Software Design

Martin Owens doctormo at gmail.com
Sat Dec 26 19:09:29 UTC 2009


On Sat, 2009-12-26 at 19:57 +0100, Markus Hitter wrote:

> To get back to the initial topic, instead of paying people to write  
> tools for taming complexity, paying people for getting rid of unused  
> stuff would be even better.

Culling is a valid job, it's just not valid to remove functionality;
move it, transform it, make it an irrelevance and remove duplicates. But
removal of the primary instance functionality is giving up on design
IMO.

> Image viewers: Instead of improving F-Spot to be the best thought  
> reference, half a dozen additional image viewers pop up, each of them  
> obviously created from scratch.

I'd pay to get an image viewer that works TBH, F-Spot's got bad deps on
my machine, and in other regards some of the functionality in F-Spot
should never have been put in there, but into a further, deeper layer
that made the functionality more far reaching and less complex, less
repetitive.

But you try asking people to write infrastructure software, you might as
well be talking to brick walls. f-Spot works as it does because it's not
been designed, much like most software. It's either a clone or need de
jour, not exactly room for design in those modes.

Martin,





More information about the Ubuntu-devel-discuss mailing list