> We'll never know how Eyetech found the Teron instead - perhaps
> looking through a catalogue of PPC evaluation boards until
> something came up that was similar to Escena's planned board.

...or by performing industrial espionage on bplan in order to learn what they were developing the Pegasos from ;-)

> Alan [...] warning people of the buggy Articia - before
> he adopted a board with it.

This recollection is flawed. Eyetech's warning started only *after* they adopted the Teron PX, with the allegedly fixed A660BNGE, as the AmigaOne XE, as opposed to the older A660BNGP part used on the Pegasos (and the Teron CX / AmigaOne SE). bplan managed to keep their northbridge a secret for a long time prior to the Pegasos (betatester) release (and even disputed to use the Articia S when inquired about it by Nathaniel "downix" Downs). It was only after Eyetech abandoned Escena and adopted the Teron CX as AmigaOne SE that bplan revealed they were using this same chip and that it was buggy (yes, bplan/MorphOS people were the first to informally disclose warnings of the chip, which got discussed on ann.lu at the time, not Eyetech, albeit Eyetech were the first to issue a formal press release weeks later, mentioning "a small problem" with the chip and usage of the chip "with fix"), thus requiring the April fix developed during the following months.