Re: Workspace/desktop switching

From: Terry Hancock (hancock_at_anansispaceworks.com)
Date: 11/18/03

  • Next message: Johann Spies: "Re: Setting fcc-hook in mutt" 
    To: debuser <debian-user@lists.debian.org>
Date: Tue, 18 Nov 2003 09:19:00 -0600

    On Tuesday 18 November 2003 01:19 am, Alex Malinovich wrote:
    > I've been seeing a lot of discussions about various WM's lately, and
    > everyone seems to be extremely concerned about easy workspace switching.
    > I'm just wondering what exactly everyone uses workspaces for?

    Use #1: Efficient parallelizing
    Obviously it's going to depend on what you do with your machine,
    but I do a lot of time intensive hand-off tasks. This often has to
    do with download/upload times, but also sometimes with disk-access,
    or even CPU time on a remote machine (won't help if it's local,
    unless we're talking about multiple processors which I don't have).

    Why wait for a long download when you can be browsing the
    web or working on your development project at the same time?
    And switching desktops is usually *much* faster (2-3X) than
    opening/closing individual app windows in actual computer
    time, and even faster when you consider that it takes fewer
    mouse clicks.

    Use#2: More space to spread out in
    [Doing this now]
    Just as an example (and this incorporates some of #1) I develop
    Zope products (web apps). I keep a master copy on the file
    server, and periodically mirror it into the "Products" directory and
    refresh as part of the development cycle. During this time, I
    usually have one desktop with two gVim windows (I prefer
    separate windows over the internal panes, BTW), one on the
    main source module I'm editing and one on to browse files it
    depends on -- usually to read, but I sometimes need to make
    edits there. Then I have an xterm logged in as the Zope user
    which I use to mirror the results.

    I use a 2nd desktop to have a browser window pointed at the
    local Zope server, usually with several tabs addressing the
    Product refresh page, an object-tree browser page, and the
    page where I actually see the output.

    I use a 3rd desktop with a tabbed browser pointing at the local
    Python manual, Zope.org, Python.org, and google for researching
    questions as they come up.

    For testing: 1) save source files, 2) up arrow and enter to run
    the "cp" command that copies the sources to the Products
    directory, 3) swap desktops, swap tabs, click "refresh" wait
    for response, 4) swap to output screen, click reload and check
    the results. It's not quite as simple as compile, link, run, but
    web apps are like that. There's also a unit-testing mode which
    hopefully I've mostly completed before getting to this level.

    I use a 4th desktop to hold my email client. I may have one or
    two emails open at a time. Sometimes I'll start writing a question
    to a list and then realize I can figure out the answer for myself
    and stop. Of course, I also use it to take a break and answer
    a question myself. Like now. Sometimes I need to swap over
    to my reference desktop to check something about my answer
    or verify a URL.

    So that's four. Right now I have two more in use, because I also
    have a separate development project that I'm just collecting
    information for. That's three xterms logged onto remote machines
    at my clients site: one each on two machines (different architectures,
    as I have to install software for both), and one with w3m running in
    the window. I'm using it to download software package files. That's
    on desktop #5.

    The 6th desktop just has XMMS in it, because I'm listening to music.

    Occasionally, I minimize apps. But as I said, it's usually more
    faster to switch desktops than to go to and from the taskbar.

    I used to be limited by the CPU speed and RAM (2 or 3 big
    apps would strap the computer), but I've upgraded, so now
    I'm mostly limited by how much I can think about at once
    (which is how it ought to be ;-) ).

    Note this is KDE 2.2.2 that comes with "Woody". My biggest
    complaint is that I haven't figured out how to sweep virtual-desktop
    style from desktop to desktop (like I could with FVWM), but
    must manually click which one I want. I figure there's probably
    a setting somewhere that controls that (or will be in KDE 3?),
    but it's not really such a pain -- I've already gotten used to it.

    Cheers,
    Terry 

    --
Terry Han*** ( han*** at anansispaceworks.com )
Anansi Spaceworks  http://www.anansispaceworks.com
-- 
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
  • Next message: Johann Spies: "Re: Setting fcc-hook in mutt"