop publishing and graphic designers, then web designers. As the web became an increasingly important platform (relative to "client-server" on Windows et al) individuals and companies who had started in web-design area became more prominent - think of the rise of 37 Signals and Ruby on Rails - and brought their Mac-ness with them. 
2) When Apple shifted to OSX they made it a real Unix. In the late 90s, one of the attractions of Linux was that it was the only way to get your hands on a proper command line and Unix tools, and to run the server-side software like databases that you needed. When MacOS became Unix, the Mac could do all that too.  
3) At the same time, Microsoft basically fell over. Believing that their birthright was to control every computer platform ever, rather than that their job was to make good tools, they spent the noughties trying to copy first Java (.NET), then Google (Bing), then Apple's iPod and iTunes (Zune), then Flash (Silverlight), then Sony's Playstation (XBox) etc. etc. The result was 5 years wasted on the appalling Vista, and a lacklustre successor Windows 7 (whose main virtue is that isn't quite as bad as Vista). (M$ clearly haven't learned the lesson, so it seems that Windows 8 will just be an inept attempt to copy the iPad while leverageing the rapidly evaporating "lock-in" they think they have in the desktop OS market.) 
4) Worse, the commodity PC market that Microsoft (and Linux) rely upon went through some rapid consolidation and price cutting. By my reckoning we expect to pay about a third of the price today for a PC compared to our expectations of the early - mid 1990s. But this didn't just happen in the nice "Moore's Law" sense. Commodity PCs got cheaper and nastier too. Sure they have faster processors, but the cheap bits often don't work together all that well. 
5) Despite Linux's maturity, the PC manufacturers have totally failed to get behind it.
Personally, I'm writing this in Chromium under Ubuntu on a beautiful Asus Bamboo laptop. And I'll resist the cult of Apple for as long as humanly possible. But the trend is obvious. Even in 2011, PC manufacturers refuse to support Linux (they won't sell a computer with Linux pre-installed, they won't help to make Linux run well on their machines and ensure that drivers are available for graphics cards etc.) 
Asus added a whole bunch of power management software for the pre-installed Windows 7 on this machine when I bought it. They offer no equivalent for Linux, so my machine runs unnecessarily hot (I have shorter battery-life and probably the machine will die sooner.)  
The combined result of the Microsoft debacle, changes to the PC industry and the refusal of PC manufacturers to support Linux is that Apple is the only company which now seems competent enough to make a decent personal computer that you can actually use for software development.  
Seriously! Think about going out and buying a computer and you think either it will be a substandard Windows 7 machine (packed with slow, buggy "extras" that the manufacturer was bribed to put there, and without the command-line tools that all professionaldevelopers need and use) or you contemplate getting the same PC and having to install Linux on it yourself and, if it's new, having to deal with driver compatibility issues etc. etc. etc. 
Or you go out and pay twice the price but get a machine which is of high build quality, you can trust will do everything you need out of the box, and where the hardware / operating system just work together.
6) Oh, and one more thing. You can't develop for iPhones and iPads on a PC or Linux machine.

September 20, 2011

Metro and the Post-Windows PC

ZDNet has a good article suggesting that Windows 8's Metro stack is going to replace the win32 and .NET stacks of legacy PCs.

It sounds plausible to me. But what should worry everyone at M$ is that it's only the legacy applications which are keeping Microsoft in its dominant position in the enterprise. Metro may be a fine new UI / operating system stack, but without legacy support it's as precarious as WebOS or Meego. Or rather, the only difference between Metro and WebOS or Meego is the inherent conservatism of M$-loyalists in the IT department.

I presume M$ will try to handle legacy apps. with emulation. Which may work on very fast new machines, but I'll be interested to see if this works out on tablets and other low-power / long-battery devices.

July 01, 2011

What happens when, suddenly, all these Chinese internet playas start going international?

Everyone is focussed on Google and Facebook and Twitter etc. But Asia is full of equivalents. At some point, one of them will achieve some mainstream success in the West. And then what?
Another thought on Google+ while I'm waiting for someone to send me an invite (hint).

People are saying Circles are good. Maybe they are, but if they are, how hard will be it be for Facebook to copy them?

Until now, FB have been pretty good at the "steal good ideas from elsewhere" game. Witness buying FriendFeed and reorganising FB to be more like Twitter. If people find that Circles has a better UI than FB for organising your friends into groups, you can bet FB will soon adopt a similar UI. I'm sure they can do that faster than Google can build up FB's userbase.

June 30, 2011

Winer on Google+.

That's pretty much all you need to know.

Update :

You know what would have been funny? If Google had made invites computationally expensive like BitCoins. That would have shown the world that they were down with the zeitgeist, still l33t masters of algorithms, and still had a sense of humour. Which is what we want from Google. It's their brand.

Instead, this is a public admission that they're scared of Facebook, a reuse of 5 year old tactics, and apparently it's internally driven by an ex-Microsoft middle-manager.

They're just building a bonfire of their credibility here. And preparing to throw all their great products (Gmail, search, YouTube) onto it.

June 17, 2011

May 17, 2011

Bill Gates was "behind" Microsoft Skype deal.

