Tiro Typeworks

Official account of Tiro Typeworks Ltd., a digital type foundry founded in 1994 by John Hudson and Ross Mills, specialising in custom fonts for multilingual publishing and computing.

#Typography #TypeDesign #Fonts #Unicode #WritingSystems #Linguistics #Palaeography #Calligraphy

#fedi22

Public Key

npub1rmg092al9kzwfgzrv57jx3frhm2efcw4eu42n8hfswjewv3autuq0wuntj

NIP-05 Address

TiroTypeworks@typo-social.mostr.pub

Profile Code

nprofile1qqspa58j4wljmp8y5ppk20frg53ma4v5u82u724fnm5c8fvhxg7797qpz4mhxue69uhhyetvv9ujuerfw36x7tnsw43qz9thwden5te0wfjkccte9ekk7um5wgh8qatzfj3dsp

Publishing to

relay.ditto.pub

relay.mostr.pub

Author Public Key

npub1rmg092al9kzwfgzrv57jx3frhm2efcw4eu42n8hfswjewv3autuq0wuntj

Show more details

Last Notes

2026-05-05 18:58:17 UTC

- reply

@nprofile…mrzw U+02BC and U+02D7 are spacing modifier signs, not combining marks. It would be best not to display them with dotted circles (which is a Unicode convention for characters with the Mark property).
Please clarify, as you have for other marks, whether there are any normalisation concerns for U+0303.
Suggest edit to opening sentence:
‘This document concerns two modifier signs and five generic combining marks...’

2026-04-17 20:51:47 UTC

- reply

@nprofile…ahkr @nprofile…wguy
‘... if there is a language expert backing it.’
Glyph names are generally script-based, not language-based, and given the technical requirements of fonts are usually most usefully derived from the Unicode character names, even though these sometimes differ from localised names in particular languages. Indeed, Unicode character names may be least useful n this respect when they correspond to language-specific names, as in the case of Sinhala, in which case the informative annotation names are more helpful.
https://typo.social/system/media_attachments/files/116/422/006/690/589/699/original/08df5913902fdfcb.png

2026-04-16 15:44:05 UTC

- reply

@nprofile…hvh8 Ross and I were sitting in a cafe one evening in late summer 1994, and decided that we wanted to have a foundry identity and small printed specimen of work in progress to take to that year’s ATypI congress (San Francisco) in September. Ross suggested Tiro Typeworks, named after Marcus Tullius Tiro, Cicero’s personal scribe and inventer of Roman legal shorthand. If we’d had more time to think about it, we might have come up with something less obscure, but I’m glad we didn’t.
I fondly remember Margaret Richardson (U&lc) taking us under her wing at ATypI, introducing us to everyone with ‘Have you met the Tiro boys?’
https://typo.social/system/media_attachments/files/116/415/138/418/787/845/original/f2e74624146ac281.png

2026-04-15 17:28:41 UTC

My choice of music isn’t usually directly related to what I am working on. But it’s Gurmukhi time, and I have a playlist of the 100 greatest Bhangra songs.

2026-04-09 22:24:02 UTC

- reply

@nprofile…sd2l Favourite left-field cover is by 1990s Toronto group the Leslie Spit Treeo.
https://www.youtube.com/watch?v=eGbFD5dZbWY

2026-04-07 16:40:12 UTC

- reply

@nprofile…05la My copy arrived yesterday. The printing isn’t great, but far from the worst I have seen. But the binding is shocking: no signatures, just a glue strip, attached to boards by the endpapers. This is not going to last.
Re. printing, while metal type could produce glorious results, which are rightly celebrated, a huge amount of it looks like this: worn, damaged, poorly fitted, and frequently over-inked type.
https://typo.social/system/media_attachments/files/116/364/395/788/067/528/original/322f4cd10108b46d.png
https://typo.social/system/media_attachments/files/116/364/407/968/326/151/original/df788045ad7bc639.png

2026-04-04 18:11:32 UTC

- reply

@nprofile…ale9 Try -30/-35. I see the illusion, but I think you have overcorrected.

2026-04-03 18:41:24 UTC

- reply

@nprofile…j074 @nprofile…f8fy @nprofile…6p0k I’ll second that. I have a collection of friends’ PhD theses printed by Lulu, and the quality is good. There is a range of options for paper weight and finish, and for binding.

