I think it was actually Shaw who said that. Who's Allan Wechsler?
Someone who invented a brilliant notation (I use Extended Weschler Notation as a standard for canonising small GoL patterns): From: Allan Wechsler Date: Wed, 30 Sep 1992 12:10-0400 Subject: Who called that HWSS a [27deee6]? Date: Tue, 29 Sep 1992 17:17 EDT From: John Conway Mainly to ACW - I loved your observation about the + and o phases of traffic lights in the early universe! By the way, I notice that nobody is taking up this term, so let me add some words about it: i) I've been using it for some time ii) It really feels psychologically "right" - one finds oneself thinking of this as happening a long time ago. iii) The "middle universe" is quite a nice term for the middle density case - it need not carry any temporal suggestions - though it can. iv) it doesn't, I grant, contain its own definition. Here iv) is negative, but not too much so. But the alternatives, such as "sparse universe", "thin universe", sound very colorless. I have been noting down life configurations in octal for quite a long time. So the trouble is that the speakers of this language have hitherto been solipsistic! Let's get everyone talking it. JHC Well then, let's get a formal description down in the archives: Life patterns that fit in fairly small rectangles can be described by short character strings as follows. Each column is represented by one character. Each row is assigned a "weight"; the bottom row has weight 1, the next weight 2, the next 4, the next 8, and so on. A column is assigned a score by adding the weights of the live cells in that column. The character that represents the column is then the nth character in the following (0-origin) string: "0123456789abcdefghijklmnopqrstuvwxyz", although encoding columns more than 4 rows tall is deprecated. (In other words, we deprecate the use of characters to the right of "f".) To mark representations of this sort, we recommend enclosure in square brackets. For example, [153] is a male glider and [532] a female one. We usually prefer the bounding rectangle of the pattern be chosen, although there are times when this is unwieldy; for example, "A blinker has two phases, [070] and [222].". Here the bounding rectangle has been chosen to retain a coordinate frame between phases. This issue should seldom arise; the recommended use of this notation is to describe single patterns, usually for purposes of naming: "We call the p2 oscillator [318c] a `beacon'.". Some notes on canonicalization. Even assuming that the bounding rectangle is chosen, there are still, in general, 8 orientations to choose from. We recommend putting the rectangle in "landscape" rather than "portrait" orientation, so that it has no fewer rows than columns. Then choose the orientation with the numerically smallest representation. When describing an oscillator, pick the phase with the smallest number of rows, breaking ties by choosing the numerically smallest representation. Thus, the blinker is [111] rather than [7]; the glider is shown in male phase [153] rather than in female [163]; the eater is [178c]; and the toad is always shown resting [1332] rather than panting [2994]. And, oddly enough, we prefer the more populous but squatter phase of the HWSS, [27deee6], to having to use big digits in [21hh197].