Announcement

Collapse
No announcement yet.

KDE Plasma 5.21 Released With Better Wayland Support, Desktop Improvements

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #61
    Originally posted by caligula View Post
    AFAIK these indexers don't perform any miracle OCR on PDFs. You can easily grep stuff in many cases by running pdftotext + grep.
    Sure, but at some point it's easier to use a search tool than to run a shell command of all the likely files each time. (Also, use ripgrep, it's much better than grep and has that Rust goodness )

    I'm not totally against the idea of indexing, but these tools often perform badly and introduce a constant background load when doing heavy lifting with files. The indexes might also be quite large, especially if you have terabytes of data and forget to exclude some unimportant files.
    Nepomuk and early Baloo had such problems. As I wrote Baloo has worked well for me for years. I don't even notice it on this 8GB laptop with 256 GB SSD. FWIW I just `git clone`d a 6,000 file repo: Baloo's file extractor used 20% of one CPU for four minutes then dropped back to 0%. My index of 45,000 files is 450 MB.

    Comment


    • #62
      Originally posted by cl333r View Post

      I always wondered what kind of people use baloo - what type of garbage case that must be to need a file indexer to find your stuff?
      Ah, someone who has pretty much every document from every personal computer since the late 80's online in a 12TB disk pool? Every interesting PDF or text file?

      The whole _point_ of a computer is to augment human brainpower - why should I remember or organise something when the computer can do it for me?

      That said, baloo is kind of piggy and does not appear to have a successful mechanism to dynamically reduce it's load - seems to grab some kind of lock fairly frequently and generally it sucks even on a machine with hundreds of gigs of RAM and 20 cores of Xeons.
      Given that it was introduced as a "performance" replacement for Nepomuk and semantic search (which I was quite looking forward to) it is a bit disappointing; but I still use it. I just turn indexing on when I expect to be away from the machine for a few hours. Putting the index on its own SSD helps - but not as much as I'd hoped.

      Comment


      • #63
        Originally posted by angrypie View Post
        Are you talking about Plasma/KDE in general? I gave up on reporting bugs to them when any report slightly opinionated (but still civil) would be labeled as a "rant" and closed. Not acknowledging reports is... progress I guess, at least they stay there.
        I've reported over 80 bugs in KDE/Plasma over the past ~2 weeks, and I've seen that attitude too... in less than 5% of the cases.

        Now for a really dumb question: how exactly do you actually use Wayland in KDE neon Plasma? And why is it so hard to find out? (I've been searching for more than an hour and only found what appear to be arcane ways while people act as if it was a simple choice on the login screen)

        Comment

        Working...
        X