Consumer Reports on iPhone; Apple snags another Delicious staffer

[AppleInsider]

[ View this article at AppleInsider.com ]

[melgross]

Too bad about Firewire. According to old expectations, 3200 should have arrived almost two years ago.

Right now, it's too little, too late. Both SATA, and the upcoming USB 3 will relegate it to a minor role.

[ajmas]

Originally Posted by melgross

Too bad about Firewire. According to old expectations, 3200 should have arrived almost two years ago.

Right now, it's too little, too late. Both SATA, and the upcoming USB 3 will relegate it to a minor role.

It all depends what you use Firewire for. SATA still uses the CPU as the controller, whereas Firewire has an independent controller. This is important for systems that are already loaded doing other things. At the same time SATA is only good for storage, whereas Firewire does a whole lot more.

As fore the Delicious Monster developers being sucked up by Apple that sucks, for DM. Delicious Library is a great tool and every time they lose someone it means that the development gets delayed. As a shareware developer there isn't really much in the way of icentives to keep someone from accepting a job at Apple.

[Louzer]

Originally Posted by ajmas

It all depends what you use Firewire for. SATA still uses the CPU as the controller, whereas Firewire has an independent controller. This is important for systems that are already loaded doing other things. At the same time SATA is only good for storage, whereas Firewire does a whole lot more.

Firewire might do a lot more, but very few people care about that (we are on a Mac forum, not discussing the latest video rendering hardware or something).

On the mac, Firewire is generally (if not solely?) used for external storage connectivity and some video cameras (unless you count the excitement that is Firewire networking). So that's how most look at it. Unfortunately, even now, because firewire requires, as you say, its own controller, firewire devices tend to cost more than USB devices, and, as such, will stay as a limited option.

For PCs, it's worse, since most desktops don't come with the port, if or they do, it's the slower 400 type.

But it doesn't matter. It's doubtful Apple will implement it (they barely care as it is) and it would be better if they went eSATA anyway, which is more likely to become the defacto standard (as USB is the defacto standard now).

Add on to that Apple's wonderful decision to only allow a select few machines (MacPros and MBPs) to have the ability to add in PCI cards to gain new functionality (most people won't even be able to take advantage of it for years to come, anyway).

BTW, when they say "the same port", I'm assuming they mean same as in "Firewire 800", not "firewire 400". Or perhaps the Firewire 4-pin port.

[Ben Huebscher]

Why doesn't apple just buy delicious monster and get the whole damn staff and another solid piece of software for its portfolio?

[JeffDM]

Originally Posted by ajmas

It all depends what you use Firewire for. SATA still uses the CPU as the controller, whereas Firewire has an independent controller. This is important for systems that are already loaded doing other things. At the same time SATA is only good for storage, whereas Firewire does a whole lot more.

I'm pretty sure SATA transfers are done by DMA. It really shouldn't take much CPU, I really doubt that Firewire's more CPU efficient to make any difference in anything. Even FW800 has a speed bottleneck compared to SATA. An eSATA link has one less drive controller as well, and it looks like eSATA enclosures are getting to be less expensive because it doesn't require that controller.

The only other thing done with consumer use of Firewire is DV deck control, and that's going away as DV camcorders are declining in prominence. They're being pushed out by hard drive, optical and solid state recorders.

[ajmas]

Originally Posted by JeffDM

The only other thing done with consumer use of Firewire is DV deck control, and that's going away as DV camcorders are declining in prominence.

What are they using for transfer instead?

[JeffDM]

Originally Posted by ajmas

What are they using for transfer instead?

USB. It's now going file-based rather than tape-based, so you can just copy the file rather than deal with a real-time import. If it's a recorder that records to DVD, the owner can just put the disc in the optical drive to import it.

[melgross]

Originally Posted by ajmas

It all depends what you use Firewire for. SATA still uses the CPU as the controller, whereas Firewire has an independent controller. This is important for systems that are already loaded doing other things. At the same time SATA is only good for storage, whereas Firewire does a whole lot more.

Face it, Firewire is dying. One big problem with it has always been the payment oriented license for it, which has sucked the life out of it.

Other than for camcorders, some hi end Tv uses, and occasional special purpose functions, there us really no more use for it.

