OT - Programming Languages

[Tim Streater]

Yes (up to five or whatever you've got it set to).

Unfortunately it depends. If there are any non-whitespace chars earlier on the line, backspace removes one space (Delete is forward delete, BTW, and always just removes one char). If the line only contains whitespace, it deletes back to the previous tab stop.

I did have a word with the developers about that, they said they'd thought about it and decided to do it that way. Anyway I didn't really mind; the important thing was to have a quick way of tabbing across without leaving a trail of tabs, whose width is entirely variable, by definition.

Five, because they're spaces.

Then you'd turn those features off or configure them differently. Personally I'd ban tabs from any software I was responsible for; if you're worried about disk space the file can always be compressed.

[news14k]

My understanding is that the intent of meaningful whitespace is to force users to indent their code, so that it's more readable. But it's a shame you weren't around with that "throwback" remark when they were specifying - python claims of a similarity with Fortran would probably have killed the idea stone dead :-)

#Paul

[polygonum]

Many, many years ago I wrote a filter that would replace multiple spaces by multiple space format effectors as files were written. And the reverse when reading them back. Had the advantage of reducing file sizes (e.g. of program sources) pretty substantially. But when using a screen editor or simply viewing them, even without the filter being invoked the display showed them all as good and proper spaces. I felt it was the best of both worlds.

Complete waste of time these days.

[John Rumm]

Its surprising what you can put up with when you are a contractor being paid a decent hourly rate and someone else keeps making the job take longer through no fault of you own ;-)

(to be fair it was quite amusing in a way - kind of like watching realisation dawn in the stone age!)

[John Rumm]

Or the "notable" phrase:

The notable doctors was not able to operate because he had no table.

[Vir Campestris]

ISTR up-thread someone said there was no significant difference in natural languages between one space and two.

The notable doctor was not able to operate because he had no table.

doesn't read any differently. And that is if course shown by variable pitched fonts and leading.

Andy

[Bartc]

For free-format text, there shouldn't be (just makes it harder to read if the spacing is odd).

But when using natural language for more technical purposes such as constructing columns and tables, then having words and numbers in the wrong place, even though the left-right/top-down flow is the same, can give totally the wrong results. Just like Python.

[Jeremy Nicoll - news posts]

Me too (that's to say, 'ooREXX' though I still mainly use its 'classic' REXX features rather than the oo aspects).

The text editor I use most is Mansfield Software's Kedit, which is a PC version of VM/CMS's Xedit, though has some of the features of MVS's ISPF edit as well. And it's programmable in 'KEXX' which is almost the same as classic REXX.

I would prefer to use a PC version of ispf edit, but the best contender, CTC's SPF Edit, has since before the year 2000 lost its REXX support (and also dropped support for ispf-like dialogs - panels, tables, skels etc) and uses a subset of C instead. The CTC programmer presumably has a PC background and doesn't understand that dropping REXX support for the editor annoyed all those users whose existing collection of macros - which worked both under MVS/TSO/ISPF and on the PC - suddenly no longer worked on their PCs.

Having to recode every existing macro in the C subset is a non-starter.

[Bob Eager]

I'm using Regina for REXX, mainly on FreeBSD. But I started with it back in 1991 on OS/2.