“Support for a faster version of USB in Linux is imminent and will become a permanent part of the Linux landscape when the next version of the operating system, 2.4.19, is introduced.” Article at News.com.
“Support for a faster version of USB in Linux is imminent and will become a permanent part of the Linux landscape when the next version of the operating system, 2.4.19, is introduced.” Article at News.com.
Keep USB 1 and use scsi/FireWire when you need something better.
—
http://islande.hirlimann.net
I still haven’t figured out how to install the last version of USB in linux. It’s no good to have these fancy features if people can’t use them. The Linux guys have GOT to focus 80% on fixing their broken ease-of-use problems, with the other 20% on apps.
It took me 25-30 hours (no exaggeration) to set up my firewall in RedHat 7.2. That is retarded. I want to try Gentoo out, but I fear having to install the firewall again. And I’ve been using Linux off and on for years, mostly off, since Windows et al. has always been “good enough”.
Linux is useless without users. There will be no corporate support, large-scale apps, masses of developers, mindshare gains, etc. until there is a sufficient mass of users. And until the system isn’t a pain in the ass with no reward but 37 text editors to choose from, Linux will be a >1% marketshare niche system, vulnerable to Microsoft’s attacks and economic woes (see the death of LWN.net).
And people say “Linux is open source, so it can’t die”, but they forget that when *everybody* ditches Linux in favor of a proprietary easy to use OS, there will be no more Free Software Community writing drivers and fixing bugs. And believe me, NOBODY can do it all by themselves. So YES, linux can die. And will – unless they fix the problems which are causing linux users to flock to Mac OS X like rats leaving the Titanic. And I’ll be on OS X with the fleeing millions, waiting for SOMEBODY to fix what’s still broken in Linux before I return. No more suffering for me.
Ludovic
Sounds like a good idea to me, but, who is going to convice the device makers to use firewire/SCSI instead of USB2. I mean, it is great to say not to use USB2, but quite another to tell me to not use USB2 despite less options and higher priced peripherals. USB2 and firewire are both emerging standards, so, it is not too late, but, with new systems from Compaq and Dell etc all shipping with USB and no firewire, the future does not look bright.
“from Compaq and Dell etc all shipping with USB and no firewire”
Good point. Another good point for Firewire is that sony fully supports it. Here in Europe most digitall from sony have a good reputation (I don’t like sony), so this could push the firewire standard.
Server side all compaq’s and dells’ have used in production where scsi equipped.
—
http://islande.hirlimann.net
Firewire has a stronghold that’ll keep OEM putting a port in their PC : every DV camera has a Firewire connector. If an OEM want to make a “multimedia” PC, it has to have Firewire
Keep USB 1 and use scsi/FireWire when you need something better.
Exactly… USB and FireWire form a dichotomy, USB for low bandwidth input devices such as mice, keyboards, and tablets, FireWire for higher bandwidth devices.
USB2 has no real world advantages over FireWire, whereas FireWire still has many over USB2. USB2’s higher bandwidth seems to be a theoretical advantage, but Firewire has been beating it in the benchmarks.
The real problem is that now there exists another division between a high performance product and a commodity product. USB2 is seen as a commodity product, is seeing a great deal of system bundling (thanks to what, Intel’s recommendation?) and is becoming widespread. FireWire, with theoretical equal cost, will be more expensive because it is not seen as a commodity product and will thus move less merchandise and will therefore have to charge a higher price.
I hate economies of scale.
Disclaimer: In my company all desktops(except mine) run RH 7.x, all firewalls run OpenBSD, and most servers run FreeBSD, and all of them run some kind of packet filter(ipf/pf/ipfw/ipTables). ~30 boxes in total, not a single MS program on them. At home I run FreeBSD, Debian, Plan9 and a few other things…
It took me 25-30 hours (no exaggeration) to set up my firewall in RedHat 7.2.
It takes 15 min max to setup a new OpenBSD firewall!
(of course it can take weeks/months/years to came up with the exact rule set you want, but that is a completely different matter, that have nothing to do with what OS/PF you use).
IIRC RedHat 7.x even let’s you setup the firewall on install time(with a very nice and useless GUI to select the services you want to allow and so on), so to setup a firewall isn’t much more difficult than just doing a normal RH install…
And, of course, if you have so much panic to use Vi, you can try something “easier”, like: http://www.fwbuilder.org/ I have never used it, but seems very popular.
Now can you explain to me how I setup a fire wall on a NT box without using (expensive)3rd paty tools? NT packet filter capabilities are absolutely useless in any real wold scenario, not to mention that to use the GUI to setup anything is a pain.
In short: either you are just ignorant, or you are spreading FUD, in any case, please be more careful before making this kind of completely misleading statements.
Thanks
\K
P.S.: Yea, I know, I’m a “MS mole” because I told somebody that to do something in Linux is easy! and it’s actually true! OMG, the sky is falling!
P.P.S.: More on topic: I’m the only one surprised to hear so much fuss about USB under Linux? USB1.x has been working fine for years now, and I have yet to see anything that uses USB2.x, which is supported in 2.4.19 any way…