Reply
 
Thread Tools Display Modes
  #481   Report Post  
Posted to rec.audio.pro
Trevor Trevor is offline
external usenet poster
 
Posts: 2,820
Default DR-40 vs. H4n (now something completely different)


"Arny Krueger" wrote in message
...
Right I'm happy without what Microsoft calls "support", but you can't get
an OEM/non support copy without buying a new computer. Just licensing
sillyness IMO.


Not true in the US, as I prove with this link:

http://www.microcenter.com/single_pr...uct_id=0361167



I'm sorry, what is that proving again? That it costs far more than what Dell
pay, and that you do need to purchase it with a new computer by the terms of
the license.
Isn't that what I said already?

Trevor.


  #482   Report Post  
Posted to rec.audio.pro
Trevor Trevor is offline
external usenet poster
 
Posts: 2,820
Default DR-40 vs. H4n (now something completely different)


"Arny Krueger" wrote in message
...
There was a time when dealers (mostly the mail order houses) would well
an OEM copy of Windows by throwing a defunct disk drive or graphics card
into the package on the basis that they were selling it along with parts
to build a computer.


One time I went that route, and the qualifying hardware was a highly
obsolete and tiny CPU fan.


Which was fine then, but most dealers won't do that any more so I can only
assume Microsoft cracked down on them and threatened their supply.

Trevor.


  #483   Report Post  
Posted to rec.audio.pro
Anahata Anahata is offline
external usenet poster
 
Posts: 378
Default DR-40 vs. H4n

On Sat, 31 Mar 2012 21:35:13 +0200, Mxsmanic wrote:


However, the 64-bit address space will be exhausted just about as
quickly as the 32-bit space, mainly because of poor software
engineering. It's a bit like IPv6, which will be exhausted about as
quickly as IPv4, because of poor allocation techniques.


I haven't yet read a reasoned critique of IPV6's allocation scheme, but I
have a suspicion that the plan behind increasing IP address lengths by a
factor of 4 instead of (as you might expect) 2 was to make sure they
couldn't run out in any foreseeable length of time. 128 bits is a lot of
addresses, however you cut the cake.

Note that even Ethernet addresses (so called 'MAC' addresses) have had a
good run. Every piece of Ethernet kit ever made has a world-wide unique
address, with manufacturers allocated blocks by IEEE, and I haven't heard
any rumours they are likely to run out soon, and they are 'only' 48 bits.

--
Anahata
-+- http://www.treewind.co.uk
Home: 01638 720444 Mob: 07976 263827
  #484   Report Post  
Posted to rec.audio.pro
Mxsmanic Mxsmanic is offline
external usenet poster
 
Posts: 805
Default DR-40 vs. H4n

Anahata writes:

I haven't yet read a reasoned critique of IPV6's allocation scheme, but I
have a suspicion that the plan behind increasing IP address lengths by a
factor of 4 instead of (as you might expect) 2 was to make sure they
couldn't run out in any foreseeable length of time. 128 bits is a lot of
addresses, however you cut the cake.


They can easily run out. The problem is that engineers think in linear terms
when looking at address spaces, but the results are exponential.

For example, the 128-bit IPv6 address has already been divided into two 64-bit
fields for some purposes. Intuitively, the people doing this "feel" as if the
address space has been reduced by half, but it hasn't. It has been reduced by
more than 99.99999999999999999%. Continued allocation of fields within the
128-bit address will continue to reduce the address space exponentially until
there is nothing left. That's exactly the same mistake that was made with
32-bit addresses.

Engineers and others often forget that a linear reduction in address size is
an exponential reduction in address space. Encoding information into the
address itself dramatically restricts the number of addresses available.

I've seen software engineers exhaust huge virtual address spaces in the same
way. Often engineers think that an address space of n bits equates to
"infinite" and allocate virtual space serially and generously, and are then
surprised when systems crash because they've run out of address space.

Note that even Ethernet addresses (so called 'MAC' addresses) have had a
good run. Every piece of Ethernet kit ever made has a world-wide unique
address, with manufacturers allocated blocks by IEEE, and I haven't heard
any rumours they are likely to run out soon, and they are 'only' 48 bits.


The 48-bit addresses are expected to last until 2100 by the IEEE. This sounds
generous until you realize that a 48-bit address space assigned serially would
accommodate 281,474,976,710,656 Ethernet cards. The address space has been
severely (exponentially) reduced by linear allocation of address bits.
Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT +1. The time now is 07:00 PM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"