ELBOX New Spider device

[Piru]

Spider is not yet available- we hope to get a scheduled Elbox order in the next few weeks.

So are these cards still rebadged NEC 720101 USB cards with a modified PCI ID? (They sell for $7 or so)
Does this particular driver erase RDB if you're unlucky?

[Darrin]

Spider is not yet available- we hope to get a scheduled Elbox order in the next few weeks.

OK, I'll wait until you get them in.  Put me down as a firm order.  I'll actually be in Wales again in November for the Swansea v Man Utd and the Cardiff v Forest matches, so if the stock comes in late I'll give you a Maesteg address to send it to or I'll pick it up in person.

[Heiroglyph]

So are these cards still regular NEC 720101 USB cards with a modified PCI ID?

You'd think they would have to be or else the old cards would stop working.

Do these drivers wipe out your RDB if you have faulty memory, a rogue app writes out of bounds or you try to fake the PCI ID?

Edit: Looks like he edited in the same question while I was posting.

[Darrin]

So are these cards still rebadged NEC 720101 USB cards with a modified PCI ID?

Probably.  A bit like MorphOS is a modified OS3.1

Does this particular driver erase RDB if you're unlucky?

Only if you're a pirate.

[amigakit]

@Piru
Harry, I am not sure- Ill let you know if we can get them in stock.  I think the third batch of cards we had a few years ago were another brand though, if I remember correctly.  I know the Spider in my A1200 cosmetically looks very different to the later Spider, in terms of layout and components.

@Darrin

It will be nice to meet you.  Swansea vs Man Utd should be interesting...

[Piru]

A bit like MorphOS is a modified OS3.1

If you're trying to find as bad analogy as you can you've succeeded.

The only way that'd work would be if Elbox would have studied some existing NEC USB card and would have created their own version with similar functionality, but with improved features. I'm afraid the Elbox contribution to the HW is the printed sticker.

Only if you're a pirate.

If only. Unfortunately AmigaOS has no memory protection, and as such the driver cannot tell a difference between a pirate and some application trashing innocent memory.

[Darrin]

You'd think they would have to be or else the old cards would stop working.

Do these drivers wipe out your RDB if you have faulty memory, a rogue app writes out of bounds or you try to fake the PCI ID?

Edit: Looks like he edited in the same question while I was posting.

As far as I'm aware the RDB "protection" issue has been over with and non existant for years and was confined to a particular driver.

[Piru]

@Piru
Harry, I am not sure- Ill let you know if we can get them in stock.  I think the third batch of cards we had a few years ago were another brand though, if I remember correctly.  I know the Spider in my A1200 cosmetically looks very different to the later Spider, in terms of layout and components.

Interesting. Do those card revisions have the same PCI ID?

Which actually reminds me: You could easily create a patch to the pci.library so that it'd report any NEC USB card as having the same PCI ID as Spider 2. Assuming the chipset is the correct one it'd work. This way the spider driver would stay untouched and RDB destruction wouldn't happen either.

[Darrin]

If you're trying to find as bad analogy as you can you've succeeded.

No, I'm just being childish like you are as you try and undermine Amigakit and Elbox sales.

If only. Unfortunately AmigaOS has no memory protection, and as such the driver cannot tell a difference between a pirate and some application trashing innocent memory.

So it is just as well that such drivers no longer exist then, isn't it?

[amigakit]

My Spider card is very old (circa 2002)- here is the PCIInfo output in the Amiga Shell:

Pci Device 9:
   VendorID: $1033
   VendorName: NEC Electronics
   DeviceID: $35
   DeviceName: udp9210 Dual OHCI controllers plus Single EHCI controller
   ClassCode: $c0310
   RevisionID: $43
   InterruptPin: 2
   SubsystemVendorID: $1033
   SubsystemID: $35
   MemSpace0: $80003000 - $80003fff

Pci Device 10:
   VendorID: $1033
   VendorName: NEC Electronics
   DeviceID: $35
   DeviceName: udp9210 Dual OHCI controllers plus Single EHCI controller
   ClassCode: $c0310
   RevisionID: $43
   InterruptPin: 3
   SubsystemVendorID: $1033
   SubsystemID: $35
   MemSpace0: $80004000 - $80004fff

Pci Device 11:
   VendorID: $1033
   VendorName: NEC Electronics
   DeviceID: $e0
   DeviceName: uPD720100A USB 2.0 Host Controller
   ClassCode: $c0320
   RevisionID: $4
   InterruptPin: 4
   SubsystemVendorID: $ee4
   SubsystemID: $89e
   MemSpace0: $80000100 - $800001ff

[Heiroglyph]

Has it been confirmed that it no longer has the RDB code in it?

That's my main worry and they would do well to put that officially to rest if they haven't already.

With Deneb gone, they are the only game in town, but I'd rather not worry about that particular problem.

[Darrin]

Considering that 100s of us use Elbox products with Elbox drivers every day then I think the answer speaks for itself.

I'd rather see Piru back up his statement by producing evidence that such code exists in the latest drivers.

I think the RDB stuff goes back to the first days when Elbox were guarding their drivers like the crown jewels and you had to get them on CD and then register their products in order to access any updates.  These days you can just download the latest drivers off their website with no checks.

[Piru]

I'd rather see Piru back up his statement by producing evidence that such code exists in the latest drivers.

Which statement is that?

[Darrin]

Does this particular driver erase RDB if you're unlucky?

and yes, I know the difference between "suggesting" and "stating", but we both know what you were insinuating so don't try and be clever.

[Piru]

I know the difference between "suggesting" and "stating"

Well it looked like as if you wouldn't have. Thanks for setting the record straight.

[Heiroglyph]

Piru and I weren't the ones caught putting potentially damaging software on the market.  It's their PR blunder.

It's a legitimate concern that I'd like to at least get confirmation on before paying 9x cost for a driver that is locked to one USB card.

If the Deneb died, I understood that replacement cost was for custom hardware.

It hurts to pay $80 for driver development (which I'm fine with BTW) and then have to pay again if the $10 hardware dies.

Adding the worry of the driver going haywire and trashing my RDB is just a step too far to ask.