ARM BASIC: SOUND extension?

The last thing I intend to do is start a discussion with the person who does this as a living, however isn’t a point supposed to be 1/72th of an inch? (wiki link)

About that. On the link you quoted it actually has some pertinent information¹:
“In metal type, the point size of the font describes the height of the metal body on which the typeface’s characters were cast.” In digital type, letters of a font are designed around an imaginary space called an em square. When a point size of a font is specified, the font is scaled so that its em square has a side length of that particular length in points."

So basically the chunk of metal the character sits on has a defined height that matches the point size, however the character may occupy any percentage of that and the horizontal is going to be totally open to interpretation.

Nicely we’ve updated things for the electronic world:
“In digital type, letters of a font are designed around an imaginary space called an em square. When a point size of a font is specified, the font is scaled so that its em square has a side length of that particular length in points.”

So, another defined space to put the same vague definition within.

Dealing with this for years may explain Nemo’s quirky behaviour. :)

¹ Wikipedia is growing up.

Even the caps height varies from font to font of the same nominal point size. The point size is the line spacing when the text is “set tight,” that is, in old metal type days, without any “leading” – thin metal strips between the rows of letters. But how big the characters are on the “body” is down to the typeface designer.

Text isn’t usually set absolutely tight, although it used to be in the early days of printing, when paper was expensive. 10pt type would nowadays typically be set on 12pt spacing, that is, with 2pt leading.

Edit. Ah, snap Steve!

isn’t a point supposed to be 1/72th of an inch?

Well, historical pedantry aside, roughly yes. But the point size doesn’t measure what you think – it does not measure the bit you can see… it measures the bit you can’t see (not least on account of it not existing any more).

The point size is the height of the sort – the block of metal which carries the face (the glyph, the letter or symbol). Faces are almost always smaller than the sort (and hence point size), but certain glyphs can protrude. Some faces are much smaller than the point size, others are so large that you need extra space between lines to balance the overall colour. That space is made of a piece of soft metal, so it’s no surprise it’s called leading.

But now we have digital fonts that don’t have a sort. However, many honour the dimensions of the historical faces as closely as possible, so the differences in optical size persist. Enough history, the take-away is that physical size is only correlated with point size, not determined by it. And in the case of decent typefaces, physical height may not be linearly related to point size.

Which is why, if you allow the font to be changed, you must allow the font size to be specified, and it’s a good idea to allow leading to be changed if necessary. The formula I use for converting from configured Wimp$FontSpacing (or Wimp$FontSize if that’s not set) to menu item height is:

SYS&6008F,,400TO,,S%
L%=VALFNvarval("Wimp$FontSpacing")
IFL%=0:L%=VALFNvarval("Wimp$FontSize")
L%=(L%*75/S%+11)ANDNOT3

Which results in 12pt (the default) working out as a height of 44 (the default).

Note the call to Font_ReadScaleFactors, because the relationship between point size and pixel size is not hard-wired.

George tried

In short, you’re both right!

Actually, a printer’s point is not 1/72 inch. But don’t let that stop you.

Text isn’t usually set absolutely tight, although it used to be in the early days of printing, when paper was expensive. 10pt type would nowadays typically be set on 12pt spacing, that is, with 2pt leading.

I have no problem with that, the one that really irks me is the electronic document habit of floaty headlines and sub-heads.

Look at any Web page or Word document¹ and you will almost always see a sub-head sitting partway between two paragraphs with little clue as to its association – is it an emboldened after thought or a prominent lead into the next paragraph?

I’m now going to simmer for a while just thinking about it.

¹ Mine have altered styles for the heads and sub-heads. Then Word isn’t so annoying.

That space is made of a piece of soft metal, so it’s no surprise it’s called leading.

Well back in the day it was a high lead content item, no doubt back into history it was lead.

So I need to alter one bit:

So basically the chunk of metal the character sits on has a defined nominal height based on that matches the point size, however the character may occupy any percentage of that and the horizontal is going to be totally open to interpretation.

