You are not logged in.
i'd still really like to know which distro gives either full or near full xfce "components"
or is only way to install debian or fedora and choose xfce during install ?
Offline
What is your definition of "full xfce components"? What do you consider full Xfce?
Is it just the apps listed here - which are the only components/apps that are developed by Xfce developers? Note: this doesn't include a display manager, any system management tools, no video drivers, no web browser....
I can speak to ArchLinux. Manually install the base system and then the "xfce4" and "xfce4-goodies" packages (don't use the arch installer - it installs other non-Xfce components). This will get most of the packages - you may need to install others like xfce4-genmon-plugin and xfce4-docklike-plugin.
Please remember to mark your thread [SOLVED] to make it easier for others to find
--- How To Ask For Help | FAQ | Developer Wiki | Community | Contribute ---
Offline
and i really like to know which distro gives the higest xfce performance?
Last edited by deepforest (2022-11-22 01:03:03)
Offline
and i really like to know which distro gives the higest xfce performance?
Its not an easy question to answer. You could conceivably take any distro, limit the number of installed apps and running processes, optimize settings and achieve relatively good performance (depending on what this means to you). But I don't think anyone has done this that I am aware of. Would be a good exercise to undertake if you're interested in finding an answer. Just make sure you compare apples to apples.
Please remember to mark your thread [SOLVED] to make it easier for others to find
--- How To Ask For Help | FAQ | Developer Wiki | Community | Contribute ---
Offline
What is your definition of "full xfce components"? What do you consider full Xfce?
Is it just the apps listed here - which are the only components/apps that are developed by Xfce developers? Note: this doesn't include a display manager, any system management tools, no video drivers, no web browser....
I can speak to ArchLinux. Manually install the base system and then the "xfce4" and "xfce4-goodies" packages (don't use the arch installer - it installs other non-Xfce components). This will get most of the packages - you may need to install others like xfce4-genmon-plugin and xfce4-docklike-plugin.
well there, you just said at the bottom, base, the goodies and this plugins
mind you i know nothing of xfce (except what i tried on few distros, mostly Mint)
is there maybe a gallery or website that shows visual what "full" xfce consists off ?
Offline
is there maybe a gallery or website that shows visual what "full" xfce consists off ?
There is Xfce's gitlab presence as well as the Xfce docs.
Please remember to mark your thread [SOLVED] to make it easier for others to find
--- How To Ask For Help | FAQ | Developer Wiki | Community | Contribute ---
Offline
deepforest wrote:and i really like to know which distro gives the higest xfce performance?
Its not an easy question to answer. You could conceivably take any distro, limit the number of installed apps and running processes, optimize settings and achieve relatively good performance (depending on what this means to you). But I don't think anyone has done this that I am aware of. Would be a good exercise to undertake if you're interested in finding an answer. Just make sure you compare apples to apples.
I have 3 distros Arch, Devuan, Arch.
On Arch and Devuan i feelling constant responce lag on all my actions. But on Artix lag is much lower and all responce is almost instant.
Were is the reason for this different behavior? What can affect to responsiveness and can it be improved without upgrading hardware?
Last edited by deepforest (2022-11-22 15:23:57)
Offline
On Arch and Devuan i feelling constant responce lag on all my actions. But on Artix lag is much lower and all responce is almost instant.
Were is the reason for this different behavior?
Compare the differences:
kernel and kernel version
video driver and video driver version
filesystem type
Xfce version
number of installed packages and what are the differences
number of running processes and what are the differences
memory footprint
init system
What can affect to responsiveness and can it be improved without upgrading hardware?
That would depend on the above, but it could potentially be anyone of them that would need to be optimized. For example, maybe Artix installs have the packages and has half the running processes of the other 2 systems? Perhaps Atrix provides additional performance tweaks that the other two don't? Asking the Artix devs may get you some answers.
Please remember to mark your thread [SOLVED] to make it easier for others to find
--- How To Ask For Help | FAQ | Developer Wiki | Community | Contribute ---
Offline
from a genmon I use this to gauge process count and business; how many active/how many 'on'
cat /proc/loadavg | cut -d ' ' -f 4
1/400 on the low side 10+/600+ on the high side
Offline
reason for poor performance xfce on Devuan its Devuan itself? not xfce?
i have never worked with Devuan so i don't have the exposure to it to guess close. dependencies can include how many services are running, how much I/O they do, page swapping burden, etc. we pretty much have to go on reports from other people and that can vary between people. someone who has used a single ultra-fast hardware who tests a distro on a slower machine can report it as slow. i've seen the same machine (no changes) reported as slow by one person and fast by another. it can depend on what they do and their experience on others.
someone might set up a common machine to test a bunch of distros and find that distro A is faster than distro B. then, a totally different machine build might find that B outperforms A. different distros can be built or configured for different setups of hardware.
when an application is slower, is it getting delayed by CPU or by swapping or by I/O? hardware plays a huge role in determining performance.
i've even seen a case of more RAM actually slowing down a particular application. apparently it discovered it had access to more memory and used it, resulting in a larger demand on swapping. i've even found that increasing the size of swap space slows down a whole machine (because it uses that space and each page take time).
these days, we see more setups with solid-state storage devices and now all new needs to configure for performance as well as wear on the SSD.
Offline
i've even seen a case of more RAM actually slowing down... configure for performance as well as wear on the SSD.
More memory is always a good thing.
Adjust 'swappiness' and install zram*.
The only time more memory has a speed impact is when the system board slows down the ram speeds in order to accommodate more 'b' and 'c' slots or higher rank modules. Even then only the benchmarks notice.
Offline
from a genmon I use this to gauge process count and business; how many active/how many 'on'
cat /proc/loadavg | cut -d ' ' -f 4
1/400 on the low side 10+/600+ on the high side
Artix
freeartist-p31ds3l:[freeartist]:~$ cat /proc/loadavg | cut -d ' ' -f 4
2/482
freeartist-p31ds3l:[freeartist]:~$
Devuan
freeartist@devuan:~$ cat /proc/loadavg | cut -d ' ' -f 4
1/594
freeartist@devuan:~$
Last edited by deepforest (2022-11-25 22:00:46)
Offline
Compare the differences:
kernel and kernel version
video driver and video driver version
filesystem type
Xfce version
number of installed packages and what are the differences
number of running processes and what are the differences
memory footprint
init system
Artix
6.0.7-artix1-1
nvidia 340.108
xfce 4.16
number of installed packages 794
number of running processes - Htop - Tasks: 96, 356 thr, 88 kthr; 1 running
OpenRc
Devuan
6.0.0-4-amd64
nvidia 340.108
xfce 4.16
number of installed packages 3183
number of running processes - Htop - Tasks: 103, 133 thr, 84 kthr; 1 running
SySVinit
Last edited by deepforest (2022-11-25 22:07:21)
Offline
What I see is Artix has a smaller installation footprint, smaller memory footprint, fewer running processes and a custom build kernel. You should look into what optimizations the Artix team put into their kernel.
Please remember to mark your thread [SOLVED] to make it easier for others to find
--- How To Ask For Help | FAQ | Developer Wiki | Community | Contribute ---
Offline
Ok. Thank you.
Offline
What I see is Artix has a smaller installation footprint, smaller memory footprint, fewer running processes and a custom build kernel. You should look into what optimizations the Artix team put into their kernel.
They say, that is all OpenRC magic
Offline
They say, that is all OpenRC magic
Interesting.
I did convert a Debian8 image to openrc a few years ago to try and access the difference and didn't notice anything significant, seemed 'equivalent'. That image is now 10 I believe, has run on a handful of boxes as a hypervisor, and still nothing worth noting noticed. I also have the sysvinit version and it has proven more stable up times, but the init is not the only variable, so no conclusion.
Offline
Thank you for a lot of interesting posts. :-)
I am beginnig to realize how difficult it is to answer all the good and relevant questions.
Maybe it is a bit like asking Toyota about the specific setup in James Petersens Corolla in Texas.
I was thinking Xfce more like a leadership for Xfce distros so they knew everything from feedback.
However it is just not that advanced at all, is it.?
I now see it more like the Corolla. If James Petersen wants to ruin the performance and good look he can just do it without asking or telling anyone, and there is not a thing Toyota can do about it. They likely won't even know it except if they make a strong investigation.
Then a bit similar I/we can ask Xfce to investigate 80 distros and tell us all about them.
I am not saying it is wrong or not relevant. I just want to bring my understanding and thanks to the administrator for keeping up the good work trying to complete an impossible mission. :-)
Maybe we could get things rolling faster if the administration here asks this page to do a full investigation:
https://distrowatch.com/
Vinifera
I can't answer your question but maybe Xubuntu can.
In their installation you can choose between a full installation and a lean one.
Offline
I think its an issue of focus. Upsteam Xfce focuses only on Xfce. Downstream distro developers focus on implementation and are free to adjust default Xfce settings as well as configurations that have nothing to do with Xfce (kernel, init system, etc). Its not feasible to ask Xfce developers to consider/test on all distros and combinations of configurations.
Please remember to mark your thread [SOLVED] to make it easier for others to find
--- How To Ask For Help | FAQ | Developer Wiki | Community | Contribute ---
Offline
if i make a distro that uses Xfce, i could set it up however i like. i would likely want to change the splash images (where you see the "Xubuntu" name and the circle portion go round and round) to something else. but i could make other changes, too. they might be consistent with the theme of my distro. or they might not.
i did make a little distro back in the 32-bit days. it was a CD that would boot on either an Intel i386 machine or a Sun Sparc 32-bit machine (like the Sparc 5/85 i had back then) and bring up a LFS Linux i made (i ported LFS to Sparc myself because no one else had done so by then). it fit on a mini-CD and i kept one in my pocket at work (we had a mix of those machines).
Offline
[ Generated in 0.018 seconds, 9 queries executed - Memory usage: 635.6 KiB (Peak: 668.45 KiB) ]