Reading some of the questions here got me to thinking... how many of the old methods, software or hardware, are still with us today?

My two examples:

HTML. It's line oriented, for crying out loud. Do you know how much development effort is wasted working around that limitation, rather than address any part of the screen at any time? Dates back to line oriented text displays that the original hypertext of the 1970's was designed to run on.

And, a lot of new desktop PC's still have at least one ISA or ISA compatible slot on the motherboard.

  • 11
    This is very broad, and also it's a bit of a rant. – Wayne Conrad Jul 2 '17 at 1:14
  • 4
    @WayneConrad - And a rant that I think is rather pointless. Computer design - like any other technological system - is an evolutionary process rather than a revolutionary process. The instruction set of an i7 3+ Ghz multicore CPU has some design connection back to the 8080. But there's nothing wrong with that. And the HTML part is entirely off-base - HTML is not line oriented, and software going back to the 1970's could address any part of the screen at any time. – manassehkatz-Moving 2 Codidact Jul 2 '17 at 5:04
  • 1
    Try using modern software integrity practices like version control in a non-textual programming environment and you might change your mind. This can be a pretty serious problem for whiz-bang WYSIWYG tools - lots of fun for a lone wolf, utterly unacceptable for a team that needs maintainability. – Chris Stratton Jul 2 '17 at 7:11
  • 3
    The simple answer is "all of them". – Chenmunka Jul 2 '17 at 7:45
  • 4
    I would really love to believe a lot of new desktop PC's still have at least one ISA slot, as I have a number of very valuable cards that need that - no, they don't. Contemporary computers with ISA slots are limited to embedded applications today. – tofro Jul 2 '17 at 7:58

I'd say many, many more than one could expect. Examples:

  • Any current computer has either a Von Neumann architecture, or a Harvard arquitecture. There possible may be other exotic architectures, but the vast majority of current day computer systems fall into any of these two categories.

  • The three major OS for desktop computers have some sort of terminal interface, despite how fancy their graphic system is.

  • The keyboard layout we use today can be dated back to the days of the PC-AT design. Maybe earlier.

  • Any 32-bit Windows version that supports MS-DOS, is still supporting some data structures that can be dated back to the CP/M days: when Windows load and run a .COM file, it loads it at offset 100h. The memory before that offset is reserved for the command line, and, if the command has one or two parameters, the OS creates two empty FCB's (File Control Block) for the command to use them in case those two parameters are files. FCB's were in fact deprecated with DOS 2.0, and they were only used in DOS 1.x . FCB's are an important part of the CP/M interface with files.

About your two examples:

  • I'd say HTML parsers are character oriented, not line oriented. In fact, an HTML parser does ignore new line/carriage return characters. And on top of that, there is the tag which allows the designer to put content on any position on the browser window (I'm not an HTML expert so it may happen that DIV is not the only one to allow that)

  • It's very rare that new desktop PCs have ISA slots. They may have ISA-style devices, for legacy OS, but ISA slots are gone long ago. They may be seen on some industrial-rated motherboards.

| improve this answer | |
  • 1
    Actually, most current computers have neither a Von Neumann nor a Harvard architecture, but a combination of the best of both. Often a Von Neumann programming model but with hardware paths optimized for the frequent case where Harvard style segregation can be maintained, especially in the caches and pipelines. These things are easy to forget about, but not entirely transparent in the case of self modifying code - including dynamic linkers. – Chris Stratton Jul 2 '17 at 7:15
  • 1
    HTML has never been line oriented. You can trace its history right back to IBM's GML (1969) and GML wasn't line oriented either. Of course GML documents were often displayed on text-only computer terminals, but they could also be printed with proportionally spaced fonts, properly justified lines of text, etc - and that was how most printed IBM manuals were prepared at that time. – alephzero Jul 2 '17 at 18:24

Not the answer you're looking for? Browse other questions tagged or ask your own question.