2026-04-02 03:54:33 UTC

- reply

@nprofile…wguy @nprofile…uvqd @nprofile…vxlc @nprofile…fnwc You want
U+202D LEFT-TO-RIGHT OVERRIDE
U+202E RIGHT-TO-LEFT OVERRIDE
If you map your flipped glyphs in the rtla layout feature, they should get automatically substituted when the text direction overrides are applied.

2026-03-23 16:39:13 UTC

- reply

@nprofile…pajl What languages are you targeting? It is worth bearing in mind that French is the only modern European language in which both æ and œ occur, but even there use of æ is rare.
In Mediaeval and Early Modern Latin texts both occur frequently, but these days ae and oe are strongly preferred.
To my knowledge, very few orthographies anywhere use both these letters.* So the possibility of confuseability of the italic forms is less of a practical concern than one might think simply from looking at the shapes.
* @nprofile…y5sf can probably identify those that do.

2026-03-23 15:00:14 UTC

@nprofile…7kma @nprofile…9ng9 Another PS/CFF hinting question:
Are BluesValues and FamilyBlues (and corresponding Other) expected to have cardinality between the pairs? It feels to me that they probably should, but I have seen output from tools that don’t write them in the same order (possible bug.

2026-03-22 15:24:27 UTC

- reply

@nprofile…9ng9 @nprofile…7kma I understood Jens to be specifically talking about the zones in the BlueValues type, which are all top zones except the baseline zone (which as he describes it is behaving like the top zones in terms of rounding direction).
Are you saying that there are other factors involved in the rounding direction. Size? Where is this documented?

2026-03-22 15:03:52 UTC

- reply

@nprofile…mrzw It’s specific to sans serif designs, and optically compensates for the fact that a shape with a longer horizontal base can appear lower than one with a shorter horizontal base. [Also true of tops of such shapes, of course, which is why the crossbar of f and t often benefit from sitting slightly below the x-height of other letter.]
It’s a *much* subtler effect than overshoots for rounds or pointy bits, but belongs in the same class of optical adjustments.
https://typo.social/system/media_attachments/files/116/273/424/076/127/451/original/7ec49977fcfb0524.png

2026-03-22 14:50:43 UTC

- reply

@nprofile…7kma @nprofile…9ng9 Yes, I was thinking afterwards that the other way to implement this would be for z to sit on the baseline and all the other (sans serif) letters would descend slightly. It feels counter intuitive, though, for the optical exception to be the norm in terms of defining the baseline position.
Your reply to Ben about BlueValues all rounding towards the top of the zone is the (other) missing piece of information in understanding this. It explains why the top of the baseline blue zone is specified as the baseline, and why it is included with the BlueValues rather than with the OtherBlues.
Also, isn’t TrueType great.😁

2026-03-22 06:02:22 UTC

- reply

@nprofile…9ng9 @nprofile…mrzw When Sumner says BlueFuzz is no longer used, do you think he means
a) BlueFuzz is no longer used in Adobe fonts
or
b) BlueFuzz is no longer used by rasterisers
?

2026-03-22 03:52:46 UTC

Happy to discover that the CoreText engine is correctly handling my akhn ligature lookups for the Unicode Vedic Extensions block Jihvamuliya and Upadhmaniya characters with Devanagari and Bangla consonants.
Thanks @nprofile…j7ax !
There is a proposal, which I support, to encode the Devanagari Jihvamuliya and Upadhmaniya separately, and reserve the ones in the Vedic Extensions block for Bangla. In the meantime, I think other shaping engines need to at least implement support for the Bangla ligation, which currently breaks in HarfBuzz and DirectWrite.
cc @nprofile…dtvq
https://typo.social/system/media_attachments/files/116/270/760/836/058/877/original/834eaeae1741d1ff.png
https://typo.social/system/media_attachments/files/116/270/786/782/934/912/original/08a5f213cbef494c.png

2026-03-21 23:25:59 UTC

- reply

@nprofile…dfhn @nprofile…mrzw Is this baseline issue the reason why BlueFuzz exists? I can’t think of another reason why one would need it and, hence, why it would exist if the baseline alignment zone acted like all the other alignment zones and wasn’t hammered to y=0.

2026-03-21 23:23:12 UTC

- reply

@nprofile…dfhn @nprofile…mrzw So in my hypothetical z @ 4000 UPM scenario, the minimum distance between all would be 25 units, yes?

