@ Novell haters
I wonder how many of you were there?
At the time of WP 5.1, the first Windows port, there were also other GUI ports of WP ongoing. I saw 3 of these run side by side, performing the same operations on the same document. There was a visible difference in the display update speed. The Display Postscript version on Next was far and away the fastest. The Windows version was the dog which wouldn't hunt. It was also clearly (and painfully so) slower than Word on comparable operations. There can only be two explanations. Either the Windows porting team was totally incompetent, or they did not have access to fast display routines - routines that Word on Windows appeared to be using. Since the management could see their 3 ports of WP, the incompetence charge can't stick - if something could have been done, it would have been. Windows APIs change - frequently - precisely to screw the competition by making them play catch-up. Once the M$ product is in the market, the APIs get published, as the game no longer depends on technical advantage but on marketing clout. This is precisely what happened - eventually the display APIs got published, the playing field was put back to level, and WP in Windows became usable, but M$ had already made their end run around it. Anyone need a lesson in M$ marketing?
Ok, a quick one. WP Corp had seen the way the wind was blowing out of Seattle. How M$ had put code in DOS to catch and slug Lotus. How M$ had killed Aldus' word processor, which had at least a 6 months lead on Word, by announcing Word would come out next month - it didn't, and couldn't, because it didn't even exist as a Win 3 product. I guess that is why WP Corp sold up to Novell. Again you may remember Novell were the leaders in networking at the time, and I guess they thought they had the clout to take on M$ even knowing full well how the Beast operated.
I maintain WP was always superior to Word from the engineering pov. The decision by Word's designers (not M$ BTW - they bought it in if you remember, like so many other things) to put formatting information in effectively a separate fork in the file, and run it in and out of memory produced quicker file read and write times, but is only any use for a kiddies' word processor that never has to format more than a couple of pages, and a poor one for anything with long document and/or DTP pretensions. (Your code ends up spending most of its time in malloc() and who said something about garbage collection? Don't you know how that's going to impact the user experience?). And as Mr Coward pointed out above, WP's file format needed only a trivial (5 minute) change to accommodate SGML and XML markup. Contrast that, if you will, with M$'s shenanigans over XML - new file format, OOXML, etc.
On days when not just the sky is grey, but the trees, fields and houses also, I think the quality metrics for software are inverted in proportion to the perceived market for it.