Here's a thought. What would it take for Microsoft to pull Gates back from retirement to take over running the company again?

May 10, 2011

Finally Microsoft does something exciting. Buys Skype.

You know, I really thought it was going to be Facebook that bought Skype. Perhaps they couldn't afford it.

Anyway, this is really the first good big move Microsoft have made on the internet since buying Hotmail. Cringley thinks it's purely defensive. And it might be, but he makes a good case that even that's a good (or necessary) idea.

But it could be so much more. If M$ don't fuck it up.

Here are a couple of observations :

1) Skype is a great brand.

I always thought that M$ had a good brand in Hotmail, but they proceeded to throw it away, continually trying to turn it into MSN / Windows / Live blah whatever. People still call it Hotmail. They still use that in the address, but M$ did everything they could to confuse and destroy the "Hot" brand.

It will be ULTRA idiotic of them to try to rename Skype as LivePhone or MSN Talk or something. I mean, really, really, really, really dumb.

Contrariwise, Skype is much better brand than anything else M$ has when it comes to cool contemporary internet stuff. Other technologies that M$ are developing could well be moved under the Skype name. For example ...

2) Skype is a social network.

It really is.

Like Facebook, Twitter and LinkedIn. And unlike either Google, Apple or any of Microsoft's previous efforts, Skype is a pretty meaningful "social graph" which can be used for all kinds of interesting experiments in social communication.

At the moment, Skype is very much focussed on synchronous chat / phone call. But it would not be hard at all to add asynchronous capabilities to the client. Some kind of pub-sub, status, wall. Allow Skype users to tag their contacts, or group them into themed lists. And then to watch the posts from a particular list. Let them add photos, links, video. I think within 6 - 8 months M$ could build a fairly plausible and compelling rival to Twitter. Especially if they allowed groups to create private workspaces and channels.

In fact, if I ran Microsoft (here it comes ...) here's exactly what I'd do. Find two or three great programmers and UX designers who are hungry to do something new. Pay Dave Winer to go and talk to them about instant outlining. Pay someone from Google's Wave project to go and talk to them about what they hoped for from it, and what went wrong. Get the designers to mock up some forward looking ideas about how a future Skype client could incorporate asynchronous communication, "narrating your work", private tweet streams, etc.

3) Skype is collaborative work

"Skype" is what people in business say when they mean "conference call".

And Skype could be another chance for M$ to get into collaborative work. Word and Excel need to support shared editing of documents. And it needs to be easy to understand. So bundle the Skype client into Office. (Not exclusively, of course). And have a menu option on Word and Excel saying "Share this document via Skype" which immediately allows you to invite skype contacts to work on a document together.

What if they don't have Office? Well, the Skype client should at least have the free document viewer built into it so that they can follow what you're doing. (I'd go further, why not allow some restricted editing facilities? And yes, this should run everywhere the Skype client runs, ie. Mac, Linux, iOS etc.)

More importantly, hello? App Stores! Have a one click "buy and install Office" built into the Windows Skype client. Make it all work smoothly.

4) Skype is a subscription service

On the subject of one-click buying, remember that Skype is a paid relationship / service. (And likely they already have the user's credit-card number.)

Apple had one of those with iTunes, and look how that worked out for them. Amazon has one, and it's managed to take the Amazon account from selling books to selling virtual servers on AWS. And it's why Amazon are a serious contender to rival Google's App Store for Android. Being able to take people's money easily is an amazingly valuable asset that none of the other social networks (Facebook, Twitter, LinkedIn) have. Even Google are struggling with this problem.

Get creative here!

5) Skype and Windows Phone

Yes, build Skype into Windows Phone. But I'd go further.

I'd immediately offer a discount on Windows Phone contracts to anyone who's put money into a SkypeOut account. It's a way of paying people to use WP7 that a) might actually encourage some undecideds but b) importantly, doesn't look (too) desperate - it is, after all, a reward for buying into the whole M$ ecosystem. Go further, a single plan for renting a Windows Phone AND SkypeOut calls.

6) More brand extension

SkypePad : it just sounds a hell of a lot funkier than Windows 8 Tablet Edition doesn't it?

Skype 360 : better than RoundTable? (Don't even start me on "Unified Communications"!)

You get the idea ...

April 14, 2011

Not updating this blog much. I've taken some time to go back to college, do art and play with a bunch of toys that I've wanted to try for a while. Frankly, my only real thinking about "Platform Wars" happens when I
read ZDNet ... so you might as well get the stories direct from them.

I did wonder whether I should retire this blog entirely. But I know that although I sometimes go on hiatus on some topics, I usually find myself getting interested in them again at some point. So consider this blog hibernating (rather than dead).

However, here's a comment I posted to a mailing list discussion a couple of weeks ago about the opportunities for LinkedIn. Nothing new to regular readers here. But I realise it deserves a permanent home here.




For years I've been asserting that the key for social networks to make
money is to enable their members to form groups that actually DO
things together. And to earn revenue as enablers of doing things.

