This is a general question but I’ll illustrate it with Eclipse. I recently reinstalled Eclipse and find that the distribs are called
eclipse-java-galileo-SR1-win32
and
eclipse-java-ganymede-SR2-win32
(I also have a “europa” from the past – what happens when we run out of Jupiter’s moons?)
I find this very confusing as there is no indication of which version is the latest and in fact I muddled them.
This is not restricted to Eclipse, and several version of software come out with version names (e.g. the Mozilla family). Personally I would much prefer the normal decimal version numbering. What other examples are there of name confusion and is there any justification for it?
update some early replies suggest some people prefer names to numbers and vice versa. Could we not have both, therefore?
update A majority view (but not consensus) seems to be emerging that names are useful for developers before release but that n umbers are better after release
Names have long been used as code names during development, so that developers could refer to a name rather than a number all the time (“version 5.67 branch 2” doesn’t roll off the tongue as easily as “bob”).
But from the point of view of the end user, they suck. They only convey information to someone if they know the list of version names used. A user doesn’t care that they are about to download “rancid cheesecake” to replace “fetid sardines”. What they want to know is that they have 2.1 installed, but there is now a 3.0 available.
Year numbers sit half way between the two. They’re numbers that are often completely made up, making them simply names – We’ve been using 3DSMax 2010 for months now. In another month the name might even coincide with the calendar! To confuse us even more, we have Visual Studio 2005 which is version 8, and Visual Studio 2008 which is version 9.
Argh!