You mean real price in static dollars? Dunno know in aussie terms, but according to the Dallas Fed oil in US dollar terms has been higher. US$105 was the real price level equivalent reached in 70′s dollars. But even that was not the historical high in real terms.

It’s a terrific article

http://dallasfed.org/research/eclett/2008/el0805.html

Question 1 – Yes, ever since early 2007 when the price exceeded the 1981 peak and kept rising. They were then and still are higher than at any other time (at least any other time shown on FMark’s graph).

Question 2 – No. The price between 1982 and 2007 was at all times lower than its 1981 peak, but has been higher than that peak continuously since early 2007.

Moreover, I reckon you’d be a naive optimist not to expect that it will continue to be higher and increasing most of the time from now on. Hence the post about cycling.

Believe it or not I bought this house in 76 partly influenced by rising fuels prices – so its near to shops and train stations but not inner city.

No. See http://www.wtrg.com/oil_graphs/oilprice1947.gif

Has there ever been a time when oil prices weren’t higher than some other time in the past?

And aren’t they always at any time in the present higher then ever in the past?

You are in unjustified presumption territory. I just chose one aspect of the Falzon’s story to comment on that intrigued me.

I have used the industry leading Nuance Dragon Naturally Speaking speech recognition program for some years (not so often now because my keyboard skills have improved a lot from when I first started using it. Now I just use it for bulk transcription of hard copy text (e.g. for quotes, extracts for study guides) where it’s a nuisance to scan the text). I get better than 99% accuracy out of it, which you actually need to achieve otherwise it’s more of a nuisance than a help – 95% for example means one error every 20 words, and each one takes time to correct and retrain – it’s quicker to type with two fingers at that accuracy level. Thus Dragon’s performance is certainly better than the TV live captioning, but then again the latter has to transcribe many different voices whereas mine is trained and customised to my voice and vocabulary.

Your links seem to contradict each other. The second one describes several Linux speech recognition solutions, but they just about all seem to work on the old IBM ViaVoice for Linux SDK. However the first document you linked says “In the late 1990s, a Linux version of ViaVoice (created by IBM) was made available to users for no charge. However, the free SDK was later removed by the developer in 2002.” Thus it appears that the solutions described may no longer be available. Moreover I lack the IT skills to cobble something together; it needs to install simply and work out of the box (as Dragon does).

I have actually explored the feasibility of shifting to Linux a couple of times including quite recently. I might be tempted by it for a secondary machine, such as an ultra-portable like Nicholas has just bought e.g. the Asus eee PC apparently comes in default configuration with a Linux OS. I’m not yet tempted for my primary machine, not just because of Dragon but also because I think OpenOffice just isn’t quite as good as MS Office. Moreover I frequently produce visual presentations and Word documents for hundreds of students, most of whom don’t have OpenOffice but just about all of whom can cope with MS Word and Powerpoint

BTW: Is the software you currently use better than the “live” (actually a 5-10 second lag) subtitling on TV news/current_affairs?