Linked by Thom Holwerda on Wed 20th May 2009 13:27 UTC
Hardware, Embedded Systems Hot on the heels of the Russians, we have another clone maker popping up, this time in fish & chips country: Freedom PC. "Powerful and versatile, environmentally friendly yet inexpensive computer systems compatible with any and all of the main operating systems: Mac OS X, Linux or Windows. So YOU can decide which one to use for what YOU want to do. And we give you a choice of models, too - from the low priced and good looking office machine, the ideal choice for business, to the high powered, sleek, gaming media centre. All, with the operating system of your choice pre-installed - or none at all - at prices accessible to all." They offer various models pre-installed with Windows, Linux, or Mac OS X.
Thread beginning with comment 364807
To view parent comment, click here.
To read all comments associated with this story, please click here.
fsck
Member since:
2005-07-06

already answered, if you do nothing to the machine, then nothing will slow it down. Just like my win2k3 VM. If you mess with it all the time, then it will get slower as crap accumulates. All three questions answered.

I dont see how "crap" accumulates in a system that removes all but text config files when uninstalling applications.

It's not just the registry, it's also drivers installed by some app, and not removed by that apps setup program (clearly not Windows fault)

Really? I dont have this problem on Linux, because it supports my hardware natively. No bs with scattering drivers with faulty install software all over the place and configs I cant remove easily.

drive fragmentation (it happens, defrag that drive once in a while)

First of all everything fragments - that's a desperate argument and not one fixable with software (but with flash drives when they take over). Secondly EXT3 and EXT4 have near legendary status in terms of preventing fragmentation. The problem is horrendous on ntfs. Sure it's in no way unique to windows but much worse than nearly any other file system.

too many startup programs

You can just turn them off on linux. a lot of software on windows is extremely insidious the way it integrates with the os (not Microsoft fault however they do make it easy to do so though hooks for example anti virus software)

Theoretically, man can't fly either, but I am flying next week.

Technically you're in a machine that can fly. That's just a ridiculous statement.

Theoretically, nothing can travel faster than light, but quantum entanglement shows that information can.
Just because it's not accepted with classical thinking does not mean it is false. I was pointing out the lack of basis(in argument) for your assertions with no reasoning behind how it is possible.


A lot of apps do more than just add or change some config files. and sometimes, changes to config files don't go away when you uninstall an app. a networking app or an update can change the duplex setting of the NIC, instant slowdown.

Which you can manually change without massive obfuscation like in Windows. Where some even problems are simply unfixable without just starting over (which is quite absurd when you think about it).

Or, a new version of xorg comes down, and because of new features, it's a bit slower than the old version. Things like this can and do happen.

So you literally have to resort to bugs (or regressions and hence mistakes - human error) as an example of how Windows is not unique in slow down situations? I think that proves my entire point.

One thing I do know for sure, is that the latest kernel available for lenny, 2.6.26, seems a bit slower than the one that was distributed when Lenny was first released. I can see the difference in speed when I boot the older kernel (2.6.24, I believe) and watch the system start up, and then again in normal usage. (I have bootsplash turned off).

There is a specific regression between 2.6.26 and 2.6.28 that cause this. It has already been fixed. Has anyone fixed the slow down in windows 95-98-me-2000-XP-Xp64-2003 ? No.

I disagree, because my Lenny desktop at home proves it. The thing was damn fast when I first installed it, and after a couple of years, it is visibly slower.

Likely the previously mentioned regression which has been fixed is the cause (there is an article on Phoronix about it if you're interested

My Ubuntu desktop went the same way, only much quicker. I've seen it on fedora core too, but it's been a while since I used it, so I couldn't tell which version.

I'm currently(and primarily) a fedora user fyi. Not experienced that.

Linux doesn't work like that, you scatter files and stuff all over, not as bad as windows, but it's not neat and tidy like one dir == one app.
That is why I said mac. Mac is not Linux. As mentioned before linux does not have this problem either due to effective package management.

I understand how the system works, do you?

