Dec 052008
 

I recently encountered a dead blog entitled “Linux Haters” and instantly thought up writing about tedious fan-boys that think that the operating system they like is the best and everyone should use it. I’ve no time for people like that as they tend to annoy rather than educate. I’ve no problem with people who prefer to use Windows, Linux, Solaris or OSX; it is their choice. Of course in the case of Windows, I do have to wonder why 🙂

But one of the links on that blog led to a place that (amongst other things) ranted about how FOSS projects always have dumb names, and that these projects need a big dose of marketing intelligence. He went on to whinge about the word-games often embedded into the project name.

First of all, he misunderstands how many open source projects start – with a geek or a group of geeks deciding they want something different. Either a new package or a variation on an existing one. There are no marketing types in sight, and the geeks involved probably have no great expectation that they are coming up with the next big thing – they are just having fun and hoping to come up with something useful for themselves. So what if they have a bit of fun playing word games to come up with a name for their project ? Not only do many such projects end up disappearing without a trace, but as marketing types have fun playing with words, why can’t geeks ?

Perhaps the names they come up with are not as punchy as a name thought up by a marketing department, but weirdness does have its own value in this area. A name such as Amarok does tend to stick in the mind more than Music Player 52. And over time, formally weird names such as google and yahoo do tend to become more normal if they are attached to popular projects.

Secondly he specifically criticises names invented by geeks for being recursive acronyms … but does that matter ? He specifically names GIMP which is admittedly particularly guilty being a recursive acronym with no termination. But most users won’t care … once they learn that GIMP does images (and most distributions will tell you so in the menu), they are not going to care that the name is an infinitely recursive acronym … they will just get on and use it.

Thirdly he overlooks the fact that some of the names may in fact have “sensible” names but are in fact sensible names in non-English languages.

Finally he tails off into a moderately incoherant rant with more insults than proper facts.

Perhaps “funny” names do put people off, but perhaps not. Most people are in fact more concerned with compatibility (they use Word because everyone else does) or features.

And of course there are more than a few commercial software packages whose name is not entirely sensible … does Photoshop have anything to do with setting up a shop to sell photos? What does Trent do ? Or Cedar ?

Dec 032008
 

At work I have the pleasure (if that is the correct word) of dealing with a number of commercial software packages which are a little more expensive than the run of the mill packages such as Word that most encounter. In fact I am probably more familiar with software costing more than £10,000 than those less than that cost because I have the curious habit of opting for free software.

I do not have any moral objection to commercial software – if someone wants to pay money for it and pay me to support it, then that’s fine by me. I am just somewhat reluctant to spend my own money on software (although I have done in some circumstances).

However I have come to the conclusion that commercial software just isn’t very good. To give some examples of poor behaviour (without mentioning any names, because it wouldn’t be fair to their competitors) :-

  • A Java application server that does the equivalent of Listen ipaddress-of-server rather than Listen 0.0.0.0 making it difficult to move the server to another server (such as cloning Solaris zones for recovery purposes). This is the kind of kindergarten mistake that any decent developer knows not to make … even I know, and I’m no developer.
  • A software package whose configuration script will work fine for 1-4 members of a cluster but breaks that cluster when you add an additional member of the cluster (taking it to 5). Not only did the configuration script break, but this was known to the vendor.
  • Packages that take months to install even with the support of the vendor.
  • Vendor supplied consultants who apparent have never encountered a keyboard. Or have trouble with basic Unix skills when they are to support a package on a Unix server.
  • Installation or configuration scripts that simply don’t work and have to be effectively re-written by ourselves.
  • Patch bundles that have obviously never been installed on the product as they completely break the service when installed. We have one vendor who consistent wants to break things and have to go through their patch bundles with a fine tooth comb to debug their scripts.
  • “Support” that takes months to respond to a query. Or only responds when hassled through an account manager.
  • “Support” that takes weeks to accept that you are entitled to support because the product whose serial number you have a photo of “hasn’t left the warehouse”.

And people pay for the privilege of this ?

The classic argument in favour of open-source software is that you have access to the source code to apply any fixes that need making. The counter to that is that places do not have the skills to be able to create any fixes, but from what I have seen, in many cases the customers have more skills than the vendors!