In addition, it is unreliable, causing data loss when upgrading OS's. I've had two Firewire drives go on me after carefully folowing all of Apple's procedures, and I know of others who have had the same problem.

Sata is the future drive standard. Get used to it!

I've converted all of my Firewire externals to eSATA, and so has almost everyone else I know. Apple would do well, and offer at least one eSATA connection on every machine they make. I certainly hope they do so this MacWorld.

[Zandros]

Originally Posted by melgross

Face it, Firewire is dying. One big problem with it has always been the payment oriented license for it, which has sucked the life out of it.

Other than for camcorders, some hi end Tv uses, and occasional special purpose functions, there us really no more use for it.

F-35 Lightning II

I've converted all of my Firewire externals to eSATA, and so has almost everyone else I know. Apple would do well, and offer at least one eSATA connection on every machine they make. I certainly hope they do so this MacWorld.

The one thing eSATA doesn't do is provide power. Can it handle daisy-chaining?

/Adrian

[hmurchison]

Firewire 3.2Gbps is important. I don't think anyone is expecting it to have a chance at supplanting USB3 but for what it does it does it well enough.

Peer to peer connections
Low cpu utilization
Isochronous communication
IP networking
cable options for long runs.

Firewire is worth it for those who want premium offerings and that's why I choose Mac. So that I'm not getting just the status quo.

Well now I know why Delicious Library is delayed. You can't have most of your team poached and maintain a good release schedule over the years. Wil I'm ready when you are bud. Just get'r done when you can.

[hmurchison]

Originally Posted by Zandros

F-35 Lightning II



The one thing eSATA doesn't do is provide power. Can it handle daisy-chaining?

/Adrian

eSATA provides no power and has a 15 device port multiplier. It's a one trick pony . With Intel adding USB3 to motherboards eSATA becomes the redundant connector on motherboards due to its narrow focus.

[JeffDM]

Originally Posted by Zandros

F-35 Lightning II

That only serves to reinforce his point. Firewire's already limited usefulness to the consumer is fading. eSATA's taking away Firewire's high performance external drive niche and USB is good enough for the home video recorder niche.

Originally Posted by hmurchison

eSATA provides no power and has a 15 device port multiplier. It's a one trick pony . With Intel adding USB3 to motherboards eSATA becomes the redundant connector on motherboards due to its narrow focus.

But USB 3 doesn't exist yet, we're apparently talking two to three years away at an unknown price, and it's using a new optical cable too, there's some potential for problems and added expense. What I really like about eSATA is that there's no question of performance, it's going to work exactly as if it was plugged in internally, there's fewer interface converters to potentially monkey the works.

[hmurchison]

Originally Posted by JeffDM

But USB 3 doesn't exist yet, we're apparently talking two to three years away at an unknown price, and it's using a new optical cable too, there's some potential for problems and added expense. What I really like about eSATA is that there's no question of performance, it's going to work exactly as if it was plugged in internally, there's fewer interface converters to potentially monkey the works.

USB 3.0 should be released by summer 2008. Intel demo'd it at IDF 2007. I believe they'll be including 3.0 support in their ICH10 South Bridge.

Hell I'll take all 3. eSATA/USB 3.0/FW 3.2Gb

Let's not rule out the darkhorse as well in Ethernet being an increasingly more popular storage connection.

[ajmas]

Originally Posted by JeffDM

But USB 3 doesn't exist yet, we're apparently talking two to three years away at an unknown price, and it's using a new optical cable too, there's some potential for problems and added expense. What I really like about eSATA is that there's no question of performance, it's going to work exactly as if it was plugged in internally, there's fewer interface converters to potentially monkey the works.

In many ways while eSATA has plenty of advantages, the main issue is that it is internal connector trying to add extra value on the outside. USB, Firewire while not being the race horses of storage data transfer offer the advantage of being good for several things. This is an important factor when you consider that portables and other compact computers are where general computing is going. Not being able to daisy chain devices is a bit of limitation.

For high performance computing your tower with its expansion slots is what you want. There you have no space limitations and eSATA finds a good home, though the irony is that by the time you have a tower you have internal space for drives anyhow.