You don't seem to know about the differences in file systems, mention already fixed bugs as a problem while ignoring endemic problems in software design that has been available (and unfixed) for years(registry, no effective software management system) and seem to be confusing Mac and Linux.

I am speaking from personal experience, but you can go ahead and discount it, just because you haven't experienced it yourself.

That's not my intention. I'm saying: If there is a problem explain where it is, what is causing it and how it is possible because it doesn't seem to be one that effects more than a small minority. For all I know it could be anything from poor hardware choices to faulty ram to some huge architectural problem but nothing verifiable is given to back up any statements - completely unfounded.

Perhaps you only surf the web and chatter on messenger, if that's all you use it for, then you might not notice any slowdowns either.

I'm currently writing a book, I play games every now and then(modern), I used to be an IT Consultant and do a lot of software development, dvd authoring, I'm currently on a graphic design course and have a lot of high end audio applications. I do other things too and if you had commented on Linux being deficient in some of those areas I would certainly agree but there seems to be nothing behind your statements other than just claims.

You haven't offered any concrete reasons why it couldn't happen, other than some muttering about only touching config files and that it "theoretically couldn't happen"

Except for the whole system of installation/uninstallation and isolated storage of config data I suggest you read my post again.
Ah so you're using the "Prove god doesn't exist" argument? I can see i've wasted my time then.

Edited 2009-05-21 17:29 UTC

Reply Parent Score: 1

BluenoseJake Member since:
2005-08-11

I dont see how "crap" accumulates in a system that removes all but text config files when uninstalling applications.


explain that to the 3 or 4 xfce apps that were still installed on my system, even though I uninstalled xfce a few months ago. If linux package management is so perfect, why didn't they go away when I removed xfce? I installed fxce using apt-get install xfce, and I removed it using apt-get remove xfce. I had to remove them myself.

Even the documentation on Debian's website states that that is how you remove packages, but it is not a complete uninstall. This is because they were dependent packages, and were added automatically by apt, but because they were separate packages, they were not automatically removed.

Are you trying to tell me that isn't leaving crap behind? Nice try though.

I'm currently(and primarily) a fedora user fyi. Not experienced that.


So why are you even talking to me? At least I have experience with the OS's I am talking about. I have already stated that I use Debian Lenny, and I use Suse, and FreeBSD for fun and profit, and have used Ubuntu and Fedora in the past. I used Ubuntu on my laptop, and have for 3 or 4 years. I use the things I am commenting on.

There is a specific regression between 2.6.26 and 2.6.28 that cause this. It has already been fixed. Has anyone fixed the slow down in windows 95-98-me-2000-XP-Xp64-2003 ? No.


Considering I was talking going from 2.6.24 to 2.6.26, I guess that means that the rest of your statements are just useless also? If you aren't going to read my posts, why should I, or anybody bother giving you any credibility?

You don't seem to know about the differences in file systems, mention already fixed bugs as a problem while ignoring endemic problems in software design that has been available (and unfixed) for years(registry, no effective software management system) and seem to be confusing Mac and Linux.


Ok, let's handle this one at a time.

Differences between filesystems:

Other than mentioning having to defrag Windows, and that Windows didn't have a dedicated swap partition, how am I seemingly confusing filesystems? I obviously know the difference between NTFS and the various Linux FSs. Really, either explain your arguments or just stop arguing, but I have offered a few examples, and other than your erroneous statements about the infallibility of Linux package management, and about a regression in a kernel that has nothing to do with our discussion, you have offered nothing but generalities and BS.

Mentioned already fixed bugs:

Already dealt with, let's keep going...

ignoring endemic problems in software design:

what that a few megs of registry can slow down a dual core with 1/2/4/8/16G of ram? Get real, it's a pretty small database, any reasonable computer should see no slowdowns with the registry. Get over it. I suppose you recommend all those bogus registry cleaners and memory optimizers too?

and seem to be confusing Mac and Linux:

I don't know how you got that, but whatever, a fanboy's got to have something to grab a hold of.

