Khyber's Journal: I caught NVIDIA being absolutely LAZY.

I have a nice shiny new HP Pavilion DV9825nr laptop. It came with Vista. As we all know, Vista is horrible.

Vista died in two days, leaving me with a brick. Ubuntu was an option, but I wanted XP back. So I go about with slipstreaming SATA and network drivers to install XP. I get XP installed, everything runs, except video.

Say that again? Video not working? It's an NVIDIA 8600M GS. All of NVIDIA's cards use pretty much the exact same driver architecture, so newer driver revisions support newer and older cards.

Apparently, NVIDIA wanted to promote the 8600M GS as Vista-only. They never included an entry in the .INF file so the card could install under Windows XP. After a little figuring out how .INF files were created, I modified the INF included with the GeForceGo drivers and the card installed and ran perfectly under XP

With this sort of laziness spotted, I wonder if NVIDIA has any real plans for the future. To be so lazy as to exclude your product from a still-viable market is marketing nonsense. Has marketing taken over NVIDIA or have the developers and engineers gotten too lazy?

Laziness, incompetence, or interference from MS

[bersl2]

Take your pick.

Shenanigans like this are increasingly making me want to take my programming closer to the hardware. For instance, I listened to the video [linux.org.au] of this talk from linux.conf.au 2008 [linux.conf.au], and if what is portrayed is typical of hardware people, they need people who understand software (especially FOSS) in addition to understanding hardware microarchitecture, even if they don't know it.

It seems like hardware companies need severe beatings too, though I know everything they do is all about business.

I perso

Re:

[Khyber]

You know what I wish I had, now that you mentioned PCI/PCMCIA? PCI-USB adapter. Do those even exist? Little box with a PCI slot inside, and a cable to go to USB port?

Back to NVIDIA. I'd definitely give some credit here to MS interference, because knowing that NVIDIA's drivers are backwards and forwards-compatible (only to some degree for the latter,) and multi-os compatible (within the Windows family except Vista which has the new driver architecture,) the drivers should've detected the card and installed i

Re:

[bersl2]

You know what I wish I had, now that you mentioned PCI/PCMCIA? PCI-USB adapter. Do those even exist? Little box with a PCI slot inside, and a cable to go to USB port?

USB 2.0 devices don't even have the theoretical maximum bandwidth to support a PCI connection. Nor does FireWire-800 or below, which is probably more suited to this sort of thing, since it can do DMA.

However, it does appear that there is such a thing as external PCI Express, so that appears to be a possible way to do this (along with a PCIe-PCI bridge).

Re:

[Khyber]

So how do we have USB sound cards if there's not enough bandwidth? USB 2.0 is what... 480 mbit? I'm thinking of this for things like a sound card, network card (Say your built-in network dies and you need to reinstall - have a USB Network card (already done) for installation. 480mbit - 100 mbit is enough for sound card, plus mouse, plus webcam, plus a slower hard drive.

Now the real question would be does USB 2.0 have the POWER to handle a PCI bus device such as a Sound Blaster LIVE!????

Re:

[bersl2]

PCI, as built into most desktop computers, is a 32-bit parallel bus which runs at 33.33 MHz, plus out-of-band signal lines. USB is a complicated serial bus with much protocol overhead. Even if you only consider the 32 data pins, that's 133.33 MB/s, which is more than double the 60 MB/s theoretical max for USB 2.0. You may be able to get some devices to work.

USB devices use that bus natively; they do not send information in the same way that PCI devices do.

As for electrical power, USB carries 5V 500mA, while