2026-03-21 23:16:06 UTC

- reply

@nprofile…mrzw I’ve never really understood why BlueFuzz exists. It extends the blue zone, and reduces the available space between zones, so in what way is it different from simply specifying a bigger blue zone?
Like other aspects of the inherited PS Type 1 hinting model in CFF, this was originally spec’d for a 1000 UPM. It isn’t very clear if the expectations are that all the math scales for a UPM of 2000, 2048, 4000, etc., or if the minimum distances remain fixed units as per the PS format spec?
So, if UPM = 4000, and bottom of z is sitting 12 units above the baseline, do I need a BlueFuzz of 12? Or of 3? And do I then need to ensure that no blue zones are closer than 36 units, or 3 units?
I recall that the BlueScale value had to be changed for larger UPMs.

2026-03-21 14:55:34 UTC

- reply

@nprofile…9ng9 This would be another scenario. Basically, any situation in which a shape doesn’t happen to sit flat on the baseline.
https://typo.social/system/media_attachments/files/116/267/736/815/479/355/original/c65e704a871de605.png

2026-03-21 14:44:08 UTC

- reply

@nprofile…9ng9 Classic example would be Frutiger’s original designs in which the baseline stroke of the z sits slightly above the baseline for optical reasons. [In digital versions, this was normalised to sitting on the baseline, perhaps because of this issue?]
In my immediate case, there is a serif flex 2 units above the baseline., which is probably ignorable but intinctively I consider something that belongs within the baseline alignment zone.
I figure as long as the first blues pair is the baseline alignment zone, that should be defined as makes sense to the designer. Tools shouldn’t be inserting BluesValues not specified by the designer.
https://typo.social/system/media_attachments/files/116/267/672/185/186/838/original/c670cae9732dd84e.png

2026-03-21 01:03:18 UTC

The Adobe spec for BlueValues states that
‘The first pair is the baseline overshoot position and the baseline. This is a bottom-zone.’
Tools I hav tested that output CFF fonts all seem to add a -20 0 bottom-zone to a font if the baseline zone in the source is not set exactly at 0, i.e. if it has an overshoot above as well as below the baseline, e.g. to catch an optically raised bar within the alignment zone.
Is it anywhere explicitly stated that the baseline is the mathematical rather than optical baseline? And if so, what is the best way to hint something that sits one or two units above the baseline that you want to align to the same height as things on or slightly below the baseline?

2026-03-08 23:40:07 UTC

- reply

@nprofile…e9hc @nprofile…ale9 @nprofile…ahkr Metal is really interesting. I know people who find it soothing and conducive to concentration, but I am very much not wired that way.
There’s some 90s industrial music for which I retain a nostalgic appreciation, but I can’t listen to it while working.

2026-03-08 20:18:36 UTC

- reply

@nprofile…e9hc THEM DID NOT SHOW ME THE MONEY

2026-03-08 17:20:58 UTC

Just received a charming email reading
‘| AM TRYING TO GET THE LATEST VERSION OF BRILL. GIVE ME THE FONT RIGHT NOW BEFORE I DESTROY YOU!’

2026-03-08 16:18:43 UTC

- reply

@nprofile…ahkr I know you’re joking, but I am genuinely interested in the different kinds and tempos of music that my colleagues listen to while working. Personally, I don’t want anything too frenetic or beat heavy.
I often sing along.

2026-03-08 16:03:21 UTC

- reply

@nprofile…ale9 I often listen to audiobooks or podcasts when working on outlines, but for kerning I find music better. I suspect this is because kerning is more of a rhythmic activity than outline editing.

2026-03-03 17:17:43 UTC

I find the ways in which AI fails a lot more interesting and revealing than the black boxes of its occasional successes.
https://typo.social/system/media_attachments/files/116/166/358/173/951/491/original/a69e6021435c0297.png

2026-02-27 20:14:17 UTC

Months ago, I planned and booked a trip to the UK in March, with plans to attend the @nprofile…dlf3 WAVE conference and meet with Fiona and the North India curator at the British Library, as well as seeing family, friends and colleagues.
Now looking at cancelling the whole trip because I don’t have time to meet the new UK dual national requirements, which we’re only just hearing about in Canada.
Even the nuclear option of renouncing my UK citizenship would take too long (and probably not something to be done while this grumpy).