Actually, a printer’s point is not 1/72 inch.

The wiki actually covers that too and says the DTP point is defined as 1/72 inch

My pet typographic peeve is inappropriate drop caps.

I can’t tell you how much I hate that.

Rick enquired

BTW, what do you use to create those animated GIFs?

Don’t forget I’m emulated. So that’s GifCam for Windows, overlaying VirtualRPC’s window.

the DTP point is defined as 1/72 inch

Originally defined by PostScript.

My pet typographic peeve is inappropriate drop caps.

I think that’s an affectation harking back to those pretty pages prepared in ye olde days by religious bods with nothing else to do.
Wasn’t the standard drop cap use one where the Drop cap and an amount of following text all emboldened?

Normally done with longer line lengths anyway, so that would likely have been “After 25 years of drought and two of too much rain, Philadelphia has its championship."

Yes, your snippet looks awful and is hard to read.

Not used such stuff for a long while (who prints anything?) but I seem to recall drop caps in either or both of Ovation and Impression so we can probably knock up a few examples to keep you happy :)

I’m trying to work out what appropriate drop caps might be. Or are they simply the pretty ones in illuminated manuscripts, nothing else?

I’m trying to work out what appropriate drop caps might be. Or are they simply the pretty ones in illuminated manuscripts, nothing else?

Possibly the latter, but I could deal with “A long time ago and far, far away" – making sure that trailing text was emboldened and the following body text was smaller.

inappropriate drop caps.

That A is far too big, it has pretentions of religious manuscripts. Surely a decent drop cap ought to just drop down one line. Anything over the top (as that example) just gets difficult to read, so emboldening the text shouldn’t be necessary.
Like… https://www.newyorker.com/books/under-review/who-owns-a-story-trust-exercise-susan-choi

My pet peeve is MOSTLY SEEN IN LICENSES WHERE HUGE SECTIONS OF SUPPOSEDLY IMPORTANT STUFF IS WRITTEN IN CAPITALS BECAUSE THAT’S SUPPOSED TO GIVE IT WEIGHT AND POWER BUT ACTUALLY ENDS UP MAKING IT DAMNED HARD TO READ WHEN ITS PARAGRAPHS OF THE STUFF AND OFTEN PRESENTED ON SCREEN IN A SANS SERIF TYPEFACE JUST TO ADD TO THE PAIN.

My pet peeve before was “the above, but on Windows in the days before TrueType fonts were able to be antialiased blurred around the edges”. [example of early TrueType]

So that’s GifCam for Windows

Thanks. It was obvious it wasn’t going to be anything RISC OS. ;-)

so emboldening the text shouldn’t be necessary.
Like… https://www.newyorker.com/books/under-review/who-owns-a-story-trust-exercise-susan-choi

Which nicely demonstrates another floaty aspect – where the drop cap isn’t sitting on the same baseline as the second of the two lines it’s set with. Just enough off to be irritating. If you can’t see it look at the “I” where the base of the I strikes across to a level above the bottom of “The question is sort…”

My pet peeve is MOSTLY SEEN IN LICENSES

Ah, that’s merely an example of the authors demonstrating a lack of functional brain cells. They are concerned that you don’t know that and do a little demonstration so that you can give the appropriate attention to what they’ve written – bugger all.

hanks. It was obvious it wasn’t going to be anything RISC OS. ;-)

My first guess was that it is yet another unreleased Nemo tool that’s ready “real soon now” :-)

Rick alleged

before TrueType fonts were able to be antialiased blurred around the edges

Now hold on one cotton picking minute. That’s not only utter rubbish, it’s embarrassingly hypocritical rubbish written on an RO5 forum. Cripes.

You may have to open this image in another tab to ensure you’re seeing it 1:1. RISC OS 5 on the left, Windows in the centre, my RO4 on the right.

Anyone casting aspersions about Windows’ font technologies should not be throwing those stones from inside the RO5 greenhouse.

Both left and right are the same Homerton.Medium displayed in !Edit BTW.

