jimi@h-three.UUCP (jimi) (07/05/90)
In article <2099@esquire.UUCP>, zuena@esquire.UUCP (Bob Zuena) writes: > In article <JJC.90Jun18150723@yquem.jclark.UUCP>, jjc@jclark.UUCP (James Clark) writes: > > The documentation on the ditroff DESC file describes the unused > > `sizescale' parameter as the `scaling for fractional pointsizes'. Are > > there any versions of troff that support fractional pointsizes? > > [stuff deleted] > I'm curious as to whether an application exists which would require type sizes > in gradations finer than 1/72 of an inch -- I realize that is not the point > of Mr. Clark's posting, but my curiosity got the better of me. Many typesetters/imagesetters have a capability to produce type at fractional pointsizes. It would be nice if ditroff could support the capability on such devices. At least a couple of DTP programs provide support for fractional point sizes. PageMaker 4.0 can support type size/leading specs to 1/10 pt; the review I cite (MacWEEK 6/5/90) calls this "sufficient for most publication work." Note that a font's point size refers to the measurement from the top of the ascender to the bottom of the descender, with a small amount of space above and below, or the measurement from baseline to baseline with no leading. It is not an absolute measurement of character size. Depending on a typeface's x-height (the height of a lowercase x), the apparent size of typefaces of the same specified point size can vary widely. For example, Garamond has a small x-height and Helvetica has a large x-height. Helvetica looks much larger than Garamond at a given point size. A 1/2 pt difference in Helvetica results in a significantly different appearance. In Garamond, a 1/2 pt difference is less apparent, but still visible. > > Does anybody think it would be useful for a troff to support > > fractional pointsizes? > IMHO, no. Yes, for at least two reasons. First, ditroff should support as large a subset of an output device's capabilitiies as possible. Second, fractional point sizes offer greater flexibility in copyfitting. For example, some text might fit in the available space at 6/7 pt (type size/leading), but not at 7 pt unless set solid. It's likely that 6 1/2 pt type would fit in the available space with room for a 1/2 pt to 1 pt of leading. In the case of a typeface such as Helvetica, the legibility gains would be important. In summary, fractional point sizes are available on many output devices, they allow typographers greater flexibility, and they can provide readers with greater legibility. The lack of fractional point sizes is one of ditroff's few vestiges of the now-obsolete output devices available during troff's early development. This limitation should be remedied. How that is best done is another discussion. -- Jim Ingram uunet!h-three!jimi | jimi%h-three@uunet.uu.net h-three Systems Corporation P.O. Box 12557 RTP NC 27709 919 549 8334