2026-02-26 23:02:02 UTC

- reply

@nprofile…nr8a I find discovering designers and foundries to be more interesting and enjoyable than looking for particular kinds of typefaces. Back in my typesetting days, I remember regularly thinking, ‘I wonder if [designer] has a typeface that would be good for this project?’ Remarkably often, they would have recently released something new that was perfect for the project.
Perfect, of course, might mean going in a totally unexpected direction, which I think is one of the joys to which you refer. The trouble with using AI or even filters and categories is that they presume you have already decided what you are going to do or, worse, are willing to accept the guidance of cliche.

2026-02-26 21:58:12 UTC

- reply

@nprofile…nr8a
“Finding the right font shouldn’t feel like searching for a needle in a haystack.”
So maybe stop burying the individual needles in an ever-expanding haystack?

2026-02-16 05:04:56 UTC

- reply

@nprofile…aswa @nprofile…cm8h I’ve lost track of the number of times I have had analyse text in random 8bit font encodings to convert to Unicode. The Indic ones are always fun, because reordering rules need to be defined along with the mappings.

2026-02-09 20:35:09 UTC

- reply

@nprofile…dfhn You can probably find the show in a podcast app.

2026-02-09 18:39:54 UTC

Shape kerning arbitrary half form sequences for what Fiona calls the ‘cat running over the keyboard’ scenario.
https://typo.social/system/media_attachments/files/116/042/117/160/011/282/original/6b6550008e6e6539.png

2026-02-09 00:52:11 UTC

@nprofile…dfhn Your font gets a couple of shout-outs in the Hypatia episode of the BBC Radio 4 ‘You’re Dead to Me’ series.
https://www.bbc.co.uk/programmes/m002r38z

2026-02-05 22:42:21 UTC

I’m on Ryobi (which makes it sound like a high blood pressure medication). The batteries and chargers do change periodically, but they remain compatible with the tools I have. I am strangely resistant to having more than one charger and brand of battery.
Interestingly, there is a Festool Canada website, but I have never seen them in any hardware stores here.

2026-02-05 20:53:29 UTC

- reply

@nprofile…e9hc I’m stuck on the power tool doom slide: ‘Buy the one that works with the same battery pack as all your other power tools, even though each is progressively worse reviewed than the previous.’

2026-02-03 03:27:11 UTC

- reply

@nprofile…aane Can I nominate a pair of values? Has to be
usLowerOpticalPointSize
usUpperOpticalPointSize
To my knowledge they were implemented in a single font family, were never used in software even for that family, and were almost immediately superseded by another technology.
But TWIP is still my favourite typographic unit.

2026-01-29 18:02:40 UTC

Does UFO include some way to record Unicode variation selector sequences, from which a format 14 cmap subtable can be written by tools processing the UFO?

2026-01-19 20:12:59 UTC

- reply

@nprofile…4t5m @nprofile…wguy Thank you for making it possible to stop it spinning!

2026-01-14 18:10:41 UTC

- reply

@nprofile…zk9d That confirms my guess that emoji handling is a special case in the font stack, and the use of variation selectors after emoji-(fiable) characters function as flags rather than something that needs to be implemented at the font level (unless a font contains both emojified and non-emojified glyphs).

2026-01-14 16:02:51 UTC

- reply

@nprofile…zk9d Interesting. I wonder if software is smart enough to fallback sensibly in this situation without needing the VS cmap mapping? I have seen examples of software falling back to an emoji font if the non-emoji VS is missing from text, but I would assume the non-emoji VS characater should act like a flag for software, rather than something that needs mapping in every font.

2026-01-13 23:53:47 UTC

- reply

@nprofile…zk9d Short answer: format 14 cmap subtable.
Long answer begins ‘What tools are you using?’

2026-01-13 23:49:35 UTC

- reply

@nprofile…grsj @nprofile…qecv @nprofile…y4xw Given how many people instinctively apply sandhi to elide s-s to s when saying it, I’d favour sanserif as the compound. But I may be influenced by Sanskrit spelling rules, which I’ve been writing about this week.

2026-01-08 19:55:05 UTC

- reply

@nprofile…w44f Otto Hupp’s Liturgisch for Klingspor (1906) has narrow caps to facilitate the dense typography and narrow columns of Bible and liturgical book setting. There’s a digital version by Dieter Steffmann from the early 2000s.

2026-01-04 00:11:16 UTC

Just finished reading _Chinese Characters Across Asia_ by @nprofile…3u23. It’s the clearest introduction I have read about how the Chinese writing system works—dispelling common, long-held misunderstandings—, and how the script was adapted to Korean, Japanese, Vietnamese, and Zhuang. Recommended.
https://typo.social/system/media_attachments/files/115/833/909/977/783/086/original/941e42efa5385b66.png

2025-12-31 19:46:55 UTC

- reply

@nprofile…l0j5 I have done laundry there.

