On 2012-10-17, Peter Chant <p...@petezilla.co.uk> wrote:
> of the features present in konqueror (KDE 3 days). Fluxbox was far too hair > shirt for me.
?????
I use fluxbox so I can use KDE efficiently. I always load full kde,
the run fb. Most slack fb menus are already tweaked towards kde.
Whatever new apps kde has that aren't already in fb menus can be added
easily. I don't understand what kde has that I can't get from fb, yet
leave all that bloat and drag behind until I need it. What?
Wallpaper? Screensavers? FB does all that. Yet whackonutty,
neposuk, and sqweegi remain off unless I need them ....for whatever
bizarre reason.
On Wed, 17 Oct 2012 19:01:06 -0400, notbob <not...@nothome.com> wrote:
> On 2012-10-17, Peter Chant <p...@petezilla.co.uk> wrote:
>> Sorry, don't understand. KDE appears to have a front end for locate,
>> but I
>> can't see the relevence to this topic.
> Then, we're even. I can't see the need for a frontend for slocate.
> slocate foo | grep fee
> That's too difficult?
What if I want to search for files based on their contents? What about
their meta data? What about their video encoding type? What if I want to
search through all PDFs that I have tagged as pertaining to a given topic?
What if I want to search based on tags, contents, types, *and* date? I use
the above locate pattern very often, but it's not always enough.
-- Aaron W. Hsu | arcf...@sacrideo.us | http://www.sacrideo.us Programming is just another word for the Lost Art of Thinking.
On Wed, 17 Oct 2012 15:50:45 -0400, Aaron W. Hsu wrote:
> Nepomuk is generally designed for indexing files that are local to the
> user, and hence, it makes sense for the index to be with the user.
That might seem like a good idea, but...
> Nepomuk only indexes the Home directory by default, and I see little
> reason to index anything outside of the Home directory. This means that
> there will be no duplication among multiple users of a system, because
> they will not be indexing the same files.
In my experience, the index files created has taken more space than the rest of the home directories. However, I have never investigated this any further, whenever I have found this kind of big index files I have quickly made sure that nepomuk has been turned off again for the user and the files removed.
What about symlinks? Most of my users do have symlinks from their home directories pointing to several terabytes of data in common areas. Could those symlinks explain those big index files?
regards Henrik
-- The address in the header is only to prevent spam. My real address is:
hc351(at)poolhem.se Examples of addresses which go to spammers:
root@localhost postmaster@localhost
On Wed, 17 Oct 2012 22:47:02 +0100, Peter Chant wrote:
> Henrik Carlqvist wrote:
>> The nepomuk approach with index files in the users home directories
>> shows that KDE is intended as a toy OS replacement for desktop
>> computers with only a single user.
> Is that fair? What is toyish about it? A set of integrated basic apps
> seems like a good plan.
Calling some OS toyish might be a little hard, but I wouldn't say that it is unfair or a bad description.
A toy OS like Microsoft Windows has its historical roots from computers known as PC where PC stands for "Personal Computer". Having matured from DOS where the OS didn't even have any kind of login, simply assuming that the user only were a single person.
Linux has its root in Unix and has allways has the ability to run multiple processes owned by multiple users logged in at the same time.
To most users Microsoft Windows does look good, and to most users the computer they log in to is their own. For those users Microsoft Windows might be an OS good enough. They don't need to log in to their computer from the network using ssh. They don't need to use X windows to put windows from other computers on their screens. But for some reason some of those users might want to replace Microsoft Windows with something else, and then Linux with KDE tries to be a replacement.
Usually, what these kind of users lack the most in the Linux environment is to run their favorite PC games that they used their toy PC for.
A set of integrated apps might seem like a good plan, but the unix way of doing things has allways been to have small independent apps doing different things but still being able to cooperate. The Microsoft/KDE way if "integrating" things usually leads to a twofold path. First the user ends up having to use more of their products than he/she first intended. Second the user gets trouble to use their products together with products from other vendors.
regards Henrik
-- The address in the header is only to prevent spam. My real address is:
hc351(at)poolhem.se Examples of addresses which go to spammers:
root@localhost postmaster@localhost
On 2012-10-18, Aaron W. Hsu <arcf...@sacrideo.us> wrote:
> What if I want to search for files based on their contents? What about
> their meta data? What about their video encoding type? What if I want to
> search through all PDFs that I have tagged as pertaining to a given topic?
> What if I want to search based on tags, contents, types, *and* date?
On 2012-10-18, Henrik Carlqvist <Henrik.Carlqv...@deadspam.com> wrote:
> A set of integrated apps might seem like a good plan, but the unix way of > doing things has allways been to have small independent apps doing > different things but still being able to cooperate. The Microsoft/KDE way > if "integrating" things usually leads to a twofold path. First the user > ends up having to use more of their products than he/she first intended. > Second the user gets trouble to use their products together with products > from other vendors.
On Thu, 18 Oct 2012 07:36:05 -0400, notbob <not...@nothome.com> wrote:
> On 2012-10-18, Aaron W. Hsu <arcf...@sacrideo.us> wrote:
>> What if I want to search for files based on their contents? What about
>> their meta data? What about their video encoding type? What if I want to
>> search through all PDFs that I have tagged as pertaining to a given
>> topic?
>> What if I want to search based on tags, contents, types, *and* date?
> find
> grep
Any suitable incantation or workflow centered around grep and find that
does what Nepomuk and it's ilk do will inevitably be a poor man's
implementation of the same, with the same problems. Firstly, you need a
way of making fast indexing possible, without needing to rescan the entire
directory structure everytime you use find, how do you do that? It needs
to index more than just filenames. Now you're basically at Nepomuk +
Strigi.
-- Aaron W. Hsu | arcf...@sacrideo.us | http://www.sacrideo.us Programming is just another word for the Lost Art of Thinking.
On Mon, 15 Oct 2012 12:22:08 -0400,
Aaron W. Hsu <arcf...@sacrideo.us> wrote:
> On Mon, 15 Oct 2012 03:57:05 -0400, Peter Chant <p...@petezilla.co.uk>
> wrote:
>> I like KDE but I've had Slack 14 installed over a week now and the system
>> _still_ seems to be indexing things.
> [...]
>> Any thoughts? Binning kmail etc and turning off nepomuk seems like one
>> option and switching to xcfe another. But I would rather that KDE
>> worked.
> There is a setting in the Desktop Search settings for > controlling how much of the currently available resources > Nepomuk and others try to use when doing their indexing, you may > want to tweak that.
Although it would tend to *increase* the total time needed for indexing, you could also use nice to restrict the system resources allowed for the process.
-- Theodore (Ted) Heise <t...@heise.nu> Bloomington, IN, USA
