On 2013-01-14, at 11:25, Groepaz wrote: > he didnt do it in "realtime" (as in: decode on the fly in a single revolution) > - he did pretty much the same as AR did (and everyone else really). I found one more claim similar to that, possibly coming from a less "controversial" source: http://www.c64.com/scene_display_interview.php?interview=178 "I de-coded a [w]hole block in real-time during reading and writing. Not first read in the buffer and then de-code." It comes from a guy who is the author of (faster than Dolphin) Prologic DOS but I still can't see how could this be done. Last night I wrote a new routine completely from scratch, using 2KiB of tables and I seem to get only marginally faster than what is inside 1571, still needing 40+ cycles in case of four out of five GCR bytes (after equalising). I disassembled also quickly DolphinDOS 2 ROM and I seem to have found some rather big tables inside but didn't extract and analysed the actual routines. Especially that Julian recently wrote he could possibly have a look at it. -- SD! Message was sent through the cbm-hackers mailing listReceived on 2013-01-14 12:00:59
Archive generated by hypermail 2.2.0.