Except for the whole system of installation/uninstallation and isolated storage of config data I suggest you read my post again.
Ah so you're using the "Prove god doesn't exist" argument? I can see i've wasted my time then.


WTF? I read your posts several times, but all I see is somebody, unlike myself, who really doesn't use Windows, and has no idea how far the platform has come since win98. But like a lot of Linux geeks, they still think that the Windows world is stuck in 1998, and continues to spout the old tired crap.

Reply Parent Score: 2

fsck Member since:
2005-07-06

Even the documentation on Debian's website states that that is how you remove packages, but it is not a complete uninstall. This is because they were dependent packages, and were added automatically by apt, but because they were separate packages, they were not automatically removed.

That is true, but as they aren't in use they have no effect on speed of the system except when accessing the package database (searching, adding or removing software). Both in theory and in reality. Nice try yourself ;)

So why are you even talking to me? At least I have experience with the OS's I am talking about. I have already stated that I use Debian Lenny, and I use Suse, and FreeBSD for fun and profit, and have used Ubuntu and Fedora in the past. I used Ubuntu on my laptop, and have for 3 or 4 years. I use the things I am commenting on.

I've been using Linux since SuSe 5.2, I've used mandrake since 7.0 Red Hat since 7.3, I've used Arch for many years, Gentoo and others. I don't see why you're just making up random ideas about what I have used with no basis? Fedora is currently my main system but that doesn't mean I use(or have used) nothing else. Please try to be objective.

Considering I was talking going from 2.6.24 to 2.6.26, I guess that means that the rest of your statements are just useless also? If you aren't going to read my posts, why should I, or anybody bother giving you any credibility?

Lol? this is quite amusing....you ignore what I said then tell ME i'm not reading your post? Wow, just wow.

I'll put it in bullet points so it's easy to understand this time:
* You upgraded to 2.6.26 from 2.6.24 to 2.6.26.
* 2.6.26 through 2.6.28 have a regression that impairs performance
* You upgraded to a kernel with a performance regression, it sucks, i'm sorry.
* It has been fixed in newer kernels.
* The older versions of windows were never updated to fix any such problems in any of them (this was a main point I made before)

I don't know how much more clearly I can repeat this. It is basic English.

Ok, let's handle this one at a time.

Differences between filesystems:

Other than mentioning having to defrag Windows, and that Windows didn't have a dedicated swap partition, how am I seemingly confusing filesystems?

I didn't say you were confusing file systems. As said above you need to read what I have typed to reply to it. I said you do not seem to know the difference between the file systems as you were mentioning fragmentation as if it is equal among all. It is not. Besides....it wont matter in the future flash drives perform the same regardless of fragmentation.

FYI i was saying you were confused between Mac and Linux as you responded to a comment I made about the Mac as if I said Linux. I did not.

Also....You can fix the swap problem by setting a registry key to make it delete the swap file on shutdown and recreate it on start up - avoids the fragmentation issue with swap files, mostly.

I obviously know the difference between NTFS and the various Linux FSs. Really, either explain your arguments or just stop arguing

As said above I do not know a way of being more clear than the previous direct examples. If you choose to read something other than what I write I cannot help that.

but I have offered a few examples, and other than your erroneous statements about the infallibility of Linux package management
I never said it was infallible, you just did. I pointed out why it was better, you relied on the argument "it's not perfect" and I never claimed that. Try again.

and about a regression in a kernel that has nothing to do with our discussion

It is directly relevant because it was concerning the kernel you upgraded to - the same one that caused a slow down on your machine.

you have offered nothing but generalities and BS.
Specific examples are hardly generalities.

Mentioned already fixed bugs:

Already dealt with, let's keep going...

ignoring endemic problems in software design:

Software design is of course subjective...granted but the results are clear. But database fragmentation is a reality, and the registry is a database....like it or not it is true. Don't believe me? http://en.wikipedia.org/wiki/Windows_registry

what that a few megs of registry can slow down a dual core with 1/2/4/8/16G of ram? Get real, it's a pretty small database, any reasonable computer should see no slowdowns with the registry.

