I love FontForge. It is what I always use for creating fonts. I tried
FontStruct once but found it highly unsatisfactory (not least because of the
geometric limitations and the licensing of the resulting fonts). I know next
to nothing about creating fonts, but for the simple things I need (letters &
diacritics, but no ligation et cetera) it works fine.

—Calculator Ftvb

P. S. I'm pretty sure http://www.kreativekorp.com/ucsur/ is not the real
CSUR; I think it is an extension of the main one (at
http://www.evertype.com/standards/csur/), as is explained at the
http://www.kreativekorp.com/ucsur/ page. I'm not sure about that though so
feel free to correct me if I'm wrong.

On 26 June 2011 00:06, Rebecca Bettencourt <[log in to unmask]> wrote:

> On Sat, Jun 25, 2011 at 7:30 PM, yuri <[log in to unmask]> wrote:
> > Hi Conlangers,
> >
> > I'm new to the list. I started creating my own conlang about 15 years
> > ago but it got put on the backburner. I'm resuming the project.
>
> Hi Yuri!
>
> > I'm interested in what software people recommend for creating a font
> > for my script.
> >
> > I have installed fontforge and inkscape, with the idea of creating the
> > glyphs in inkscape and importing the SVG for each glyph into
> > fontforge.
>
> For the price, you can't beat FontForge. It can do everything you
> need, though it gets a bit technical. A tip if you're going to be
> importing from SVG files: put a rectangle of a constant height around
> each glyph so they remain the same size when imported, then remove the
> rectangle after importing.
>
> The only other font software I've used is something called ScanFont,
> but only just to trace outlines from a scanned image; it's pretty
> terrible at actually producing a valid font. :P
>
> > Also, I understand the unicode address space includes areas for
> > private use. Can anyone tell me more about that?
>
> Yes, code points E000 to F8FF and F0000 and above are set aside as
> private use. That means Unicode itself won't ever define what those
> code points mean; whoever is exchanging text using those code points
> has to come up with their own agreement as to what those code points
> mean.
>
> The conlang community has their own agreement about what private use
> code points mean. It's called the ConScript Unicode Registry. More
> information is here:
>
> http://www.kreativekorp.com/ucsur/
>
> If you'd like some Unicode code space, find some unassigned space in
> that list, put together a proposal including a list of all the
> character names and the code points you'd like to assign them, and
> then I can add it.
>
> > One required feature is the ability to compose any combination of base
> > letter and diacritics.
> > I'd like to avoid having to have a font full of precomposed characters
> > because the diacritics above each base letter is the vowel, and the
> > diacritic below the base letter indicates voiced/unvoiced and
> > aspirated/unaspirated, so the combinations could run into the
> > thousands.
> >
> > The base letters themselves are all consonants.
>
> If the base letters are all the same width, this is easy. The
> diacritics are just zero-width characters with the glyph to the left
> of the origin. If they're all different widths, though, you'll need to
> do a lot more work, but it can be done. Something about GPOS tables or
> something; I've actually never done it before.
>
> -- Rebecca Bettencourt
>