Why not incorporate/adopt E17 as DM/WM?

Hey there,

There's a lot of things that go into choosing a toolkit. For us, GTK+ is choice for several reasons:

* It does all the things we need it to do. GTK+ is rendered with Cairo, giving us beautiful, crisp, and scalable widgets. It's themed with a CSS-like syntax which lowers the barrier to entry in creating great visual design. It supports RTL languages, a11y, etc etc.

* It has a really big 3rd-party developer community. A platform is only as good as the apps that run on it, and the biggest names in free software target GTK+ desktops. Additionally there are a lot of resources to help developers work with GTK+.

* All our apps are already written in it! We originally started out on GNOME, so all of our apps are already written in GTK+. Moving to another toolkit would be a really big process and take a lot of time. It would have to have some kind of massive pay-off to make it worth doing.

As far as enlightenment as a desktop environment or window manager, we already have our own desktop environment called Pantheon that includes our Gala window manager. So we already have something custom made. It's not GNOME, XFCE, or LXDE.

Hi Daniel:

  Thanks for the prompt and thoughtful reply. I'll look more closely into Pantheon and Gala over the weekend to see how its less of a re-branding and more of an original work. Specifically, how it differs in configuration and interacts with Xorg more than the look-n-feel of its widgets and sprites.

Note that implementing E17 in no way detracts from using GTK-based apps; its simply (IMO) a
better DM/WM than the others (TWM/SLim/Mdm/LXdm/XFCE/KDM/GDM/...) both in terms of resources and aesthetics.
My only reason for using KDM/XDM and my primary complaint for EDM is that it can't handle XDMCP (neither can LXDM).
But there is always NX,X11VNC,SSH-X as alternatives. "Many ways to do one thing" - is very embraceable.

  And for future reference, E is better suited to slim/mobile environments. After more than a decade I have come to seriously dislike Gnome in any version, despite still using it on some hosts. Certainly, one could always make the argument that a DM is just a DM (same for WM).

It's not a putdown, since i use and depend on it, but in my experience, GTK/gtk2-engines/libs is no prize either. In terms of both freeze/lock ups, front-end buggyness, ..., and metacity.
One need only look at the host of GLib-GObject-CRITICAL/Gtk-CRITICAL in .xsession errors
and stuff like ~/.themes/blah/gtk-2.0/gtkrc:62: error: invalid whatever
then follow the related forum threads to realize that some ppl are going to be left hanging. Add to that the number of apps that get appended to export GTK_RGBA_APPS for anyone wanting some eyecandy and that onion gets layered with even more complexity.

Simply put E, for all its warts, handles Xorg/X11 resources better, is faster and more attractive. Not that I'd ever advocate Wayland (as I see Xorg running quite speedily on my N900), but I also suspect that E will be more suited to that env as well.
I believe that EOS may find its future hinges on tablets,netbooks... mobile devices than on the staid desktop and if that's the case then GTK may prove a challenge. E17 can handle both GTK and QT apps equally well.

Anyhow, to each their own; and I'm sure it will only be a matter of time before someone puts E on EOS anyhow. My suggestion was in hopes of making it the default DM/WM out of the gate.

My true desire is that I can make a persuasive argument for making an EOS dist that contains both full client AND a server, out of the box; touted it as the most "secure distro available". If you want an email server, then RBL/SA/DKIM and ports 465/995 by default; email clients setup to use gpg,light- httpd running SSL only, jitsi/wildfire to provide XMMP, TLS being a no brainer, etc...

Currently there are no distros that offer this capability, and none are anticipating end-users wanting to run home-based servers. But (I believe) this is going to change as DPI increases and CDN/Citadels (facebook, twitter,et al) treat their members as a profit-source. Giving users full control over their own data (and methods of sharing it) will mandate
distributed end-user computing as the only solution.

Any distro that enables each and every user to decide how/what they want to communicate and the ability to white/black list who they want to allow in their circles is the deliverable for a free-net.

In closing FWIW, my bent is toward slackware workstations and CentOS servers, though I install ubuntu on other's machines and if doing again, would probably opt for mint. But for the past two years I've been advocating that any distro which takes the (mostly trivial) efforts of providing user-friendly GUI's that let non-technical, non-geeks, configure their client/server apps-n-services and associated security measures will be taking Linux to its next evolution.

If EOS can rise to this undertaking (as a community effort) then, barring all else, it know it would become the preferred dist of choice in very short order.

Be well, and remember to take walks and breathe deeply

Rion

Dated, but even more relevant today:

http://jeffhoogland.blogspot.com/2010/10/x-reasons-to-give-e17-try.html