2025-12-30 22:11:47 UTC

- reply

@nprofile…wtpt @nprofile…d6ff The last book I read was _Hanns and Rudolf_, so it would be either the commandant of Auschwitz or the war crimes investigator who tracked him down and captured him. I’d opt for the latter, obviously.

2025-12-19 19:40:56 UTC

- reply

While we’re on the subject: if anyone has a spare copy of Typography Papers 8 that they would like to sell, please message me.

2025-12-19 19:30:08 UTC

I love how completely random even the existence of Typography Paper seems: less a journal with an irregular and infrequent publishing schedule, more of a magical alignment of signs that flash into being and then disappear again.
https://typography.network/typographypapers10/

2025-12-14 17:48:04 UTC

- reply

@nprofile…t9uc The one in the Birmingham Museum of Art seems better/wetter.
https://typo.social/system/media_attachments/files/115/719/167/681/107/205/original/ec1456e033611774.png

2025-12-14 04:40:45 UTC

- reply

@nprofile…7kps In this case, it’s instantiating nominal fonts and picking up their metrics, but failing to correctly instantiate the glyph shapes and instead falling back to the default instance glyphs.
https://typo.social/system/media_attachments/files/115/716/070/915/660/870/original/6f9da10517d93152.png

2025-12-13 17:37:05 UTC

- reply

@nprofile…n60g @nprofile…qppg Just noticed that the Export PDF with the broken font embedding is 1.4MB, while the Print PDF with the correct embedding is only 257 KB. I am guessing that the complete roman and italic variable fonts are being embedded.

2025-12-13 00:59:29 UTC

- reply

@nprofile…n60g @nprofile…qppg HOWEVER...
If I use the print function, and the ‘Microsoft Print to PDF’ function, I get correct output in the resulting PDF.
But the print-to-PDF output doesn’t preserve the dimensions of the slides: you have to define a paper size (!), and I’m not sure a large international company can rely on all users remembering never to use the export function and always remembering to set special page size to match slide dimensions.
https://typo.social/system/media_attachments/files/115/709/544/606/709/771/original/90a89785fee5187a.png

2025-12-13 00:51:13 UTC

- reply

@nprofile…n60g @nprofile…qppg Powerpoint on Windows (Windows 11 Pro and Office 365 latest release versions). I did my own tests today. Test set in MS system version of Sitka variable font. Banner optical size subfamily members selectable in PPT font menu, and display correctly in the slide. Used Export ‘Create PDF/XPS’ function, with PDF ‘Standard’ option. Result is as reported by client.
https://typo.social/system/media_attachments/files/115/709/509/087/141/518/original/d44f7d0c4140cf31.png

2025-12-12 18:40:12 UTC

- reply

@nprofile…qppg The unexpected default instance is only in the PDF it seems. The Banner variant is selectable in the PPT font menu, and the client reports that it displays correctly in their slides.

2025-12-12 18:37:23 UTC

- reply

@nprofile…qppg It seems to support named instances. Microsoft switched from shipping Sitka as 36 static fonts to two variable fonts. The named instances show up in Office apps, including PPT, as if they were static fonts. Backwards compatibility with the static fonts was a major concern for MS when making the variable version.

2025-12-12 17:05:57 UTC

Am I right in thinking that PDF still doesn’t play well with variable fonts? Specifically, I have a client running into problems exporting a PDF from Powerpoint using Sitka variable fonts and getting the default instance on the nominal Banner opsz metrics.
https://typo.social/system/media_attachments/files/115/707/676/078/619/896/original/175493b902980e84.png

2025-12-10 03:36:26 UTC

- reply

@nprofile…zaf0 Please read the typo.social instance rules regardoomg advertising
https://typo.social/privacy-policy#advertising-standards
and remember to include the ad hashtag when posting such content. Thank you.

2025-12-05 21:08:35 UTC

My favourite Unicode script encoding proposals are the ones that begin ‘The script is based on a vision the inventor received in a dream’. They make me wonder what lost stories people might once have told about the creation of the Latin script.

2025-12-04 19:27:27 UTC

- reply

@nprofile…auj3 What does SF have for U+1D1A?

