Hello,

i own 2 acer notebooks,one i3 4000u,4 gig,500hdd, and 1 with celeron dual core 1.7ghz sandy bridge 2 gig ram, 500 hdd.

I have tried Debian-Ubuntu-LXLE-OpenSuse-Mint in both of them.
No tweaks, normal install,no tweaks-only percentage of swap file.
Although linux distros with the exception of ubuntu run at fewer
resources CPU-RAM than windows 7 and 10, they are less responsive
under load.Load for me means open tabs in chrome and eclipse used for
C++ and java projects for University.
Windows cpu-ram are close to full when i use both of them but it is still responsive.
Swap file is 8 gig.tweak at 10 percent of usage.
Fedora can not even run without lag.
Debian-Ubuntu unusuable after 3-4 chrome tabs.can not run Eclipse with not lag.it lags even when writing some times.
LXLE-Mint mate is ok for a few tabs meaning light html pages and eclipse,it hangs at quick switch between the apps,needs some time for the app to get responsive.
OpenSuse kde is exceptionally responsive almost like windows.strange for kde,i Keep opensuse installed as dual boot at the moment.
I can not understand why the bad responsiveness.I know that my notebooks are not so decent in terms of hardware but they are quite responsive in windows under load.On intel celeron i can have 10 tabs open with some youtube videos open while i run eclipse and its still ok.There are some minor lags here and there but they are quickly resolved.
Am i missing something?Or this is how it is supposed to be.
I can not find many tweaks for latest versions except the swap file one.
sry for bad english

I use lubuntu 15.10, which is fast and lightweight. If you want something more responsive, give it a try.
Avaiable here.
https://help.ubuntu.com/community/Lubuntu/GetLubuntu

There is a lot going on in your test and your question. Doubt we can offer guidance unless test is smaller and more facts are known.

In general, linux distro's are built to serve a wide array of users, and systems and programs. They do make a trade off in many cases because linux developers are forced to use secondary means to make programs work. For most users, things like responsive may be from a number of sources. You'd have to test for those bottle necks.

Problem with windows is the use of virtual swap. And random memory laid down in chunks with all over the drive. it will only get slower as you go unless things have changed.
and with linux the longer it runs the more cache is created and the faster it will become that is the theory. Windows when they used windows 3.1 with dos 5.0 was the last time they took it serious maybe the NT OS's but the facts are apples and oranges.

Hello there,

If you have time, please try Slackware and tell us about it.
Slackware has a reputation for being light on hardware, fast,
stable and you can configure KDE or XFCE to use very little
graphic effects.

I would say that depends on how the OS switches among processes. I think (and you need to go for it - if interested) Windows is tuned to be more user friendly, therefore its GUI is maybe more responsive. On linux X is just a process and probably the processes "in the background" run more smoothly.
You also need to check the bottleneck: is this the cpu, ram, swap, io, or ?? to understand it.

You are not testing the OS under load. You're testing your internet browsing speed.
To properly compare the OSs under load, you'll need benchmarking software designed for that purpose.

Chrome browser is owned by Google. It is possible that Chrome is tweaked so that it performs better with Windows.
Furthermore, Google may have agreements with sites like Facebook, Twitter etc so that using the SPDY protocol enables faster browsing speeds (Google_chrome#Speed. Such agreements may not be available for linux distros.

Have you tried Chromium and Firefox with linux to see if there's a better browsing experience?
Essentially, it seems since Google owns and influences a large part of the internet, then this enables them to gain greater internet browsing speeds with their proprietary browser.

Since linux uses less CPU and RAM, then objectively this means it will perform better under load.
All supercomputers have linux OSs - none of them have Windows.
Thus your browsing experience is more to do with internet browsers rather than the performance of linux under load.

i can try another 100s of distros and i will still face the same perfomance issues.
No matter how much fewer resources linux uses ,under some load it always stays behind.
I guess that perfomance increase is noticed mostly on servers rather than on desktop usage.
For desktop usage,stability and perfomance is lacking.
Maybe i will try arch.
Hint:decreasing swap usage makes things worse.i experience better overall responce with a swappiness close to 90
sry for bad english.

That is true,but it is not about the internet speed.Its the lag,bad responsiveness switching tabs and eclipse/other app in my situation.Spikes in CPU and Ram are under 90% so it should handle it better.
Servers are not desktops.I do not need to argue about anything,the reality is that for desktop usage ,linux lags seriously.
Internet browsers are vital for desktop usage and can be really heavy in resources.why do i need something more to benchmark?
I did this post as to see if there is something ,a tweak,setting that can improve responsiveness.

You are experiencing slower internet browsing speeds using Google Chrome with linux compared to Google Chrome with Windows.
This does not mean linux has a performance issue. An OS is not just an internet browser.

How do you know this? You have not put linux under load. The CPU and RAM usage is much lower than Windows.
Your internet browsing speed is to do with how the Google Chrome program works as opposed to the efficiency of an OS.

This is a computer program issue not an operating system issue.

That is also a linux distro.

This may indicate you need more RAM.

I could see m$-server or xbox vs GNU\Linux but only one is freely tweakabial... and out of those "100s of distros" how many 5, no tweaks?

You can test all the distros in the world. Until you find out what is causing the issue there is little to be gained by it.

Did you look at any metrics to determine why your issue exist.

As long as you want to convince yourself of that, it will probably remain true for you. But as you have not provided any actual measurements or test cases, it cannot be judged to be so by others.

That is only the reality between your own ears, and eyes, input device and screen.

I am typing from a 1.7Ghz Intel Celeron single core Toshiba laptop, 2GB RAM - what I call my perfect laptop running Slackware 14.1 - nothing could be much improved! And while it has been configured for my needs in many ways, that does not include any GUI performance tweaks that I recall (I use Fluxbox).

I currently have 9 tabs open in my browser (Firefox, uBlock Origin), 81, yes, eighty-one currently open/active terminals (Tmux'd, this is my all-day/every-day environment), multiple open remote sessions, a number of open files in Vim, listening to ELO comeback tour through VLC (video window minimized)... I have nothing to complain about other than my slow DSL internet!

With htop running: CPU 31% (7% if I shut down VLC), RAM 621MB used, 0% swap used of 2GB. No lag noticable in anything.

Well, you need something benchmarked in order for anyone else to make a remotely meaningful comparison.

Your dual core lags, my single-core does not... na-na-na-na-na-na! But that is utterly meaningless for comparison purposes between GNU/Linux distros or between GNU/Linux and any other OS without a common test case and a few metrics.

Benchmarks give us something to compare. If not an actual numerical benchmark, then at least some reproducible test case.

What you have stated so for is meaningful only to yourself, really.

I have not sat at a window$ machine for years so cannot offer a current comment, but I do recall that it had a different "feel" overall than my Linux machines of the same generation. The window$ machine (sometimes) felt smoother, even thoug it actually had much slower useful throughput. As I recall, the reason was that window$ was all about the GUI, and that GUI was designed for smoothness at the cost of performance, whereas the GNU/Linux machine had much better performance but could show small visual ticks between processes. I suspect that is less important today, although I really don't know.

Hello,

Shouldn't it be swapping to a partition?
cheers

it can be either