Tuesday, September 4, 2012

Fit for Purpose

Just before I get started on a post, there's a situation developing around software patents in New Zealand. For more information, and a petition around it, go here. It's a worthwhile cause.

I've been meaning to write this particular post for a while now. I've had a few other things on my plate - the future of "The Image" i.e. how can I get it resourced (I'll talk about this another time)? My place in the Manaiakalani project. My identity outside of that project. My identity inside it for that matter.

Something that always gets to me is how people see the whole "fit for purpose argument". Take Apple's stance on the iPad. The iPad is sold as a personal device BUT is marketed to the education sector as something desirable for schools. But is it fit for that purpose?

I am of the strong belief that it's not. Why? Not for any technical reason but because of legal ones. The user agreement for AppStore and the like is that each device will be associated to the user's account. What if:
  • The user isn't old enough to have an account?
  • The device isn't owned by an individual but rather, by an entity?
The management becomes troublesome. Not so bad for cellphones (at the moment). Terrible for computing devices. So you buy an App (short for application) off AppStore (short for Apple *moan*) and it syncs with all of your devices? (I'm not sure this is true). But what if you need 3, 4, 100 distinct licenses of a piece of software? The AppStore, in it's current form, doesn't allow you to buy multiple licenses for the one account. You would need 3, 4, 100 distinct accounts to do so. To not do so would be a violation of their terms of usage.

So is it fit for purpose if using it for a particular purpose seeks to criminalise you even if that purpose is being pushed by that particular company?

While we're at it, Android also isn't that great a fit. It too is a personal device but not marketed in the same way as it's Apple counterpart. Android really does lack the management tools to make it easy to use on a mass scale and we still talk about exploits and rooting...

I got told about the savings on Microsoft's 365 over Office recently. I have my suspicions that these savings won't actually be savings as you're talking about a completely different way of licensing - previously you licensed per machine. Now you'll be licensing per user. This might not sound vastly different BUT what happens when you have staff turn over? Do you then have the admin costs around deactivating an account and activating another one? This sounds a whole lot more complicated to me than throwing a computer with Office already installed at the new employee. But more importantly, does MS 365 have the functionality that users are already using?

In the deep dark recesses of Office there's VBA. And scaringly, it's used quite extensively. Various Windows business orientated applications plug in to Office using VBA. I heard that the financial market use it quite a lot (it's not a great platform - it works .... mostly .... but can be temperamental). This puts me in mind of South Korea's reliance of ActiveX. MS dropped the support for ActiveX after Explorer 6. That article linked to there does not paint a grim enough picture. Not only is the nation stuck with Internet Explorer - they're also stuck with a horribly outdated web browser. And this from a nation whose on top of Internet connectivity and speed.

Is MS 365 fit for purpose? For a lot of people, yes. For a lot of people, they'll find that things they took for granted just don't work quite the same. Things like Cognos. I came across a whole skew of them while doing some work for New Zealand's biggest company. I'm sure there are companies using a bunch of unsupported applications that make use VBA features. Why would they still use unsupported software? Because often a viable alternative simply doesn't exist yet.

The question for me as an I.T. person, do I criminalise myself or knowingly sing the praises of something knowing that it might not be a good fit (often trickier when no good alternative exists) in order to try and shoehorn in a solution for a client/user/company/customer? And while these questions of ethics are about, is it fair that the expectation by those clients/users/companies/customers is that these solutions are fit for purpose without understanding the underlying problems?

Flash is a perfect example of this. Macromedia (and now, to a lesser degree, Adobe) made their money by certifying a device as fit to run their software. It can't legally be distributed (as part of an image or in a package for example). It can be downloaded off their site. This has caused me all sorts of headaches. The Ubuntu way of doing it is to include a package that then, during it's installation, downloads and installs the program from Adobe. If that installation is interrupted for whatever reason, the package manager doesn't usually know and so you end up with the packaging system being broken or a dysfunctional "installation" of Flash. Packages should be self containing and not need to download ANYTHING during installation - or risk breaking the packaging system "when things go wrong" (That's so going to be a TV show).

Knowing that flash is absolutely essential to A LOT of great online educational applications, what is the ethical road? Telling the user that they'll have to go through the god awful experience of downloading and installing the plugin themselves even though there's no technical reason why it couldn't be distributed to their machines? Or breaking the license agreement and criminalising yourself? Perhaps manageable when you're talking about ... less than 20 machines. Not so much if you're talking about thousands.

The technical reasons for Free software often hold a lesser place for me than the philosophical - but they're there and very much real.

No comments:

Post a Comment