There is a poster on Reddit criticizing dnf as being too slow.
I do not use Reddit.
But here is my question.
The terminal emulator is running in wayland. Right ?
So is it fair to compare that to a distro or DE not using wayland ?
Obviously, I have no clue what I am talking about.
Also, if dnf speed is important to you, should you not reduce unnecessary processes or do it in the tty ?
D.L.
On Sat, 2020-04-18 at 15:33 -0500, David wrote:
There is a poster on Reddit criticizing dnf as being too slow.
I do not use Reddit.
But here is my question.
The terminal emulator is running in wayland. Right ?
So is it fair to compare that to a distro or DE not using wayland ?
Obviously, I have no clue what I am talking about.
Also, if dnf speed is important to you, should you not reduce unnecessary processes or do it in the tty ?
I seriously doubt that the effect of the terminal emulator on dnf speed is even measurable. The whole thing is dominated by network I/O and local processing of package files.
poc
On Sat, 18 Apr 2020 at 19:02, Patrick O'Callaghan pocallaghan@gmail.com wrote:
On Sat, 2020-04-18 at 15:33 -0500, David wrote:
There is a poster on Reddit criticizing dnf as being too slow.
I do not use Reddit.
But here is my question.
The terminal emulator is running in wayland. Right ?
So is it fair to compare that to a distro or DE not using wayland ?
Obviously, I have no clue what I am talking about.
Also, if dnf speed is important to you, should you not reduce unnecessary processes or do it in the tty ?
I seriously doubt that the effect of the terminal emulator on dnf speed is even measurable. The whole thing is dominated by network I/O and local processing of package files.
poc
I also use Ubuntu and Debian distros. The "apt" tool is quite a bit faster than "dnf" on the same hardware, and they have aptitude, a text-mode front-end that is fast and gives quick access to package descriptions and can download change logs with a single keypress.
On 2020-04-19 07:28, George N. White III wrote:
On Sat, 18 Apr 2020 at 19:02, Patrick O'Callaghan <pocallaghan@gmail.com mailto:pocallaghan@gmail.com> wrote:
On Sat, 2020-04-18 at 15:33 -0500, David wrote: > There is a poster on Reddit criticizing dnf as being too slow. > > I do not use Reddit. > > But here is my question. > > The terminal emulator is running in wayland. Right ? > > So is it fair to compare that to a distro or DE not using wayland ? > > Obviously, I have no clue what I am talking about. > > Also, if dnf speed is important to you, should you not reduce unnecessary > processes or do it in the tty ? I seriously doubt that the effect of the terminal emulator on dnf speed is even measurable. The whole thing is dominated by network I/O and local processing of package files. pocI also use Ubuntu and Debian distros. The "apt" tool is quite a bit faster than "dnf" on the same hardware, and they have aptitude, a text-mode front-end that is fast and gives quick access to package descriptions and can download change logs with a single keypress.
I suppose more digging is actually required..
Is the speed difference due to network/mirror access or local processing? Does apt cache data/info for a longer period than dnf or in a more efficient manner?
Rather than comparing dnf to apt I think it is more important, if the speed irritates one, is to identify the steps dnf is taking and which of those steps are more or less responsible.
Analogous to "systemd-analyze blame".
On Sun, 19 Apr 2020 at 01:14, Ed Greshko ed.greshko@greshko.com wrote:
On 2020-04-19 07:28, George N. White III wrote:
On Sat, 18 Apr 2020 at 19:02, Patrick O'Callaghan <pocallaghan@gmail.com
mailto:pocallaghan@gmail.com> wrote:
On Sat, 2020-04-18 at 15:33 -0500, David wrote: > There is a poster on Reddit criticizing dnf as being too slow. > > I do not use Reddit. > > But here is my question. > > The terminal emulator is running in wayland. Right ? > > So is it fair to compare that to a distro or DE not using wayland ? > > Obviously, I have no clue what I am talking about. > > Also, if dnf speed is important to you, should you not reduceunnecessary
> processes or do it in the tty ? I seriously doubt that the effect of the terminal emulator on dnfspeed
is even measurable. The whole thing is dominated by network I/O and local processing of package files. pocI also use Ubuntu and Debian distros. The "apt" tool is quite a bit
faster than "dnf" on the same
hardware, and they have aptitude, a text-mode front-end that is fast and
gives quick access to
package descriptions and can download change logs with a single keypress.
I suppose more digging is actually required..
Is the speed difference due to network/mirror access or local processing? Does apt cache data/info for a longer period than dnf or in a more efficient manner?
Rather than comparing dnf to apt I think it is more important, if the speed irritates one, is to identify the steps dnf is taking and which of those steps are more or less responsible.
Analogous to "systemd-analyze blame".
That's a well considered response. One thing to bear in mind is that 'apt cache' part. Bear with me. So when we say 'dnf is slow'. The correct question is dnf is slow at what?
Because apt is not apt-get, nor is it apt-cache or even aptitude. (Synaptic anyone?) (apt-mark, and so on, lots of individual tools)
Each of the above, is a separate tool with it's own 'rules'. apt for instance, iirc doesn't honour apt-get blacklists. synaptic probably ignores both.
Now if you want info on a package. it's apt-cache show. where dnf is an all in wonder tool.
Also while i haven't looked at what happened for a while, doesn't dnf-dragora have a text based interface? (is that still the gui?)
So dnf is slow at what? :)
Also there are a lot of features dnf/yum have over apt that might introduce overhead, being able to 'step back in history' for instance.
On Sat, 2020-04-18 at 15:33 -0500, David wrote:
There is a poster on Reddit criticizing dnf as being too slow.
Compared with what?
-
Anthony F McInerney -
David -
Ed Greshko -
George N. White III -
Patrick O'Callaghan -
Tim