After a bit of a user revolt on the openSUSE-KDE list (never under-estimate the pathological need of KDE users to configure every little knob) this advanced search configuration should make the 1% very, very happy... Other users can continue to use the 'old' dialog. Now, how do we expose this advanced one without cluttering the interface...
19 plus ones
Shared publicly•View activity
View 84 previous comments
- Sorry, but I think everyone should always be "allowed" to disable it, when for example they simply don't need an indexer or maybe they want to use a another indexer instead...
Anyway, I have big files in my home directory because I'm messy and that's just how I work with stuff. My point is probably clearer when I switch the perspective around: should users have to alter their workflow to fit to how the file indexing service of their desktop environment works?
Again, I have no idea how to solve this, other than disable indexing for "big" files by default... Maybe there could be a counter of how much was indexed for a specific file and set limits on that because the raw file size looks like a bad criterion when I think about documents with some text, but lots of images... I just think it's a real problem.May 5, 2014
The problem if we make Baloo optional is that we cannot deeply integrate it into applications without problems. For example in Plasma Active, when we decided to put Nepomuk (because there was no Baloo at the time) front and center, it opened whole new opportunities for us. We didn't need a classical "File Open" dialog, because we could just directly use the Nepomuk index. We could make a file browser which operated on file types, tags and content instead of a folder hierarchy. We could make a "bookshelf" application for elegant browsing in ebooks based on metadata from Nepomuk. None of that would have been possible if we couldn't count on Nepomuk being always there.
I don't say we plan such a deep integration for Plasma Desktop, but without being sure that everyone has Baloo running, we always need to provide an alternative for every feature which uses Baloo (or leave users with half-broken applications). For example, KMail uses Baloo for auto-completion of email addresses. If we can't rely on it being there, we'd have to implement an alternative method for auto-completion with its own index just for people who don't like Baloo. This way we can never use Baloo's full potential, but we want to!
That's why it's Vishesh's goal is to make Baloo so good and light-weight that we can make it mandatory without causing people trouble.
Your idea to just stop indexing and internally blacklisting a file after a limit of X indexed characters in it may actually work, though even then it would still save you trouble if you just put your OSM files into a separate folder.May 5, 2014
- I already have a kde dev account, thought about submitting the adv kcm somewhere, but wasn't sure where it should go (extragear?) and didn't want to tread on anyones toes.May 5, 2014
- With regards to apps needing baloo to enable all their features - if people want to disable baloo, then they can just wear not having the features, e.g with akonadi - no address completion, or not use akonadi/kmail at all.May 5, 2014
As for where to put the KCM: I'm not a developer, so I'm no expert in the repository hierarchy. I'd suggest asking on the Plasma list where they think it should go.
And no, as long as you don't start a big campaign for replacing the current KCM with yours by default, you won't be treading on anyone's toes.
The goal of the new KCM was not to piss off advanced users, but to make life easier for "casual" ones. What I would like to see, though, is continued cooperation with both advanced and casual users in order to make the default KCM suitable for more users without overburdening the majority.
Those users who absolutely need very fine-grained control, your KCM works well and nobody has a problem if they use it.
I have an idea for how to handle disabling of Baloo without losing information for its improvement. I've already talked to Vishesh about it but he hasn't decided whether he wants to implement it. I'll tell more about it if he decides to adopt it.May 6, 2014
- It needs to go to the playground first. After that, when you feel it is ready, you can submit it for review.
After that, it could go to extragear/utils or similar.May 6, 2014
Add a comment...