[JeffDM]

Originally Posted by hmurchison

USB 3.0 should be released by summer 2008. Intel demo'd it at IDF 2007. I believe they'll be including 3.0 support in their ICH10 South Bridge.

Hell I'll take all 3. eSATA/USB 3.0/FW 3.2Gb

Let's not rule out the darkhorse as well in Ethernet being an increasingly more popular storage connection.

The Ars Technica story I read said that the spec should be finished mid-08 and compatible devices appearing in '09 or '10.

Ethernet is good if you need to access the information from several systems, though that's how I do it, I share one Mac's drives to other computers, but the local connection is still native.

Originally Posted by ajmas

In many ways while eSATA has plenty of advantages, the main issue is that it is internal connector trying to add extra value on the outside. USB, Firewire while not being the race horses of storage data transfer offer the advantage of being good for several things. This is an important factor when you consider that portables and other compact computers are where general computing is going. Not being able to daisy chain devices is a bit of limitation.

Needing to daisy chain suggests that you're dealing with non-portable situation anyway. I'm hoping for cheaper multiplier bays. All my USB drives do daisy chain if that's what I need.

The only other real consumer use for FW is camcorders and that's dead-end too. So I'm not sure what Firewire can do or what other consumer devices benefit from it. So I'm just not seeing it. I still do use FW for my backups and my legacy HDV camcorder, but any future purchaces will wipe out both uses.

For high performance computing your tower with its expansion slots is what you want. There you have no space limitations and eSATA finds a good home, though the irony is that by the time you have a tower you have internal space for drives anyhow.

The internal space might not be enough for me. I have to go external for backups as well, I'd like to have mirrors go external.

[melgross]

Originally Posted by Zandros

F
The one thing eSATA doesn't do is provide power. Can it handle daisy-chaining?

/Adrian

You can put a number of drives in one case if the case has conglomeration. It's not really a problem.

As to power, you're right, but even there, solutions are beginning to appear.

One of the most important part of the specs though is that these are native drives, no translation required, which has been the bane of Firewire's problems, and is responsible for many of the corruption problems.

An advantage other than that is that with one eSATA connector, and a four rive case with conglomeratin, you can get up to a practical limit of 350 MB/s.

Try that with any Firewire combo, including 3200 when it comes out. It just won't happen.

[melgross]

Originally Posted by hmurchison

Firewire 3.2Gbps is important. I don't think anyone is expecting it to have a chance at supplanting USB3 but for what it does it does it well enough.

Peer to peer connections
Low cpu utilization
Isochronous communication
IP networking
cable options for long runs.

Firewire is worth it for those who want premium offerings and that's why I choose Mac. So that I'm not getting just the status quo.

Well now I know why Delicious Library is delayed. You can't have most of your team poached and maintain a good release schedule over the years. Wil I'm ready when you are bud. Just get'r done when you can.

I used to love Firewire. But, let's face it, Apple, and it's partners dropped the ball.

Considering that this was, more than anything, Apple's initiative, they were much too redolent about applying it to their own machines, and the association's continued insistence in charging per socket has also kept it down.

Also, the continued problems with adapters has haunted it.

[melgross]

Originally Posted by hmurchison

eSATA provides no power and has a 15 device port multiplier. It's a one trick pony . With Intel adding USB3 to motherboards eSATA becomes the redundant connector on motherboards due to its narrow focus.

I don't agree. USB 3 will never become as efficient as a drive interface. And, SATA will also get faster. Remember that USB 3 won't be available widely until 2010. A lot can happen before then.

[melgross]

Originally Posted by hmurchison

USB 3.0 should be released by summer 2008. Intel demo'd it at IDF 2007. I believe they'll be including 3.0 support in their ICH10 South Bridge.

Hell I'll take all 3. eSATA/USB 3.0/FW 3.2Gb

Let's not rule out the darkhorse as well in Ethernet being an increasingly more popular storage connection.

10Gb/s Ethernet is a reality, but is currently far too expensive for the likes of us. When I was still working professionally, sure.

But, it will drop just like all other Ethernet standards did before. By the time USB 3 is really here, in a useful way, 10 Ethernet will be much cheaper, though not yet as cheap as some other services.