Current issue

Vol.26 No.4

Vol.26 No.4


© 1984-2024
British APL Association
All rights reserved.

Archive articles posted online on request: ask the archivist.


Volume 16, No.1

Guest Editorial: The Walls are Closing in ...

by Adrian Smith

Stef is taking a well-earned break in Italy, so Vector Production is taking its chance to trail some of its wilder ideas and fears across the bows of the APL community. Firstly, I must thank both our new Editor and the rest of the working group for taking the deadlines seriously, and getting Vector back on schedule. This time, we had plenty of good material to choose from, and it was all assembled and ready for typesetting by the beginning of June. In fact there is a lot of good stuff already lined up for 16.2; this makes it so much less hassle to get the final copy prepared, and allows us to get much of the material on the Vector web site in parallel with the physical publication.

One clear consequence of keeping the web site well maintained and indexed has been a rise in demand for back numbers. Today we parcelled up the 3rd complete set of back numbers in as many months, and there has been a steady trickle of requests for odd volumes. Yes, the stock is still there to do this, but only just! We are down to single-figure stacks of several early editions, so if you want to complete your set, better get in touch soon!

On to the fears and hopes! Did you ever drive up one of those narrow lanes in the Yorkshire Dales where the walls seem to be closing in on you? The effect is most marked late at night, and the experience is enhanced when you come upon a herd of Aberdeen Angus (the black, woolly ones with green eyes) in the middle of the road. What has this to do with APL? In particular, what has it to do with the APL character set and the Internet? Consider for a moment what we are asking the browser to do ...

  • <code>2+2</code> tells it that what lies between these tags is computer code, so use a fixed pitch font and don’t muck about with silly tricks like smart quotes. In IE5 it is essential to use these tags, unless you want results like ‘f’ ŒWC ‘Form’, or are happy to see all your subtraction signs rendered as .
  • <font face="APL2741"> says ‘use this font please’. In IE4 all was fine, but in IE5 you discover that unless the font is marked as ‘symbol encoded’ any characters in the range 128-160 are rendered with Courier as the browser assumes that the font you requested cannot produce them! This makes APL code pretty unreadable, as a quick glance at Charmap will confirm!

So far, not too serious, until you set the appropriate switch to get it all working in IE5, sit back contentedly, and drive slap bang into the back of Netscape-4 which cannot get its tiny brain around the notion of a symbol-encoded code font, and refuses to use it at all! It really is all our fault ... computer code uses the low ASCII characters, everyone knows that! It’s not that the browser-writers are out to get us intentionally, it’s just that we are so far outside their mindset that the walls they are building are closing in fast and there may already be no escape.

So much for reproducing the character set, now what about the opposite problem – typing it in to an editor! Just how many of us old mainframers have to die before someone finally has the courage to dump the old 2741 layout and make a rational mapping of symbols to keytops? OK, we learned to type APL when there was much less to learn – you got to know the set of base symbols quite quickly and the funny ones were composed with judicious use of the backspace key. OK, {jot}backspace{shoe} was a bit bizarre for {comment}, but in general you got all the common stuff in single keystrokes and many were quite easy mnemonics.

Then we had the old 3270 screens, with a pretty accidental scatter of overstruck symbols generated by various Alt+key combinations. At least we still had them engraved on the key tops! On to APL*PLUS, and the random scatter was legitimised as the ‘union’ keyboard – Dyalog even preserved the mapping of quote to Ctrl+K, but at least they gave us stickers. Now we have the worst of all possible worlds – a stupid mapping and absolutely zero assistance from the keytops or the APL environment.

And you wonder why no-one is learning APL any more.

script began 9:05:45
caching off
debug mode off
cache time 3600 sec
indmtime not found in cache
cached index is fresh
recompiling index.xml
index compiled in 0.1832 secs
read index
read issues/index.xml
identified 26 volumes, 101 issues
array (
  'id' => '10008840',
regenerated static HTML
article source is 'HTML'
source file encoding is 'ASCII'
read as 'Windows-1252'
completed in 0.202 secs