Why XFCE gona die if no action taken

Manoa · 2013-08-08 12:19:28

- Porting everything to QT

The only suggestion from the OP I would support is porting to Qt.

why ? it sucks, if you feel the need to port to another UI framework, FLTK smokes anything in the linux existence by a huge margin, both in terms of size and in terms of speed (look at Dillo - no GUI browser comes even remotely close to it's size or speed)

Xwayland/Xmir
why ? it sucks, every few weeks there are benchmarks published on phoronix about them and EVERY time X.Org smokes them (HARD! by between 50% and 30% in terms of speed), and there is NO WAY anyone is going to rewrite all the X.Org applications from scratch to work with the native versions of those X servers, X.Org: linux is stuck with it, forever

tablet/phone
why ? it sucks, those are the most claustrophobic, uncomfortable, limited in functionality devices ever (defective by design), it's TOTAL crap "hardware" and I wouldn't take it even if I was offered them for free, make no mistake - these are NOT computing devices (and just because many people buy them and use does not make them such, they are 95% flash and 5% photo [and that means they are for show to other people "look ma, I can play doom on my phone"] and you tablet/phone owners ask yourself this: have you ever done serious work on them ? such as opened 100's of web pages, programmed any application ? wrote a couple of web pages ? wrote a thesis ? performed data analysis ? wrote VB scripts ? - it proves how useless they really are, and the users ? are all mouse aimers, I have yet to see a single keyboard capper using them for anything other than show "ma, look how rich I am to have a 1000$ phone that does absolutely nothing useful")
and also this:
* want to run more applications ? PC: add some more memory, tablet: "LOL, WTF DUDE ?"
* out of space for data ? PC: add hard drive, tablet: "LOL, WTF DUDE ? you already used the SD slot"
* some program running too slow ? PC: change necessary hardware, tablet: "she's given it all it got captain!"
* want to use a mouse ? PC: sure !, tablet: "LOL, WTF DUDE ?"
* monitor sucks, want to replace with better/bigger ? PC: sure ! and as many as you whant !, tablet: "LOL, WTF DUDE ?"
* want a reliable network connection ? PC: plug a cable (and as many as you have network cards for), tablet: "LOL, WTF DUDE ?"
..."LOL, WTF DUDE ?"
..."LOL, WTF DUDE ?"
..."LOL, WTF DUDE ?"

as for everything else, xfce is not going to die, tablet/phone are not going to take over pc, and xfce does not require any changes (and it is not at a point for, anyway), +1 for GNOME sucks, +1 for Microsuck Dildows 8 sucks, ubuntu sucks ? in terms of size and speed ? sure ! but I can't deny the fact that it is a major contributor to linux as a whole (more linux users is always better for everyone both now and in the future, even if that user base comes from ubuntu)

Last edited by Manoa (2013-08-08 13:52:10)

bwat47 · 2013-08-12 23:23:33

Manoa wrote:

- Porting everything to QT
The only suggestion from the OP I would support is porting to Qt.
why ? it sucks, if you feel the need to port to another UI framework, FLTK smokes anything in the linux existence by a huge margin, both in terms of size and in terms of speed (look at Dillo - no GUI browser comes even remotely close to it's size or speed)
Xwayland/Xmir
why ? it sucks, every few weeks there are benchmarks published on phoronix about them and EVERY time X.Org smokes them (HARD! by between 50% and 30% in terms of speed), and there is NO WAY anyone is going to rewrite all the X.Org applications from scratch to work with the native versions of those X servers, X.Org: linux is stuck with it, forever

You do not know what you are talking about.

1. Porting to FLTK would be pretty silly, its not nearly as well supported as GTK or QT. I don't necessarily agree with porting to QT though (they are already working on porting to GTK3 anyway afaik).

2. Xwayland/Xmir is not finished yet, there are still several parts that need to be completed (either in xwayland/xmir or in other parts of the graphics stack) before performance will be good. once it is finished performance is expected to be nearly identical to just X. For example here is a post from an xmir developer: http://blog.cooperteam.net/2013/07/xmir … mance.html

And regarding xwayland, xwayland won't run whole DE's like xmir, it will just focus on running single apps for compatibility purposes, and again once it is finished there should be little to no performance hit.

Most applications will work fine on wayland because most of the work is done on the toolkit level. Most applications that uses GTK3 ot QT5 will work fine with no modification. Yes there will be certain apps that will probably take a very long time to port to work natively, but that will not be a big deal because xmir/xwayland will provide seamless compatibility.

You are bashing these projects by looking at benchmarks of unfinished software on moronix.

I'd love XFCE to port to wayland someday personally, it would solve some major issues that it has now (such as video-tearing). Wayland provides far more efficient compositing and IPC than X does. If XFCE never ports to wayland they will probably eventually be forced into a "Legacy DE" niche. I can understand them waiting until wayland and the underlying infrastructure is finished and mature before even thinking about porting though, because XFCE has always been conservative and stable and that's one of the things I like about it.

I do agree that XFCE targeting tablet/phones would be silly though. People use XFCE because it is a sane DE focused on desktop users. We already have gnome and unity for people that are interested in tablet/phone focused DE's. XFCE should stay focused on the desktop now more than ever because other DE's are starting to lose that focus. However, as I mention above I don't think this means that XFCE should stick woth outdated tech like X and GTK2 under the hood. If XFCE never moves towords wayland then I may switch to MATE in the future (they've announced that they plan to fully support systemd, GTK3, and wayland which I like).

Last edited by bwat47 (2013-08-12 23:28:21)

Manoa · 2013-08-13 16:47:00

1. so what now dude ? be stuck with bloated crap that QT/GTK are instead of porting to something that PWNS and just because of support ? what support is needed ?
I guess we are waiting for the new X servers to be complete and then make some tests

bwat47 · 2013-08-14 02:55:38

Most applications are GTK or QT, porting XFCE to an obscure toolkit like FLTK would result in inconsistant theming and other such integration issues.

Last edited by bwat47 (2013-08-14 02:56:03)

Sideburns · 2013-08-14 03:14:43

I think that the important thing is that almost everything in Xfce works just fine the way things are. Porting everything to a new toolkit just because it's new would cause a whole lot of problems and fix nothing. As things stand, Xfce isn't broke and doesn't need fixing.

Xfce Forum

Sub domains

#26 2013-08-08 12:19:28

Re: Why XFCE gona die if no action taken

#27 2013-08-12 23:23:33

Re: Why XFCE gona die if no action taken

#28 2013-08-13 16:47:00

Re: Why XFCE gona die if no action taken

#29 2013-08-14 02:55:38

Re: Why XFCE gona die if no action taken

#30 2013-08-14 03:14:43

Re: Why XFCE gona die if no action taken

Board footer