2010-04-12

Software on speed

Yesterday I tested Ubuntu 10.04 Lucid Lynx beta 2. Within half less than an hour testing I found 4 bugs that I do not consider to be minor (and I filed them accordingly) - not talking about the design change putting window buttons to the left (which is considered as a bug by the community). All those 5 points I do consider making a bad impression on new users considering a switch over to Linux. By writing this they have only 17 days left to fix. This is a very few time left I would say.

The twice-a-year release schedule has been discussed already lately at a number of places and I would also say that such a schedule is too short for comprehensive tests and high quality software development.

Time-to-market is an important parameter in business and especially in software business where there is such a huge amount vendors and sites. Other than the supermarket around the corner with limited reach, software can easily distributed around the world - nowadays (it has not always been that way, think to times before internet has been available to the vast majority of people). So from the competition point of view software is a hard field.

As it is widely accepted that software has bugs, vendors do not worry much about the bugs and do not give much on quality - that is my impression.

Although none of the bugs I found can be considered as a real show-stopper, it could be so annoying for people that they don't use it (or stop using it). I have experienced a lot of bugs and bad implemented software that alltogether made a lot of software unusable for me considering the software usage from the economical point of view and thinking of optimizing productivity.

Software is there to improve productivity and make things easier - it is not just for geeks playing around.

That said, Linux (and Ubuntu in this case) is driven a lot by the community and it is given away for free, so one should not complain. But this post applies to a lot of other software products too.

Anyway, quality should be an important priority in software development and from that point of view I understand the debian guys, where an eternity can pass until a new version is released. I one thing why I don't use a debian stable release is because they are so far behind in features or new software releases respectively.

Maybe the optimum is somewhere in the middle...

Related posts: New Year's IT resolutions, Going Linux.