Limits of beta culture
A journalist asked me today, “is all software going to be done in the beta culture way?”
I replied that no - there need to be limits. A line that you should not cross.
Coincidentally, Ivan Kuznetsov had written decisively against such movement yesterday:
Beta culture seems to be spreading from internet startups to bank systems (although in this case I would be more inclined to say that this was a huge screw up on Danske Bank IT department’s part). I’m not sure I like it. Gmail beta, Flickr beta, but Sampo Bank beta? It’s definitely fun to participate in debugging of the new web service, but not when it deals with your real money.
I totally agree.
I would say that “beta culture” is irresponsible and unacceptable whenever:
- You charge the users directly, or
- The downside risk is more significant than the expected benefit of running a beta, like:
- Losing your money or something you have paid for (say, software for Mars rover spacecraft)
- Losing your precious data (say, software that screws up your lifetime photo collection)
- Losing your life / health (say, software for your pacemaker)
- Causing collateral damage to others (say, software for a nuclear power plant)
- Breaching ultra-sensitive private data
At least this is how I see it. What do you think yourself, what should be the limits of beta culture?
gmail has been in Beta for how many years now? What a load of tripe. Continuing to call it ‘beta’ software/application after all this time is a cop-out that lets Google refrain from taking any responsibility for the application.
Most reputable software houses, large or small, have some sort of private beta/alpha cycle, and often they will have a more public beta cycle too. There’s nothing wrong with that, and generally it is an excellent system for ironing out the wrinkles.
I agree with your list of software we should never see in beta, but I’d add that anything that is still in beta after a year or more is operating under a dishonest license to abstain from responsibility for the product and any failure it may demonstrate.
Some things, for instance Operating Systems are so huge that I can see why it can take years to develop them, but gmail??? flickr? These applications may be clever, but they are not monolithic applications requiring rocket scientists to build and maintain!!!
Comment by Steve — April 28, 2008 @ 10:31 pm
… Or death of Betalabs
It started out so good, lots of exiting applications and action on this website.
NOW - update of existing applications, very few comments. Important software like the transition of Lifeblog to Nokia photos went into production without a single word from betalabs.
I started checking this site 2-3 times per week. Now maybe every second week during my usual fortnight scrolling through favorites.
Get up to speed, beta-labs
Comment by Lars — April 29, 2008 @ 9:19 am
[…] on the Nokia Beta Labs blogs asks what can and can’t be classified as beta software. He concludes with a list of types software that probably aren’t suitable for releasing as […]
Pingback by Mobile Phone Development » Blog Archive » Beta Applications — April 29, 2008 @ 10:32 am
Steve: agreed, some kind of a time limit would be a sensible addition. It’s a bit difficult to carve that 1-year deadline in the stone, however, as it partly depends on the nature of the projects. For example, some research-driven initiatives might have no commercialization plans, but it might make sense to keep them available for years. Should we call them “betas” or something else?
Lars, thanks for the reminder. We need to get back to speed. Let’s see if we have something nice coming up…
Comment by Tommi Vilkamo — April 29, 2008 @ 12:18 pm
The term Beta has officially been ruined since the phrase Web 2.0 was first uttered. There is no beta software anymore, just software that works and software that doesn’t. Bugs will always be there, whether beta or not, collecting feedback is something you always should do, beta or not, there used to be a time when beta software was genuinely unstable and was not advertised to the masses.
Now companies advertise beta’s to everyone, that is where I see a problem. You offer beta’s to dedicated users, users who evangelize your products to get a sneak peak at new features, but you don’t offer beta’s to the entire world.
Comment by Stefan Constantinescu — April 29, 2008 @ 2:25 pm
[…] post by Tommi Vilkamo Posted in Uncategorized | Leave a […]
Pingback by Nitrio Mobile Blog » Blog Archive » Limits of beta culture — April 29, 2008 @ 3:06 pm
All beta software comes with an agreement that they do not take any responsibility for what the application does. I believe that as long as the user accepts this condition, there should be no limits. It’s the user’s fault if he/she prefers to use the application in a risky situation.
A responsible beta tester would know what is risky and what is not. Moreover, with new virtualization tools, setting up sandboxes is very easy.
If the user is taking risks with beta software, that user is either too amateur (using the software just for the sake of showing off) or too professional (exactly knows everything and is ready to overcome any problems).
Open source software development is rising and this requires beta testing to be open to everyone as well. In open source development, you provide the code from the beginning, so why not provide the application from the beginning?
Comment by Burak — April 30, 2008 @ 4:50 pm
The connection to open source software is interesting, because in that context the term “beta release” doesn’t seem to be used very much, if at all. I guess the problem is that the term “beta release” is so fuzzy that all it really says is that the product isn’t finished. Typically, open source developers are techies, so we want people to understand what we’re talking about. Terms like “release candidate” tell you the whole story, while “beta” could mean anything.
Comment by puterman — May 1, 2008 @ 6:18 pm