September 24, 2003

Every so often, miracles happen.

And when they do, they always bring a sense of awe, and a warm fuzzy feeling of trust in a higher power to the unsuspecting onlooker.

I experienced such a miracle yesterday.

The higher power was Microsoft, and the awe was triggered by the fact, that after gently nudging them (i.e. bitching at them) for a few months now, finally, like with an all-encompassing light and a booming voice, they spoke to me.

(And for anybody who hasn't detected my sarcasm yet, I suggest a thourough head examination....)

Of course it was in regard to the "NMI: Parity Check Error" that has too often been the topic of monologue in this blog (like here, here, and here).

For all interested parties, here's the beat by beat from the MS engineer who got stuck with the task of investigating this (and who for his own safety, shall remain anonymous). It should be noted, that the message was graciously forwarded by Robert Scoble, who deserves a pat on the virtual back for keeping his promise and producing results. (Note: Despite requests for more general information, the MS response focuses solely on my particular problem with the NMI Parity Check error in conjunction with the Nikon LS 2000 Film scanner):

We've had troubles with this scanner because of the lack of Nikon offering XP drivers for the device back during the XP testing days.

I read his links [he means ME by that - never mind, that I'm not a "he"] and it looks like he was having several problems using Windows in general with Norton, Adaptec, Nikon problems. (Generally it's a rant describing a process that fails.) [You got that right, Buster. Rant it is, and fail it did.]

As for the Nikon LS 2000 issue - beware what you buy. I would not recommend using any legacy peripheral device for XP or post XP (including serial, parallel, and SCSI devices). The description about his SCSI problems are indicative of the SCSI experience using XP (in that Windows had no Adaptec drivers inbox and the using had to troll the Adaptec site for the correct drivers.

Also, the original mail came out only 4 months after XP released. I remember a lot of hardware vendors still were not up to speed (XP-driver-wise, WHQL-driver-wise) until the following summer.

Today, I found the following:

http://www.nikonusa.com/usa_tech_center/dtc_options.jsp?cat=1&grp=2&prod uctNr=9295

I'm sure here a user can get their Nikon film scanner working for this device on XP and be able to update the firmware for the device.

Accord to what I read, he gave up on his Nikon LS2000 a while ago. Today, I think he would have less trouble. We didn't have any trouble with SCSI for XP or Adaptec: remember that KB article that recommended users goto the Adaptec site? The rest was up to the vendor, Nikon. I don't see any XP problem illustrated here: I see vendors failing their customers 4 months after XP shipped.

If I've missed the point or something, let me know.

Uhhhmmmm ... sort of. The bit about not using any serial, parallel, or SCSI devices was certainly a revelation. Yet I have a flatbed scanner and printer daisy-chained off my parallel port, and they've been running with zero problems under XP. So far, anyway.

As far as the Nikon scanner goes - well, given that it took Microsoft 3 months to come back with the keen observation that Nikon is slow to keep up on their compatibility development, it smells to me a bit too much like the pot calling the kettle black.

The fact remains that I still get tons of email from distressed people in regards to this error, looking for information on how to solve the problem.

To make it easier for them from here on out though, I'll be dedicating a section of this site to collecting error reports, resources, and individual solutions in the hope that one day, a clear picture about the root of the error and its resolve may emerge. (I will be posting the link to the new section in this blog as soon as I'm done with building the page.)

Update 2:35pm: It's done.