Hi decoy,
I agree that the wording of the KB is quite silly. The underlying reason of why we don't have comprehensive error codes listed in our header files is due to the interactions among many different software layers involving internal components from NI that don't always have a customer-facing header file.
The error code you received is coming from an internal platform abstraction layer (NI-PAL) that is sitting between IMAQdx and Microsoft's firewire driver. In general, many error codes that this layer returns are translated into IMAQdx errors before being returned by the API. However, not all of them are translated, usually because they are unknown/unexpected or possibly new.
Now, when you are using LabVIEW, each product group at NI (including for internal components such as NI-PAL) installs an error description file into LabVIEW that allows it to translate errors from any NI product. This lessens the impact of non-translated error codes for LabVIEW customers. As far as I know, we don't have such a component available under CVI or C to achieve similar functionality. Thus, when an error code is returned by an underlying component that IMAQdx doesn't know how to translate, it will return it directly and unfortunately I don't know of a programmatic way in CVI to decode that error.
To address your original problem, I think you are just simply running out of RAM. When acquiring from firewire cameras, IMAQdx is going to want to allocate DMA'able, contiguous physical memory as large as your image size. With that stingray returning a somewhat large 1MP image, it is quite possible the OS is not able to find enough contiguous pages to create a ring of several of these buffers if you only have 1GB of memory total in the system. Windows maintains various pools of different types of memory and while you might have plenty of normal memory available, it might be getting exhausted of the type needed by the firewire driver.
Hope this helps,
Eric