2025-11-30 00:56:40 UTC

- reply

@nprofile…66h2 I’ve not played D&D since some time in mid 1980s, but lately I have been falling asleep listening to this guy’s spectactularly nerdy deep dives into the early history of desktop role-playing games.
https://www.youtube.com/@daddyrolleda1

2025-11-25 17:01:02 UTC

- reply

Laconia Cherokee is a companion to the Laconia typeface family, designed by Ross Mills, with assistance from Anna Štepanovská and Paul Hanslow. Laconia supports a wider range of Latin script languages.
https://www.tiro.com/fonts/laconia
#ad
https://typo.social/system/media_attachments/files/115/611/391/501/468/287/original/210fed4b0963d317.png
https://typo.social/system/media_attachments/files/115/611/395/173/288/087/original/5e16a6d1b352a5ab.png

2025-11-25 16:45:26 UTC

- reply

In support of Indigenous language use, Laconia Cherokee is available for license under our pay-what-you-want model. You decide the price.
https://www.tiro.com/fonts/laconia-cherokee#purchase
#ad
https://typo.social/system/media_attachments/files/115/611/335/756/036/495/original/ff51535fa24d71e1.png

2025-11-25 16:42:57 UTC

New Laconia Cherokee specimen images, by Paul Hanslow. See these and more at
https://www.tiro.com/fonts/laconia-cherokee
#ad
https://typo.social/system/media_attachments/files/115/611/301/050/455/853/original/8f19b3e3e640b007.png

2025-11-20 02:08:07 UTC

- reply

@nprofile…gh9h So what was supposed to be optional (‘if’, ‘may’) and legacy-oriented (‘...arbitrary instances of a variable font must be exported as a nonvariable font for legacy applications and for printing’), ended up becoming a de facto requirement for any variable font, hence enshrined in FontBakery checks etc.
The added requirement that roman and italic variable fonts have distinct ID 25 entries in order to work in InDesign seems to further undermine the original intent.

2025-11-19 15:57:25 UTC

- reply

@nprofile…gh9h I presume you are testing in Adobe apps? Distinct name ID 25 entries—indeed, presence of ID 25 names at all—is not a requirement of the OpenType spec, but something imposed by Adobe’s daft implementation.

2025-11-07 22:32:02 UTC

I should have kept track of all the time I’ve spent designing glyphs for the long vocalic ḹ in various scripts. I’ve chosen to consider it a form of meditation.
https://typo.social/system/media_attachments/files/115/510/760/461/905/329/original/0f0547f18f894e99.png

2025-11-03 17:06:46 UTC

Riccardo Olocco’s lovely tribute to James Mosley.
https://fontstand.com/news/essays/james-mosley-19352025/

2025-10-28 15:31:36 UTC

- reply

@nprofile…06ag Struck but unfinished matrices?

2025-10-24 12:24:16 UTC

@nprofile…00ep explaining the distinction between machine learning and artifical intelligence.
https://typo.social/system/media_attachments/files/115/429/114/559/858/247/original/b6e83f7eae859f4c.png

2025-10-22 15:48:32 UTC

TIL that HP Lovecraft wrote _At the Mountains of Madness_ on the back of scrap paper, letters from friends (and the editors of Hairdressers’ News), notices from his bank, etc., and there is now a facsimile edition of the manuscript that includes all those verso pages.
https://www.spbooks.com/155-at-the-mountains-of-madness-9791095457152.html
https://typo.social/system/media_attachments/files/115/418/586/020/470/482/original/cc437473621bbbfb.png
https://typo.social/system/media_attachments/files/115/418/591/914/556/143/original/65ec9cb47169aa11.png

2025-10-22 14:55:50 UTC

- reply

@nprofile…r44m @nprofile…fxs0 @nprofile…qppg The featureparams are mapped in the GSUB table under the common OTL script->langsys-feature hierarchy, which means that they share the same independent behaviour: if a feature has different functionality for different scripts, it makes sense that it also has different labeling.
At the name table level, these are just separate strings, which may or may not be localised (most often not).

2025-10-22 14:50:02 UTC

- reply

@nprofile…r44m @nprofile…fxs0 @nprofile…qppg
We’re talking about featureparams, which is mechanism of name table strings mapped from the GSUB table for some specific OTL features (stylistic sets and character variants). Font makers can use featureparams to provide descriptive labels for these features, which software *may* display in the typography UI.
Like other name table strings, featureparams names can be localised using platform, encoding andf language identifiers.
https://typo.social/system/media_attachments/files/115/418/368/024/768/473/original/f737e7835b241fa4.png

2025-10-21 19:59:03 UTC

- reply

@nprofile…fxs0 @nprofile…qppg It should do ‘something’. It could show all applicable featureparams, or it could show whatever is associated with the DFLT script entry, or it could show a ‘Multiple stylistic substitutions’ message, or it could fall back, as InDesign and Wakamai Fondue now do, to a generic label.
I’m okay with any of those somethings. I’m not okay with assuming that something implicit in OpenType Layout architecture doesn’t occur.