Database fragmentation recently brought down the sql server hosting windows 7 RC requests. This is a much larger scale database of course but database fragmentation effects the registry also and causes much of the slowdown. Microsoft have said themselves this has been a problem. See this is where I feel I'm hitting my head on a brick wall. Lack of knowledge about how databases work is just one example here.

I suppose you recommend all those bogus registry cleaners and memory optimizers too?
Only the ones published by Microsoft ;) I'd definitely recommend hijack this(trend micro) for start-up programs and spybot search and destroy for the obvious though.

and seem to be confusing Mac and Linux:

I don't know how you got that, but whatever, a fanboy's got to have something to grab a hold of.


You said
Linux doesn't work like that, you scatter files and stuff all over, not as bad as windows, but it's not neat and tidy like one dir == one app.

When I used the word Mac at the start of the sentence specifically talking about Macs, but you answer with "Linux". They are not the same. Nice try though.

WTF? I read your posts several times

Not enough seeing as you misread almost everything you quoted, clearly.

but all I see is somebody, unlike myself, who really doesn't use Windows

I'm posting this on windows and I have to use it daily for the obvious audio software, Photoshop (though, this, specifically does run though wine) and various games. See the funny thing is you seem to have this hugely biased view of me when in reality I can concede many windows advantages (mainly applications) such as Visual studio. Nothing on Linux has a patch on it, nothing, not even close. But you'd rather just write off everything I say with some personal bias because you didn't understand anything I wrote, I really despair at this.

and has no idea how far the platform has come since win98. But like a lot of Linux geeks, they still think that the Windows world is stuck in 1998, and continues to spout the old tired crap.

I think I have a good idea. XP-64 is alright on the whole, but i'm not a fan of Vista. I think you'll find that's quite common among the tech crowd though. If we're speaking on a technical or architectural level I think it's a mess(as demonstrated by my previous points) but it works for something quick, easy with a huge breadth of application support. A computer is just a tool for most people anyway.

A final note:
In future please leave your "oohhh you're a fanboy" BS out of replies unless you actually understand the words I have used to respond. There's no need to be childish, its just a friendly debate on a website.

Edited 2009-05-21 22:24 UTC

Reply Parent Score: 1

aahjnnot Member since:
2008-07-24

"I dont see how "crap" accumulates in a system that removes all but text config files when uninstalling applications.


explain that to the 3 or 4 xfce apps that were still installed on my system, even though I uninstalled xfce a few months ago.
"
This was historically a problem with most Linux package managers - in installing XFCE, you hauled in a load of dependencies that wouldn't automatically be removed when you removed XFCE in case they'd subsequently become important for other software (possibly hand-installed) on your system. Newer versions of Ubuntu and, presumably, Debian advise you to run apt-get autoremove if you want to recover the space, but that message is only visible to command line users.

Ubuntu 9.04 takes this further by giving you a GUI Janitor program that looks for redundant software and makes recommendations for removal. That seems to me to offer the best of both worlds - intuitive, easy and relatively risk free.

Reply Parent Score: 1

Hoodlum Member since:
2009-05-22

Considering I was talking going from 2.6.24 to 2.6.26, I guess that means that the rest of your statements are just useless also? If you aren't going to read my posts, why should I, or anybody bother giving you any credibility?


You didn't even read what he said....

I don't know how you got that, but whatever, a fanboy's got to have something to grab a hold of.

Just huge bias...all over your whole post

all I see is somebody, unlike myself, who really doesn't use Windows, and has no idea how far the platform has come since win98. But like a lot of Linux geeks, they still think that the Windows world is stuck in 1998, and continues to spout the old tired crap.

And here is the crux. Typical bitter fanboy retort. When someone presents a rational argument with examples and facts to back it up you cry foul. People have different experiences, just because yours wasn't the same as others doesn't mean suddenly they eat babies.

I just had to sign up to comment on this - reminded me of a tabloid journalist.

Reply Parent Score: 1