Vintage Computers — 8

Some years go on the 486 system the CMOS battery began the predictable decline. This is a barrel battery soldered to the board and I have been too lazy to replace. The battery retains some function for many hours after powering off but eventually drains to forgetfulness.

Before that battery began deteriorating, when I was still using the system in my early days of creating a house network, I had a written a time sync script on the Windows for Workgroups (WFWG) side that used the net time command to sync the system clock to the house network time server. At the time I considered that trickery pretty cool.

Despite that improvement and because the system sat powered off most of the year for the past several years, often when I first powered on the system I would forget to set the date and time in the BIOS. Not that MS-DOS or WFWG are time critical operating systems — and the net time script corrected the problem anyway. To remedy my forgetfulness I revised the AUTOEXEC.BAT to manually request the correct date and time. This avoids needing to remember to enter the BIOS after the system has been idle for long periods.

    rem The CMOS battery is old.
    rem Manually prompt for the date and time.

unrelated but a quirk that bothered me with this system is the notorious fragmentation of the FAT file system. The WFWG swap file needs to be contiguous but when creating, the file gets inserted anywhere on the disk. I always thought the file should be out of the way — at the end of the partition space. Functionally this is irrelevant but cosmetically “seemed right.”

I booted into Windows and deleted the swap file. On my Slackware office system I created empty (/dev/zero) dummy files on a network share. The 486 has network access. I copied dummy files to the local disk. I repeatedly ran the Norton Speedisk utility to defragment the disk until there was barely sufficient empty space at the end of the disk partition to create a new swap file. With most of the disk partition filled I created the swap file. I repeated the disk defragmentation until satisfied my trickery succeeded. I deleted the dummy files, repeated the speed disk exercise, and had a swap file at the end of the disk partition. If I was going to use the system in production there would be little to no file fragmentation caused by the swap file.

More to come.

Posted: Category: Usability Tagged: General

Next: Vintage Computers — 9

Previous: Vintage Computers — 7