I have to say I don’t think any of them are that great. RO5 is at least uniform weight, but yes, it’s blurry. Both the Windows one and yours have noticeable variations in stroke weight. You pays your money and you takes your pick: do you prefer things sharp, or uniform weight? With modest pixel sizes, you can’t have both.

In your SuperSample, the l.c. Ts and Fs and the u.c. H all stick out like sore thumbs to me. Windows’s u.c. H has one upright heavier than the other. But no, I don’t much like the blurriness either. It’s the price you pay for finite pixel sizes.

takes your pick

Have you not noticed that the RO5 rendering is completely the wrong weight?

Oh hang on, what are you viewing the image on? Is it a RISC OS machine with no colour management, a large gamma, and software that has no idea what either of those things are?

No, I’m viewing it at 1:1 on FireFox on a Mac. It looks much the same at 1:1 on XNView on the Mac.

And yes, I know the weight’s wrong on RO5.

I don’t actually have much of an issue with any of them, but I don’t think any of them are perfect, nor could they be given the pixel sizes.

A completely different thing, that could be better, in some (many? I don’t know) italic or oblique fonts on both TruType and PostScript (and probably RISCOS too, but I don’t do a lot of text on RISCOS these days, and when I did I had control over such things, easy bit of hacking) is the kerning around apostrophes. Terrible!

I’m probably weird (don’t answer that!), but I prefer the original RISC OS version. Perhaps it is to do with my myopia, however it is possible to view it clearly both close and further from the screen.

Yours looks quite pleasant, but perhaps ought to be a tad heavier. One thing I do notice is the kerning is a mess – “Ho(space)merton”, “cu(space)stom”, etc…

Windows? While I’m none too keen on that, I would ask which version of Windows. XP?

RISC OS 5 on the left, Windows in the centre, my RO4 on the right.

Presumably captured from a Windows machine with emulated RO instances.

What emulator in both cases and have you compared with an instance on a real machine into its own display?

Just asking because the display of the emulator window has a rather large contribution from the emulator and the treatment of the host OS of that emulator – that may be the issue. May not as well, but one should avoid the exercise derived from jumping to conclusions.
IT tech – consider all the factors and remove each in turn until something changes (but start with the most likely to make a big change.)

While I’m none too keen on that, I would ask which version of Windows. XP?

XP!! Cough, splutter.
Does Nemo strike you as one to invite disaster?

Got a request to allow the install of a system with XP the other day.
Everyone in the office is contemplating the issue to come up with something polite to say.
“No” (end sentence) is the politest so far.
Including the information that they are free to choose left or right at the traffic lights at the front of site is considered a bit much, but it’s implied.

XP!! Cough, splutter.

The latest Windows font rendering that I am familiar with is ClearType on XP. It is that which I’m comparing with RISC OS (might have been a clue in my posting a screenshot of Win3.11 as a prior example?).
I would imagine things have improved in the decade or so since then, but I would like to point out that RISC OS originally had a far inferior font mechanism too, the old bitmap method.

allow the install of a system with XP the other day

Well, that stands to reason. The only reasons I’m sticking with XP is…

1. My PC is an old P4 box, probably incapable of doing anything useful with newer versions.
2. I pretty much only use the machine for ripping DVDs and uploading stuff to my site with WinSCP.
3. Because of #1 and #2, I don’t find any need to justify the time/expense/hassle of updating an ancient machine that works to a newer machine that’ll do the exact same things…

But I’m a home user who knows that XP is an insecure crock. Your use case is more “nuke it from orbit”, and rightfully so. I still see the electronic checkout machines in the supermarket using XP (obvious when it reboots) and I cringe… That said, the last time the cash machine at the bank got itself stuck in a boot loop, it was flippin’ OS/2!

And – if it isn’t something of the XP era, it’s a little unfair to compare yesterday’s font technology with something that hasn’t been touched since the very early ’90s.

I’m viewing it at 1:1 on FireFox on a Mac

Is this on your TV?

Try this test