And what everyone forgets ...
And what everyone forgets ...
Posted Nov 5, 2013 21:40 UTC (Tue) by Wol (subscriber, #4433)In reply to: Seigo: on introducing new ideas to free software communities by nhasan
Parent article: Seigo: on introducing new ideas to free software communities
KDE4.0 should never have been released to the community in general, because most of the necessary apps WEREN'T there, because it's hard to develop on quicksand if the API keeps changing underneath you.
Unfortunately, as is the nature of open development, naive users did get their hands on 4.0 (or had it foisted on them) and of course they were upset because most of the superstructure (unsurprisingly) didn't work.
Not helped, unfortunately, by the disaster that was semantic-desktop/strigi/nepomuk. Why the f*** would I want something *enabled* *by* *default* that gives me HALF AN HOUR boot times (if not longer) and is touted as "speeding up all those things I never do".
That is why I now have both XFCE and LXDE now installed, although I rarely have cause to use either (apart from my old Y2K era system which is far too long in the tooth really. Maxed out with RAM in the *mega*bytes).
Cheers,
Wol
Posted Nov 5, 2013 23:29 UTC (Tue)
by pboddie (guest, #50784)
[Link] (2 responses)
Meanwhile, Tracker still manages to saturate the CPU and I/O bandwidth in its relentless quest to make absolutely everything on my system searchable. No wonder Debian will switch to XFCE: I bet they've had a lot of bug reports about people not being able to interact with their computers and that "Linux has become slow".
One related thing that annoys me about KDE 4, erm, Plasma Desktop is the way that you can't launch applications for a while after the panel has loaded. That feature of Windows surely didn't need to be copied. I really liked the way KDE 3 was interactive immediately after the splash screen went away.
Posted Nov 7, 2013 19:42 UTC (Thu)
by ovitters (guest, #27950)
[Link] (1 responses)
Posted Nov 7, 2013 23:23 UTC (Thu)
by pboddie (guest, #50784)
[Link]
I remember loosely following the furore about Tracker and whether it was written appropriately. Having had some experience with database and indexing systems, there are situations where you just want to crunch the data as quickly as you can, and you get used to seeing big numbers in things like vmstat.
Quite how Tracker manages to tie up lots of resources while established data processing solutions generally do not might have something to do with a "perfect storm" of sub-optimally stored data - rather likely given the "scan the filesystem" nature of desktop search - and processor-intensive data extraction tools. Database and indexing systems tackle the former issue head on, of course, as I suspect Tracker also does once it has populated its index.
The solution is to dial down the aggressiveness of the indexing instead of assuming that the scheduler will keep things civilised. Sadly, I rather suspect that people have experienced Tracker's "warming up" period and have written off Debian and other distributions as a result.
Random Internet page showing that this isn't just me: http://forums.opensuse.org/english/get-technical-help-her...
And what everyone forgets ...
And what everyone forgets ...
And what everyone forgets ...
