It looks like the laptop running Windows 3.1 has at least 16 bit color support, I rarely see Windows 3.1 with a powerful enough video card to support that!

My first "real" computer came with Win3.11 and an upgrade to Win95 when it was released... It had 1 MB of VRAM, so it supported 16bit color at a low-ish resolution...

I should start posting pictures of Project Dinosaur Browsing Ticalc.org

Project Dinosaur was A friend's challenge that I couldn't do anything with an 8088 anymore. Not only did I do something with a Zenith Transnote 8mhz 8088 machine, but I put it online with E-mail and Internet from a Dos Based Browser. It even Runs Win 3.0.

ticalc.org isn't too fun from DOS... (see Web Page link) ;-)

Ah, I see you've got Lynx for DOS! How fun!

I personally prefer Links (or ELinks) over Lynx, though.

Links is a trifle flashy for my tastes ;-)
I recently tried to see how many levels I could build up between me and a browser, and ended up running Lynx on a remote Linux server using PuTTY. PuTTY was running on Windows 2000, which itself was itself running inside MS Virtual PC. That was running on Windows XP.

It's the only way to surf!

I'm thinking of writing a terminal emulator for my Commodore 64, then I can hook it up to the serial port of a PC (I haven't figured this part out yet), to which I can then redirect Links or Lynx! Then I can surf from my C64 at a glorious 40 x 25 characters!

I miss fun things like QBASIC and MEMMAKER.EXE

I wouldn't exactly call MEMMAKER a lot of fun ;-)

QBasic was cool. remember the game where u throw a banana at the other gorrila? there is a ver'n in the archives...

IBM Gorillas. A true gaming classic!

I love that game! I made my own version for my Physics class (only with tanks). QBASIC is oldschool.

Yeah, I would play that all the time when I was 5 or 6, and then have a lot of fun writing really crappy programs in QBASIC...

I didn't care for Gorillas as much, but I played Nibbles a lot. I was rather annoyed at the scoring display bug that put commas in the wrong places. You'd think Micrsoft's QBASIC developers would know how to use a PRINT USING statement.

I don't think I ever saw that bug, but there is a bug that causes the program to have a division by zero error when I ran it on modern hardware. The speed loop goes so quickly that the timer doesn't even increment at all, so dividing by that difference (0) causes it to bomb. I just had to increase the delay loop by a few orders of magnitude and adjust the calculation line the same way.

I don't understand why it needs a speed test. It could just as easily do something like

SUB delay(t)
start = timer
WHILE timer < start + t: WEND
END SUB

to get a delay regardless of the speed of the machine. Maybe I'll take a look at the program again to see why it does the speed test.

I think I remember that, because I made a game that used a very similar delay calibration loop and got a division by zero error when trying on a (then) modern computer.

The problem with TIMER (if I remember correctly) is that the "resolution" is limited. I think it gave a couple of decimal places, but in my tests, it wasn't all that precise sometimes for very small delays, so I had to use a delay loop instead.

That's what I kind of figured... though the speed test should wait for the timer to increase before it ends, then it can be sure it doesn't divide by zero.

Of course, the DBZ problem was almost unthinkable back then.

You could probably get really advanced and use ON ERROR code to handle the error and try a longer speed loop.

Or probably better (if I'm thinking this through right), use a particular amount of time (x seconds every time) and see how many times a variable can be incremented in that time. Unless the variable overflows, that shouldn't be a problem, I don't think.

Another small problem with TIMER is that it resets at midnight, so if you were running a loop on the switch-over it would throw the program into an infinite loop. I had a BASIC book that had a TIMER delay routine that took this problem into account (I don't remember how, though), and I always used it in programs using TIMER.

*benryves invokes the evil spell ON ERROR RESUME NEXT*

I never found the ON ERROR stuff very useful because it acts more like a GOTO. I like the Try...Else...Endtry implementation better.

I did a lot of messing around with QBASIC, since I didn't really have much useful software at all. Sometimes I had to make my own (nothing fancy, though, since I didn't have that much programming experience at the time.)

I even ran MEMMAKER once or twice. I don't remember why, but it wasn't always a lot of fun. :-)