[PATCH] Making Soprano optional (again) in kdelibs

Sebastian Kügler sebas at kde.org
Wed Dec 30 13:31:27 CET 2009

On Wednesday 30 December 2009 00:43:31 Jorge Manuel B. S. Vicetto wrote:
> On 29-12-2009 18:30, Sebastian Kügler wrote:
> > I get your point (bloat), but being vague and sarcastic is not the way to
> > go about it. The best is to provide numbers to back it up and show the
> > real costs of running two databases. I believe they're rather easy to
> > attain, given the right tools, so I don't really get why you're coming
> > with a long rant how relational database systems bloat KDE (my
> > interpretation), instead of just listing those numbers.
> If "common sense" won't cut it, do I need to start a vote, forum thread
> or something else asking KDE users (I can do it for Gentoo alone) if
> they are worried/upset about this? Should I ask how many "former KDE
> users" stopped using it because of the MySQL dep?

So the next step would be quantifying the performance, and comparing it to 
alternatives. To get you going, here's a couple of measurements that would be 
interesting to quantify this bloat you're talking about:

- How much memory does kmail take using akonadi, and not using akonadi (on the same 
  email data set)?
- How much time does it take to start up in both scenarios?
- How much time does it take to load a big folder with emails?
- How much time does it take to search for a particular email?
- ...

With that information, you can tell the developers: Look, my email application now 
takes twice (randomly made up figure that would support your claim) the memory and 
isn't any faster. That would be a good start. "I can find people who don't like it on 
the Internet" is useless to us, however.

Note that the above comparisons completely ignore any gained features by the use of 
akonadi, so it's not completely fair game. For the exact same use case, the data 
would still be interesting though.

One particular point that's often forgotten is that the alternative is not running 
MySQL or not, but running MySQL vs. using another storage mechanism. There's no free 
lunch involved here, it's about offsetting alternatives.

> If you want, I can point you all to the user rants in Gentoo bugs about
> MySQL dependency for KDE as a whole and amarok as one particular KDE app.
> Why is it so hard to understand that KDE is getting more and more bloat?
> By the way, no other distributions / packagers have issues with the KDE
> dependencies?

That would give us a good grip on how much hand-waving half-knowledge there is 
around. I can easily "go on the Internet and find people who want to complain about 
something", especially during slow days around christmas.

I'm starting to get really amazed how people are willing to condemn the use of 
databases without even coming up with a single performance measurement backing up the 
claims of bloat. And then people go wild based on wrong assumptions. :/

We can do better than that, no?

http://www.kde.org | http://vizZzion.org | GPG Key ID: 9119 0EF9

More information about the release-team mailing list