hi
i’m a member of the italian forum of opensuse…
during the last days, a user on our forum asked us this thing:
“why opensuse compared to ubuntu, kubuntu or other linux distro seems to be slower and less reactive?”
we don’t know an answer and, since none of our users is a developer, we can only make some hypothesis.
steluciani wrote:
> hi
> i’m a member of the italian forum of opensuse…
> during the last days, a user on our forum asked us this thing:
> “why opensuse compared to ubuntu, kubuntu or other linux distro seems
> to be slower and less reactive?”
> we don’t know an answer and, since none of our users is a developer, we
> can only make some hypothesis.
And so you just took his word for it, without questions, asking for proves,
or (at least) trying it for yourself?
no, we speaked about that problem with a discussion of 3 pages on the forum… also other user have this problem… with opensuse and a single core processor like pentium4 or amd sempron…
i have a sempron 3100+ with 1gb ram and yes also in my opinion opensuse is a bit less reactive compared to ubuntu or fedora (for me just a bit, so it is not a problem, but for some users is more than a bit slower)
an example: it’s possible to see that compiz and kwin are less fluid or that firefox, konqueror, dolphin and openoffice loading take more seconds and an higher cpu % compared to ubuntu or fedora,
one of the hypothesis is the number of services that opensuse loads on startup, maybe more compared to other distributions…
steluciani wrote:
> no, we speaked about that problem with a discussion of 3 pages on the
> forum… also other user have this problem… with opensuse and a single
> core processor like pentium4 or amd sempron…
> i have a sempron 3100+ with 1gb ram and yes also in my opinion opensuse
> is a bit less reactive compared to ubuntu or fedora (for me just a bit,
> so it is not a problem, but for some users is more than a bit slower)
> an example: it’s possible to see that compiz and kwin are less fluid or
> that firefox, konqueror, dolphin and openoffice loading take more
> seconds and an higher cpu % compared to ubuntu or fedora,
>
>
> one of the hypothesis is the number of services that opensuse loads on
> startup, maybe more compared to other distributions…
There lies the problem to start with; it’s almost impossible to make the
same install with all the Debian-, Redhat- and SUSE derivatives out there
and be able to objectively compare them wrt speed issues.
I’m not saying *buntu isn’t faster in some areas, but you can’t draw any
conclusion before knowing much more about the DUTs.
be sure to uninstall or turn off beagle…that dog really chews up
the cycles…
and, openSUSE updater will also suck up the cycles from about 15
minutes after boot, and on for five to 15 minutes, depending on how
busy the update repos are…
> steluciani wrote:
>> no, we speaked about that problem with a discussion of 3 pages on the
>> forum… also other user have this problem… with opensuse and a
>> single core processor like pentium4 or amd sempron…
>> i have a sempron 3100+ with 1gb ram and yes also in my opinion
>> opensuse is a bit less reactive compared to ubuntu or fedora (for me
>> just a bit, so it is not a problem, but for some users is more than a
>> bit slower) an example: it’s possible to see that compiz and kwin are
>> less fluid or that firefox, konqueror, dolphin and openoffice loading
>> take more seconds and an higher cpu % compared to ubuntu or fedora,
>>
>>
>> one of the hypothesis is the number of services that opensuse loads
>> on startup, maybe more compared to other distributions…
>
> There lies the problem to start with; it’s almost impossible to make
> the same install with all the Debian-, Redhat- and SUSE derivatives
> out there and be able to objectively compare them wrt speed issues.
>
> I’m not saying *buntu isn’t faster in some areas, but you can’t draw
> any conclusion before knowing much more about the DUTs.
iirc saw this discussion somewhere else and it had to do with how the
kernel was compiled - which modules were included (openSuse was both a
server and desktop distro while *buntu was primarily desktop focused)and
which optimizations were passed to the compiler (supposedly *buntu’s
used more aggressive options). When I had tried kubuntu (on my previous
pc)I too felt it was snappier but didn’t feel it was as good.
thanks very much for the answer
i also think that opensuse is a better distro compared to ubuntu (also if it’s quite slower, yast is insuperable and the system is very stable)
I did some comparative tests with openSUSE versus Kubuntu and a couple of other distros. This was on a 64bit Sempron 3100+ processor with 512MB RAM. The versions of openSUSE were both 32bit, but Kubuntu was 64bit. On that system, I have never noticed much difference between 32bit and 64bit versions of the same distro.
My impression of openSUSE 11.1/KDE 3.5 versus Kubuntu 8.04/KDE3.5 was that after KDE startup, Kubuntu initialized with more RAM used (+100MB) than openSUSE.
Comparing openSUSE 10.3/KDE3.5 against the same Kubuntu as above, Kubuntu initialized with more RAM used (+135MB).
I believe that is because Kubuntu is pre-loading programs used by e.g. openoffice. This will shorten the apparent time taken to first load openoffice on kubuntu, so it would appear to be more reactive. However once openoffice is loaded on openSUSE, the difference in responsiveness between the distros is far less significant, and they all respond quickly.
If you pre-load RAM with programs for one application (e.g. openoffice) you are obviously depriving other applications (e.g a video player) that may need that RAM. Initially openSUSE is leaving more ram free for any application that may need it.
steluciani wrote:
>> I’m not saying *buntu isn’t faster in some areas, but you can’t draw
>> any
>> conclusion before knowing much more about the DUTs.
>
> sorry for my ignorance: what is the meaning of DUTs??