This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ComputingWikipedia:WikiProject ComputingTemplate:WikiProject ComputingComputing articles
This article is within the scope of WikiProject Brands, a collaborative effort to improve the coverage of brands on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.BrandsWikipedia:WikiProject BrandsTemplate:WikiProject BrandsBrands articles
Besides the tiny display and limited disk capacity, a less obvious Osborne shortcoming was the use of memory-mapped i/o chips. I believe these chips were from the 6502 family that was fully used (with the 6502 CPU) in the Apple II.
Memory-mapped i/o in general and the 6502 family in particular were poorly documented in the CP/M world, so it was much harder for amateurs to write and debug software in assembly language or BASIC for I/O through the Osborne's non-standard chips. Although the competing KayPro is often thought of as a reverse-engineered Osborne clone, in fact the KayPro used I/O chips from the Zilog-80 family that used conventional I/O addressing. These chips were well-documented and much easier to program without a Ph.D. in electrical engineering.
Yet another Osborne shortcoming was that serial i/o only worked reliably up to 1200 baud (bits per second). In an office setting where 5-1/4" and 8" floppy disks coexisted and LANs were still years off, file transfers via serial ports were often necessary to move files between computers with incompatible disk formats. At 1200 baud, moving the mere 90K bytes of a full Osborne single-sided/single density floppy disk to a computer with 8" floppies took 12.5 minutes, plus overhead for disk reads and writes. KayPros and other CP/M computers with Zilog chips could run their serial ports at at 19,200 baud, 16x faster!
DaveM, Los Angeles
The serial chip in the Osborne 1 was a Motorola 6850 ACIA. This did not have an on-chip bit rate generator, but could be switched between 300 and 1200 baud by writing different values to the on-chip clock divider. The user manual documented a change that allowed a jumper to be moved, giving 600 and 2400 bits/second rates - 600 was useless, but 2400 was handy when 2400 modems started appearing. Back in the day I don't recall any problems getting terminal software for the Osborne and there were overlays for popular programs such as Modem7 and MEX. There was a very nice terminal program called OTERM which had an attractive split-screen feature and on-screen help (rare in those days). Probably a bigger limitation than memory-mapped I/O was the fact that the I/O was also bank-switched. This required considerably more understanding of the system to get any serial communications to work - you couldn't just find another overlay and tweak the port addresses, you had to write bank-switching code. Interrupt handling was challenging, as I recall. The floppy disk controller was a Western Digital 1793 (I think...one of that family, anyway. This [1] says it was a Fujitsu second-soure of the 1793) but did not include an on-chip double-density data separator. Double sided drives would have been very nice, too...available in the aftermarket, but not from Osborne Co. I recall reading at the time that Osborne considered single-sided drives to be more reliable, but I speculate the real reason was lower cost. The double-density "kit" for the Osborne was a rather clumsy overlay for the main board, as was the "80 column" video board. Strangely enough, the Osborne didn't use the ubiqutious 6845 video controller. The parallel printer port used a Motorola 6821 - this could be made bi-directional for use as an IEEE 488 controller, though I never met anyone or read of anyone who did this. --Wtshymanski (talk) 16:53, 2 June 2009 (UTC)[reply]