2025-10-21 19:54:21 UTC

- reply

@nprofile…qppg @nprofile…k4zy v5.00 is the new variable-build derived version, and we updated our build tool at the same time to enable featureparam names to be added to VOLT-derived GSUB tables.
For some reason, Brill have not updated the version available under the no-fee non-commercial use license. So you can currently only get v5.00 under commercial license from Tiro or, possibly, the Apple distribution. [I’m not sure which version Apple is currently making available. They do have v5.00.]

2025-10-21 14:37:53 UTC

- reply

@nprofile…qppg @nprofile…k4zy Localisation of name strings is a different issue, though. In this case, the strings are all in English, but they differ across the script tags, because e.g. ss01 does different things in different scripts.

2025-10-20 21:10:25 UTC

New tools for manipulating FontLab .vfj (and MS VOLT .vtp) data:
vfj-skew.py
Applies slant (and optional x,y offsets) to anchors in a .vfj source.
vfj-copy-anchors-from-vtp.py
Copies anchors from an .vtp file (or .ttf containing a TSIV table) to a target .vfj file.
With thanks, as usual, to @nprofile…7kps
https://github.com/TiroTypeworks/TiroTools/tree/master/VFJtools

2025-10-20 21:02:51 UTC

- reply

@nprofile…k4zy As I guessed might happen: WF is returning only the stylistic set featureparam names that do not have multiple entries for different scripts, so for Brill Roman it is displaying names for ss05, ss06 and ss20, but not for ss01–ss04. This is similar to what Adobe InDesign does when confronted by this font.
Ideally, the name would be selected based on the script property of the text, and only fall back to a generic label if more than one script is present in the text/selection.
https://typo.social/system/media_attachments/files/115/408/504/163/303/043/original/b02aac09f5da05f1.png

2025-10-17 16:27:00 UTC

- reply

@nprofile…k4zy Have you considered exposing feature parameter names for stylistic set and character variant features?
https://typo.social/system/media_attachments/files/115/390/437/805/488/897/original/36072d0d11ed0b57.png

2025-10-17 16:21:07 UTC

- reply

@nprofile…k4zy The list of supported languages is very sparse and missing *many* languages that the font supports.
https://typo.social/system/media_attachments/files/115/390/415/819/366/445/original/0e3b3606044b345f.png

2025-10-16 23:09:01 UTC

What could possibly go wrong?
https://typo.social/system/media_attachments/files/115/386/352/665/832/236/original/4cdabc3e18c20cce.png

2025-10-16 22:57:34 UTC

- reply

@nprofile…2fal @nprofile…vk0r @nprofile…m8r8
ATypI does have a code of conduct now, but it is a fairly generic ‘don’t be a racist, sexist asshole’ code of conduct.
https://atypi.org/about-atypi/rules-regulations/code-of-conduct/
It says nothing about specific professional conduct relating to the type business, e.g. copying typeface designs, unlicensed font use, etc. So far as I know, the code of conduct is no part of the association statutes, in the way the old Code was.

2025-10-16 22:53:40 UTC

- reply

@nprofile…2fal @nprofile…vk0r @nprofile…m8r8
I’ve spoken with some people who think the Code should not have been simply dropped but replaced by some kind of code of professional conduct.
The problem I always had with the idea of ATypI as a ‘professional association’ is that no one could tell me what the profession was. People who designed typefaces? People who manufactured fonts? People who sold fonts? People who used fonts? They were all represented in the ATypI membership, along with non-professionals.

2025-10-16 22:50:26 UTC

- reply

@nprofile…2fal @nprofile…vk0r @nprofile…m8r8
I think dropping the Code, as it was, was necessary and a good thing. It was really an artifact of an earlier industrial model of organisation, and didn’t make much sense in the context of hundreds of small independent foundries. ATypI didn’t evolve into a standards organisation, but it did evolve, away from a business/professional organisation and more towards an academic/research conference organiser.

