Hi Jean. On Tue, 2005-12-20 at 20:52 +0100, Jean Delvare wrote: > You seem to think that using more than one bit for sign is odd from > National. This is actually the standard way of doing though. Think > about it. [...] Now that you say it: you're absolutely right. And as a result: > Your patch is probably correct but also needlessly complex. All we > really have here is a unsigned vs. signed issue. What about this more > simple fix instead? Tried your patch, with a test program I wrote (testing against the positive and negative temperature values given in the LM77 data sheet) as well as with a live system (only with positive temperature readings, though). The patch works fine, thanks. Bye, Mike