In a sense, Groupon is a great step forward although a) it doesn't
come out of an existing social network, and b) it seems like Groupon
finds the deals itself rather than lets members find them.

If I ran a network like Facebook I'd add "one click" Kickstarter-like
functionality so that organisers of any event could instantly set up
an account to collect donations, or an escrow account to collect money
for things that only happen if a threshold is reached. In the long
run, Facebook's 500 million members are going to be more effective at
finding Groupon-like deals than Groupon's paid staff.

In the case of LinkedIn, it's slightly trickier, because the essence
of LinkedIn is to NOT be Facebook. By which I mean, LinkedIn's
strength is that it works as as low-traffic, low frequency site for
"serious" uses (as opposed to the fripperies of Facebook). You can't
expect LinkedIn users to be on the site every day (or even every
week).

Here's a personal perspective. I'm an artist / musician and software
developer. I've been paid to write music for modern dance productions;
and I've been paid to write stock-control software to manage
warehouses. And, of course, depending on what kind of gig I'm applying
for, I may want to send a very different CV of my recent projects and
achievements. I basically keep two Word files and manually keep them
up-to-date and in sync. Now what I'd love LinkedIn to do for me would
be to let me create richer CVs with detailed *tagged* sections and let
me generate printable docs based on a selection of tags. Eg. if I'm
applying for an exhibition I want to include five recent bits of
web-art I've done and links to a YouTube channel. But won't bother
mentioning the e-commerce system I built in Perl all those years ago.
For a different potential employer, I may like to generate the
opposite.

I probably wouldn't pay directly for that functionality alone. But I
would spend more time on the site if it was a full CV management
solution. And I'd accept targetted advertising. (LinkedIn would, of
course, have a lot more data about me too, that it could use for
targetting.)

Similarly, people aren't changing jobs every day, but freelancers
*are* pitching for jobs every month. I'm not saying that LinkedIn want
to buy oDesk or should copy it directly. But I am saying that the
growth opportunities for a "professional oriented" social utility will
include a lot more support for people doing short term contracts. And
yes, that means helping people manage outsourcing to freelancers
around the world, letting people manage more sophisticated CVs and
portfolios, keeping a reputation system, facilitating international
payments etc. etc.

LinkedIn still thinks we do a series of "one job after another" rather
than have a portfolio of different jobs, freelance contracts, non-paid
activites etc. which overlap in time. I went back to college this
year, and yet I can't tell LinkedIn that I'm in full time education
again and therefore not to show my most recent paid job as if it's my
current one. The biggest improvement they could make is to keep the
seriousness (ie. don't try to copy Facebook) but to understand that
work is evolving and to help members track and navigate their
trajectory through this increasingly complex space.

February 25, 2011

February 02, 2011

Dare Obasanjo has a good post up, analysing the failure of several formats and protocols. Useful thoughts.

January 19, 2011

Microsoft release OneNote free (as-in-beer) on iPhone in the US.

Slowly crawling back towards being a software company? (As opposed to disappointed wannabe heir-apparent to all classes of computer company)

January 13, 2011

Google Wave is dead. But EtherPad (bought and shut-down by Google, but then released as open-source and sponsored as a non-profit by Google) thrives.

What's the lesson here?

January 06, 2011

ZDNet : 2011 is the year of the enterprise iPad.

To repeat what I've been saying for a while, Microsoft are utterly fucked if they don't get Excel on the iPad now! They will lose their best and most valuable business brand. And more importantly they'll lose their de facto ownership of the majority of the world's "semi-structured" data.

And once this is lost, the rest of the Office house of cards will come tumbling down too - I "need" Excel because all my business data is in XLS files; but once my business data isn't in XLS files (once its in iPad todo-lists and executive dashboards) then I won't need Excel; and if I don't need Excel, what else in Office do I really care about?
ZDNet predicts a massacre of traditional Mac developers when the app-store comes online.

Update : this is good too :

Thinking about this, I’m predicting my own behaviour with the Mac App Store. I’ll probably start trying out all sorts of free and low-cost “apps” if they look like they can provide me with instant gratification. (Especially if I can use an external hard drive to store them.) And I’ll probably buy a few “apps” that I can justify, in terms of effort and cost. But I might give up quickly on these if my initial experience isn’t optimal (if the apps in question aren’t worth the cost or effort). And I’ll try different things associated with these apps I do enjoy.
GigaOm : Has Google given-up on social?

January 05, 2011

Amazon cleverly create their own Android AppStore.

This is a smart and interesting move. There's no reason you need to own the operating system to own the appstore. And Amazon as innovative online retailer with size and technical has a good profile to do this.

There will, of course, be a tension between Amazon the curator / publisher and Amazon the network-market enabler. Will they censor content they don't like? Would they exclude a wikileaks app?

It also raises the fascinating question (at least to me), if Android, why not Windows? Is any company bold enough or foolhardy to build an app-store and take the role of curator / gatekeeper of the Windows eco-system rather than wait for Microsoft to lumber up and do it badly in about three or four years?

Some people have kind of been in this game for a while : Tucows, Download etc.

Imagine what some visionary leadership and chutzpah could do with one of these companies at this moment.