Linked by Thom Holwerda on Wed 26th Mar 2008 21:30 UTC, submitted by ohxten
Thread beginning with comment 306914
To view parent comment, click here.
To read all comments associated with this story, please click here.
To view parent comment, click here.
To read all comments associated with this story, please click here.
Yeah, what he said! There's way too much bloat in OSes these days -- for very little in the way of visible advantage. What's wrong with having a slick, bare-bones OS that zips along and lets you add stuff to it as you please?
I could do my daily tasks in DOS on a 386 all those years ago much faster than my Vista-laden Pentium-something-or-other can do them today. And this is progress? All you get in return is some fancy graphics. Big friggin' deal.
Who says an OS has to be one massive, complex beast of code with bleeding-edge 3-D graphical effects and every "feature" imaginable in order to make it past pure hobby status?
If you're not writing a desktop OS, then it can be streamlined. But desktop OSes need to have all the features that users expect these days.
Sure, it needs to look nice enough to to be able to comfortably use, but even that doesn't need the latest Quartz, Compiz or Aero with a recent graphics card. As long as it works, and it works good... that's what really matters. If it has the most important features, but not little-used ones, again... what's the problem? Smaller code size, leaner OS, still functional. Too many unused features are just bloat.
A common misconception. There is no magical list of features that are "bloat". Despite what lots of people think, if you remove the 20%/30%/50% of features that are least frequently used, you will lose a far higher percentage of your users.
I have yet to find a practical reason for an OS to explode in size like the OSes of today are, other than sloppier programming.
Features, features, features. As the features expand, programmers require more frameworks to keep up with development. Those frameworks require more resources. Sure if we all still developed in C and assembly everything would run a lot faster, but it is completely impossible to write complex software with such basic tools in a reasonable time frame. Sorry. Just not possible.
In conclusion, I think it's the other way around. The typical operating system these days is *too bloated*... if any OS manages to make its way out of hobby status while having good hardware support, having all the main features that are expected in an OS today, and still being lightweight, that would be great.
It would be great. But unfortunately completely impossible.





Member since:
2006-12-05
Who says an OS has to be one massive, complex beast of code with bleeding-edge 3-D graphical effects and every "feature" imaginable in order to make it past pure hobby status? Sure, it needs to look nice enough to to be able to comfortably use, but even that doesn't need the latest Quartz, Compiz or Aero with a recent graphics card. As long as it works, and it works good... that's what really matters. If it has the most important features, but not little-used ones, again... what's the problem? Smaller code size, leaner OS, still functional. Too many unused features are just bloat.
I have yet to find a practical reason for an OS to explode in size like the OSes of today are, other than sloppier programming. As processor speeds increase, it should be spent on optimization, to make the most of that extra power... not wasted by lazy programmers, making that extra power useless (yet at the same time, just to be able to run their POS program). In some of the worse cases, a new version of a program on newer hardware can still run slower than an older version of the same program, on *slower* hardware... with absolutely NO new features (at least, none that are of any use). Mostly just a shiny new resource-intensive GUI. That is beyond disgusting.
In conclusion, I think it's the other way around. The typical operating system these days is *too bloated*... if any OS manages to make its way out of hobby status while having good hardware support, having all the main features that are expected in an OS today, and still being lightweight, that would be great. From what I hear, Haiku appears to be approaching that point, and I hope they succeed.