doc/sphinx/techref/UniqueID.rst
.. epigraph:: | C makes it easy to shoot yourself in the foot; | C++ makes it harder, but when you do, | it blows away your whole leg."
-- Bjarne Stroustrup
| From: Thomas Phinney of Adobe | Subject: [OpenType] Why you don't need UniqueIDs | Date: Fri, 04 Mar 2005 10:04:41 -0800 | | Neither Type 1 nor OpenType CFF fonts require PostScript UniqueIDs. | | Back in the mid-80s, when printing might be done on a 57K serial connection | and printers might have 8 MHz processors, the caching of device-generated | bitmaps enabled by UniqueIDs made a noticeable speed difference. With today's | connection bandwidths and printers, the printing speed difference was | insignificant in our tests (conducted around 2002), and certainly not enough | to outweigh the risks of collisions between UniqueIDs for different fonts and | the trouble of tracking the ID numbers. | | For these reasons, Adobe stopped using UniqueIDs (and XUIDs) in our own | OpenType CFF fonts. If we still made Type 1 fonts, we wouldn't use UniqueIDs | for them, either. | | You may of course continue to use UniqueIDs, it's just that they are not | necessary. | | Regards, | | T | | Thomas W. Phinney | Program Manager | Fonts & Core Technologies | Adobe Systems