2025-10-16 22:43:03 UTC

- reply

@nprofile…2fal @nprofile…vk0r @nprofile…m8r8
The biggest push to review and reform the ATypI statutes came from Adobe’s legal department. Someone I had been talking with at Adobe suggested that maybe the association might become responsible for the OpenType format specification, so ran the idea past their lawyers. The lawyers looked over the ATypI statutes and said ‘This is a cartel; we can’t have anything to do with it.’

2025-10-16 22:38:43 UTC

- reply

@nprofile…2fal @nprofile…vk0r @nprofile…m8r8
Some missing context on the decision to ditch the Code Morale in 2004:
Cynthia Batty had been working on getting ATypI official non-profit status in the US, and I had been working on the possibilty of ATypI evolving into a standards organisation that could coordinate industry involvement in font formats, character encoding, and Web standards. The Code Morale was identified as problematic for these initiatives because it made ATypI look like a cartel.

2025-10-15 15:56:58 UTC

- reply

@nprofile…00ep You’ve made an explicit distinction between machine learning and artificial intelligence, so your dummy title is already more insightful than 90% of what’s said on the subject.

2025-10-15 03:29:16 UTC

- reply

@nprofile…82xt Paul Hanslow designed the Skeena and Skeena Display graphics (as well as the various specimens on the site). He does lovely work.

2025-10-14 23:24:17 UTC

The Univerisity of Reading quietly published a playlist of James Mosley’s lectures on the history of (Latin) letterforms to YourTube.
https://www.youtube.com/playlist?list=PLh8Q6TpPxtgxSG6DEuOrH9iKx9bh6Pr7G

2025-10-14 22:48:47 UTC

- reply

Check out these and other Tiro Typeworks fonts on our website, and be sure to look for the specimen PDF link at the bottom of each font page for more samples.
[Below: Gabriola, by John Hudson.]
https://www.tiro.com/fonts
#ad
https://typo.social/system/media_attachments/files/115/374/951/820/399/586/original/f0068041d19daf31.png

2025-10-14 22:46:17 UTC

2025-10-14 22:39:47 UTC

- reply

New releases include Skeena and Skeena Display, designed by John Hudson and @Paul_Hanslow for Microsoft Office, and now extended to pan-European Latin character coverage.
https://www.tiro.com/fonts/skeena
https://www.tiro.com/fonts/skeena-display
#ad
https://typo.social/system/media_attachments/files/115/374/908/951/849/734/original/99e61eec413d2538.png
https://typo.social/system/media_attachments/files/115/374/916/016/427/641/original/9a8e4b2db84d445d.png

2025-10-14 22:21:36 UTC

- reply

In addition to tyefaces created by John Hudson, Ross Mills, and Paul Hanslow, we’re proud to be the licensing source for Matthew Carter’s Sitka family of optical size fonts (available as 36 individual fonts, or as two OpenType variable fonts).
https://www.tiro.com/fonts/sitka
#ad
https://typo.social/system/media_attachments/files/115/374/845/239/014/725/original/e7122afda3fdf03f.png

2025-10-14 22:16:43 UTC

- reply

Microsoft fonts available from Tiro Typeworks include older designs like Gabriola, Cambria, and Constantia, as well as the latter’s brand new companion type: Constantia Display.
https://www.tiro.com/fonts/constantia-display
#ad
https://typo.social/system/media_attachments/files/115/374/824/935/229/137/original/9665aeb0044a84c7.png

2025-10-14 22:10:14 UTC

Select fonts that we’ve made over the years for Microsoft are now available for commercial licensing from our website (including webfont use licenses). The first tranche of these fonts are now live, with more to follow later this year.
#announcement
https://www.tiro.com/articles/microsoft-fonts

2025-10-07 20:13:11 UTC

- reply

@nprofile…90r6 @nprofile…fpqe I’m not sure whom I writing for, either.
I think for the first one, I am going to keep it simple: announce the things and provide some links.

2025-10-07 20:01:12 UTC

- reply

@nprofile…jlpt I’m leaning towards the latter. not least because the SVG graphics on the website look so much better than the PNG graphics in emal.
Feeling my way—after 30 years in this business!—in the unfamiliar world of retail font marketing.

2025-10-07 19:58:55 UTC

- reply

@nprofile…x70x Thanks, Ivo. Very insightful.

2025-10-07 19:39:03 UTC

- reply

@nprofile…qx50 Okay, a different question, for everyone:
What are some of your favourite foundry newsletters?