L'uso di questo sito
autorizza anche l'uso dei cookie
necessari al suo funzionamento.
(Altre informazioni)

Friday, August 8, 2008

Il paese della regina rossa

Tanto, dubito che Slashdot me lo pubblicherà (non lo fanno mai). Pr chi non lo sapesse o non lo ricordi, nel Paese della Regina Rossa (da "Attraverso lo specchio e ciò che Alice vi trovò") bisogna correre più che si può per restare nello stesso posto. Così è in informatica....

So, I started to evaluate the latest XenSource (Xen 4.10) after having spent quite a bit of time with the free Xen version (3.2.1). Most everything has changed. Gone has 'xm' (replaced, but not quite, by 'xe') - meaning all my xm based tools will be broken. The config system has undergone a revolution - more tool breakage.


In short, several man weeks of knowledge invested in the former release will be wasted and for what? When I'll get on top of this, my converted system will look exactly like the one I have now (therefore I will probably NOT do the transition unless I find compelling performance reasons, and so Citrix will lose some money - however little - on this).

Why this blatant disregard (contempt?) for the time your user base puts into learning a product? This is not unique to Xen: it is the same mindset that - for no apparent good reason - squirreled away a number of icons under the "Adminstration Tools" pseudo-folder of Windows' control panel. I have thousands of other vendor neutral examples.


Now, it is not like these people don't know how annoying this is. In fact, the tools
they use are almost immune from this pointless interface churning: editor shortcuts and compiler switches have amazing staying power (emacs' shortcuts and gcc's switches have been cast in stone for more than ten years, as I remember)


So why not showing the same respect for the tools they inflict on their users?

Cheers,
alf

P.S: I will be doubtlessly told that I am just a luddite resisting change etc. So please explain why 'xm create' is worse than its moral equivalent ('xe vm-install ...xe vm-start', presumably).

No comments: