Jot Down Memories

~(already recalled some. but jotted these down, just now... 11:16 PM 10/8/2011)


1.) MSDOS 1.0, 5.5, 6.22 almost done long before marketed.
*Scandisk
- my textmode graphics was better,
led to a better addressing scheme;
- error detection and correction, adler32 = adler64
(i think i found a better one, better recovery, fewer errors);
- resulted to CMOS/BIOS modifications/improvements, patented;

2.) MSDOS interrupts. Ours. Maybe the one painstakingly done by her almost without sleep (i wanted her to enter the binary opcodes by hand). Or our own BIOS, which was definitely what we wanted. I added two interrupts of my own devise, which started it all.

3.) GUI for windows. "Double menu-ed" XP start menu from 95. It should have been for Win95, but i wanted it for XP. Someone of higher influence decided on the blue left side of the menu in the "All Programs" option of the Start menu. As well as the width. Perfect.

"Parallel development of windows versions," i said. Should boil down to just the graphical user interface.

The standard left selection pane we now see in most windows I also conceived and decided upon. For windows 95 first, but then it was decided for XP. "Maybe in Win95 wizards. Then XP later." The stacked, calendar-like selection menus inside were mine; not the usual tiled toolbox. "Just text strings!" was hers, but she was just faster to say it, as well as the left icons beside them. Very crude graphics before. This is where we tinkered with improving screen "colors", and saw how the used "data structure" really is crucial to faster graphics. A little background dithering was done here.

Internet browser-like windows, for users to easily go back to the previous "context or content," and for code re-use and programming ease.

"Icon menus" idea was mine, not the usual right-clicked context menus.

4.) I remember Steve Jobs. I asked him if he liked Win95 GUI but he badly wanted Windows 3.11 tech for faster graphics. That's why we came to analyzing Apple hardware configs, too.

I was distributing compilers. We developed a language and compiler creation toolkit. In a snap, with keyboard inputs and an "interference" from the user, they can have a Pascal or C compiler product. I made Quick C more powerful (you can extract all the files inside the Quick C executable) and Turbo C faster. I think I created a batch file for this task. Quick C was later weakened for distribution, and this feature was omitted.

I asked for Turbo C and Turbo Pascal to be marketed by Borland, but Borland's were Borland Pascal and Borland C/C++. (The future of the IBM-PC line was still uncertain at the time. Bill knew about the Turbo C/Pascal deal with Borland.)

They went to me because i changed the IBM-PC BIOS, having introduced the MSDOS interrupts. They of course wanted to know why those new interupts were introduced at all.

The product names "Windows", Windows 3.1/3.11, Windows 95, Windows XP, and Windows Diamond already set from the start.

"Internet explorer." Explorer "e" logo. Explorer will "roam the globe."

"Opera" was named there.

"Microsoft Office", "Word", "Excel", "Access", etc., also ours. "Word." I really liked the way i named Word. And Office.

As well as the "Visual" and "Studio" names for the compiler suite. C++. And Basic, in honor of Bill's Altair Basic. He might want to "still program."

I remember Windows XP and Office 2007 GUI came from one group only. Very neat. It is assumed, of course, that in the following years they had been developed by a group of international programmers and developers, for international character sets (Unicode) and languages supported, as started predominantly in Windows 95.

"Ribbons and bundles." The groups inside we called "bundles."

"Pointers for Visual Basic later."

5.) We reached higher-color quality for every existing hardware graphics color mode. Started with Windows 95 selection pane dithering. Then it led to color screen hardware innovations, apart from existing monitors, such as new LCD tech.

Speed improvements in MP3 and H.264, thus H.265. All speed improvements were dictated by hard disk "bandwidth" capacity. As well as in networking. "Networks and Internet bandwidth are governed by hard disk speed."

I had to take charge of the computing landscape. It was the upcoming internet that made me concerned.

PowerPC up to Apple G5. Dual Core, then "Quad-core is enough for all users."

"Variant arrays" was my tech, for everything. "Skip Lists." But variant arrays, better. (The two of us were the first Windows ergonomics engineers, too). Many new innovations from her because of the variant arrays. A design principle for user interfaces I shared with her; well, she actually "discovered" it. So I articulated the principle more succinctly. And I think she liked it. I told her the variant arrays data structure is perfect for her area of expertise in all aspects of "windows" development. "Where are you good at?"


(10/22/2011)