The macosxhints Forums

The macosxhints Forums (http://hintsforums.macworld.com/index.php)
-   Hardware and Peripherals (http://hintsforums.macworld.com/forumdisplay.php?f=21)
-   -   Alright, So USB 2.0 is garbage and........ (http://hintsforums.macworld.com/showthread.php?t=54612)

CAlvarez 04-19-2006 05:00 PM

Quote:

would they be better than the same devices standardized in the USB format? Yep.
No, they would not be better. They would be at best the same technologically, but cost a lot more. So much for $25 keyboards. And you'd need more FW ports, which also cost more. And you'd have to avoid mixing asynchronous devices like keyboards and mice which would drag down the performance of the isosynchronous devices on the same bus. Etc, etc.

Please try to understand the technology before you make blanket statements about it.

sulo28 04-19-2006 05:11 PM

Yes, but in the case of 802.11 and ethernet, overhead is to blame for the diminished throughput and eventual goodput. Could that be the case with USB? I suppose garbage is a bit harsh, but I've always been one to tell it like it is. If something doesn't live up to its advertised expectation, I tend to lash-out at the stupidity behind such claims. I don't blow sunshine up people's butts or cushion things just to appease other people. It is what it is. But you raise a good point; could overhead be a problem?

Unlike ethernet or WiFi, USB can be thought of as a closed system interface, meaning that you hook up a USB cable directly to a drive, and since it's not used for anything else, guarenteed delivery is expected. This of course excludes wireless USB mice because once communication leaves the confines of a guided medium, interference becomes a much larger problem, especially in the 2.4GHz region of the spectrum. However, for the sake of argument, we'll focus on wired USB devices since those are predominantly wired in nature.

hayne 04-19-2006 05:15 PM

Quote:

Originally Posted by sulo28
Yes, but in the case of 802.11 and ethernet, overhead is to blame for the diminished throughput and eventual goodput. Could that be the case with USB?

Yes.
That seems like a more reasonable explanation than magic. :)

sulo28 04-19-2006 05:15 PM

Quote:

Originally Posted by CAlvarez
No, they would not be better. They would be at best the same technologically, but cost a lot more. So much for $25 keyboards. And you'd need more FW ports, which also cost more. And you'd have to avoid mixing asynchronous devices like keyboards and mice which would drag down the performance of the isosynchronous devices on the same bus. Etc, etc.

Please try to understand the technology before you make blanket statements about it.

Thats why this post was in the "Help Requests" section. I'd like to understand it, thus the questions/statements. Please, by all means correct me; but do so in a civil way.

sulo28 04-19-2006 05:16 PM

Quote:

Originally Posted by hayne
Yes.
That seems like a more reasonable explanation than magic. :)


Thanks hayne. A simple explanation and some intelligent conversation was all I was looking for. Take care.

hayne 04-19-2006 05:24 PM

Giyf
 
Googling for: USB overhead
or: USB throughput
will find many useful articles explaining what you seem to want to know.
For example, look at this technical white paper on USB 1.0: http://www.usb.org/developers/whitepapers/bwpaper2.pdf

styrafome 04-19-2006 08:25 PM

Quote:

Originally Posted by sulo28
Yes, but in the case of 802.11 and ethernet, overhead is to blame for the diminished throughput and eventual goodput.

Even that example is much more complicated than is stated there. Ethernet may have to contend with overhead, but 802.11 also has to contend with noise and interference from other completely unrelated devices, range issues, radio blockage by obstacles... If someone wanted to protest USB for not reaching its full throughput, then to be proportional about it, they would have to start a major public campaign against how 802.11 often fails to reach 50% of the throughput advertised by Apple and everyone else.

It could be that companies advertise the ideal not just because it sounds better, but because it's simple. Ads would have to say "USB port could be really fast or really slow, depending; and FireWire port is usually a lot faster, most of the time, but not with a slow drive, etc..." I'm not sure how much of an improvement that would be.

CAlvarez 04-19-2006 11:41 PM

Quote:

Yes, but in the case of 802.11 and ethernet, overhead is to blame for the diminished throughput and eventual goodput.
Once again, I urge you to learn the technology before making such assumptions. Ethernet signaling at a simple level is not that different from USB (remember, I said SIMPLE level). I don't know what exactly you mean by "overhead" nor do I think you really do, but if you understood ethernet I doubt you'd make the statement above.

I'll use Ethernet vs. Token Ring as the example, simply because I understand it to a far deeper level than USB/FW details, but the same principles apply. Token Ring is like a well-moderated debate, where everyone waits to be called upon. Everyone must agree to this, and must wisely negotiate the process and how others may join. Firewire is like this also. Ethernet is a crowded room where people shout randomly, and if two people shout at the same time, both wait a random time and shout again, hopefully not colliding again. There is no intelligence or agreement. Therefore, Ethernet (USB) is far cheaper to implement and produce, and it also works extremely well when used properly. USB does a lot of things very well, and indeed, could do most things FW can do but with more demands on the system and user.

You definitely would not want a world with only FW, for the reasons I noted before regarding a FW keyboard or mouse.


All times are GMT -5. The time now is 11:33 AM.

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2014, vBulletin Solutions, Inc.
Site design © IDG Consumer & SMB; individuals retain copyright of their postings
but consent to the possible use of their material in other areas of IDG Consumer & SMB.