Endeavor

I found this old manual for a 1993 PC Workstation called the Epson Endeavor, which claimed to be distributed with a utilities disk that enabled a 100x31 text mode in 1-2-3.

How could such a thing be possible?

I was able to find the drivers on an old ftp site. Sadly they didn’t work, presumably they needed a real Cirrus Logic VGA card, but the fact that these drivers exist at all made me curious. It seems plausible that 123 can internally generate an arbitrary sized canvas, and it simply asks the display driver what resolution it wants?

I know that dosemu can simulate arbitrary text mode resolutions, and some DOS software is able to use it. Is it possible that I could write a dosemu display driver, and 123 would just work in a fullscreen xterm?

History

Lotus 1-2-3 was ported to a bunch of systems, including OpenVMS, Xenix, and even System/390. In 1991, Lotus released a version for SunOS4 on SPARC.

A few years later, a Lotus engineer uploaded a patch for a file management bug in the XALERT component that affected the 1.2 update. Luckily, they forgot to remove the STABS data, leaving incredibly rich debugging data.

typedef short int lmbcs_type;
typedef short int platform;
typedef short int memory_type;
struct envblk_ { /* size 68 id 9 */
  short int envsize; /* bitsize 16, bitpos 0 */
  short int pfid; /* bitsize 16, bitpos 16 */
  short int csid; /* bitsize 16, bitpos 32 */
  short int pad; /* bitsize 16, bitpos 48 */
  short int maps; /* bitsize 16, bitpos 64 */
  short int type; /* bitsize 16, bitpos 80 */
  mptr nulh; /* bitsize 32, bitpos 96 */
  long unsigned int maxsize; /* bitsize 32, bitpos 128 */
  unsigned char *applid; /* bitsize 32, bitpos 160 */
  lhdl gabinfo; /* bitsize 32, bitpos 192 */
  char *regfile; /* bitsize 32, bitpos 224 */
  mptr (*get) (/* unknown */); /* bitsize 32, bitpos 256 */
  short int (*free) (/* unknown */); /* bitsize 32, bitpos 288 */
  lhdl (*map) (/* unknown */); /* bitsize 32, bitpos 320 */
  void (*unmap) (/* unknown */); /* bitsize 32, bitpos 352 */
  lhdl (*alloc) (/* unknown */); /* bitsize 32, bitpos 384 */
  short int (*dealloc) (/* unknown */); /* bitsize 32, bitpos 416 */
  short unsigned int (*load) (/* unknown */); /* bitsize 32, bitpos 448 */
  short unsigned int (*unload) (/* unknown */); /* bitsize 32, bitpos 480 */
  short unsigned int (*syscall) (/* unknown */); /* bitsize 32, bitpos 512 */
};
typedef struct envblk_ /* id 9 */ envblk;
typedef long int LpiSTATUS;
...

This was the only file (trust me, I checked them all) that had any debugging data left on the whole ftp!

Now, this file was for a different architecture, operating system, compiler, release and was a long obsolete format. But, the clues it contained about the driver API were invaluable, and saved me so much time.

$ file alert.so1
alert.so1: SPARC demand paged shared library not stripped

The file was so old that to read the data I had to find an older version of binutils that still supported a.out-sunos-big. Luckily that worked and it was able to reconstruct a ton of useful data.

Here is the full recovered data, if you’re interested.

More Luck

I got another stroke of luck, I found a third party printer driver on an old SUNET archive for the Siemens Highprint 7400. Remarkably, it had some ancient Codeview debugging data left in it.

Here is a picture of a lady printing some rorschach tests with her new Highprint 7400, for some reason.

The debugging data was so old, I struggled to find a tool that could parse it. The data had a version code NB02, which means they probably used link.exe version 5, released around 1988. I tried TDUMP, CVDUMP and even SYMDEB, and none recognized it. I did eventually find an old enough version of Codeview that could read it, but it had no way to save output to a file except “printing” your backlog, and the backlog wasn’t big enough to hold all the data!

I had to script a DOS emulator to scrape the data out a page at a time! I can’t tell you how much time I wasted getting this working, but it was worth it, it gave some valuable hints about how the DEVPRIM (Device Primitive) API worked.

Publics for: _DVTBL!*
Symbols for: L13PSI74!*
3CD5:016E struct           psd
3CD5:006E struct GdvEntryTbl1_ pcf
3CD5:01DC unsigned int     print_head_height_dots
3CD5:000A char             grph_init_string[]
3CD5:0158 char             gr_eol_str[]
3CD5:016C unsigned int     col_res_num_pos
3CD5:0020 char             start_grph_str[]
3CD5:0050 char             last_grph_lf_str[]
3CB1:001C unsigned char far * graph_data_ptr
3CD5:0064 char             gr_eog_str[]
3CD5:0006 unsigned long    graph_data_size
3CD5:0000 unsigned int     scan_width
3CD5:001E unsigned int     stripNum
3AC8:0006 unsigned int PROC  hmu_len()
[BP+0008] unsigned char far * lmbcs_str_buf
[BP+0006] unsigned int     lbuf_len
[BP-0006] unsigned int     return_hmu
[BP-000A] unsigned char far * bundle_ptr
[BP-0004] unsigned char far * orig_ptr
3AC8:0148 void PROC        finish_down()
[BP+0006] unsigned int     distance
3CB1:000C char             max_feed[]
[BP-0006] int              i
[BP-0004] unsigned char far * p
3AC8:01A4 int PROC         graphics_init_printer()

Here is the full recovered data.