What on earth possessed me... (2008) | Hacker News Reader

What on earth possessed me... (2008)

fontforge.github.io

105 pointsevolve2k34 comments

> In July a friend of mine, who is a mac user, said she wouldn't even consider looking at fontforge on her mac unless it behaved more like a mac application. So I figured out how to build a mac Application... > > My friend still (November) has not looked at fontforge. Ah well.

This is a great example why user feedback needs to be taken in close proximity to a salt mine.

We, or at least most of us, being human, have a horrible job anticipating our future emotional state. So when his friend said that she'd be interested if FontForge was available on her Mac, she probably meant it, but her request didn't reflect the underlying reasons she held for not using the software.

Not out of dishonesty or malice, but because she probably didn't know herself. It's like all those times somebody swore they would love to have an app that frobnitzed the foobars, but when said app became available, it was met with a chorus of shrugs and mehs.

There's an excellent book on this topic -- The Mom Test -- that anybody building products should read.

Reminds me of one day in the 90s I was out jogging with a colleague. He said that what the world needed was a Java compiler that compiled to native code. He needed one right now, and I should work on it.

I said it was a great idea, and I had already implemented one. He could start using it today.

He promptly lost all interest in it.

(It turns out, on the complex plane the demand for native Java compilers is on the imaginary axis.)

Did he say what he needed it for, or was it a vague aspiration?

Indeed, gcc recently nuked its support for this due to lack of interest (by users or maintainers).

That only happened, because the majority of GCJ developers left when OpenJDK was announced back in 2009, and almost no one cared about it since then.

Companies selling commercial native compilers for Java are still in business since the early days.

The follow-on question to an ask like that is nominally a variation on "So, what would that enable you to do?"

From there, I can then start digging into what the real problem is. Maybe it was a desire for greater performance, or easier deployability. Maybe his concern was about protecting intellectual property, and he thought that binaries would be harder to reverse-engineer than bytecode.

All of those could be legitimate problems, and none of them would really be solved by a Java-to-native compiler.

https://www.aicas.com/cms/en/JamaicaVM

There are/were a couple others, but the customer base is very... exclusive

> It turns out, on the complex plane the demand for native Java compilers is on the imaginary axis.

Quite the contrary.

Excelsion JET, IBM, PTG, Aicas have supported native Java compilers since the early days.

Oracle added AOT compilation for Linux x64 to OpenJDK on Java 9, with the remaining platforms being supported later. The Java 10 fork, already has Windows and macOS support as well.

The Graal team recently announced the Polyglot VM, which is actually another AOT compiler for Java.

The demand for native Java compilers is going quite thanks to Fintech and embedded development.

Indeed, Excelsior JET is really a fantastic product. It's a really straight forward, low friction way to deploy Java desktop applications. Their commercial license prices have risen precipitously in recent years, indicating they must be doing quite well for themselves. It's been a few years since I was a user, but I was very please with the product.

@Walter: did you generate code form the AST or from bytecode? Seems like a lot of the solutions around now generate code from the bytecode, my understanding at least.

From the AST.

The trouble with a native compiler is dealing with dynamic class loading.

Yeah, I'd leave that out if I were doing it :) otherwise you pretty much have to include an interpreter as well.

Including the interpreter, and then the JIT, made a native compiler pointless.

But that's all beside the point, which is people will often tell me what they want in a product, but don't actually want it.

> So when his friend said that she'd be interested if FontForge was available on her Mac, she probably meant it, but her request didn't reflect the underlying reasons she held for not using the software.

Nitpicking, but that's not what OP said; they said she "wouldn't even consider looking at fontforge on her mac unless it behaved more like a mac application". That's different from affirming that she would be interested in it if it looked native.

Another subtle is about asking not just "Would you use it" but "Do you already use something like this?". I bet if his friend is already interesting in fonts they already use something else.

Also friends are not reliable product testers. They might feel guilty saying "this sucks" or "there is better stuff than this". They just don't want to hurt you feelings. Even if it is someone you just met at a conference or even on the street if you talk with them long enough, you've already formed a rapport and they probably wouldn't want to upset you.

Well it's 2017 now and FontForge still doesn't behave like a Mac app (it uses X11 still) so the friend seems to have an excellent justification.

Thus sounds more like issue of not listening to user and replacing her response with own wishful thinking.She did not said she will use it if it looks like mac.

She was saying no and identified one easy to see thing she dislikes. There is no implication that it is only reason or that she has general interest in it. There is nothing to imply that she would even find the app/topic useful.

By anything measurable, the "I don't even look at apps that don't looks mac" have been absolutely honest and true. It did not however meant "the only reason I don't use it is UI look". That part was wishful thinking.

The account wasn't that his friend would use the software if. It was wouldn't use the software unless. Those responses imply different things.

Semi-related, I recently got into type design as a hobby.

I've primarily been using an app called Glyphs https://glyphsapp.com/ as well as FontLab VI https://www.fontlab.com/font-editor/fontlab-vi/

Both are excellent. FontForge is pretty good but I found it slightly overwhelming. I found it's great if you need to fix parts of existing fonts though.

There are some excellent guides out there on how to use FontForge. I really liked this guide about adjusting for optical illusions: http://designwithfontforge.com/en-US/Trusting_Your_Eyes.html .. I found it kind of mind blowing.

For those interested, I'm currently working on a 19th century French Didot style font, here's my progress: http://dn.ht/didhot/test.html .. feedback welcome. :)

AN and BW kern a little loosely, it looks like? :)

Good work!

Thanks! Yeah heh.. it turns out the spacing and kerning is probably more work than the lettering.

I've also been looking at this long enough now that I often can't tell what looks "normal" anymore.

I'm wanting to get into type design also.

Any recommendations for good beginner resources/ philosophies / courses to get started on the right track?

The fontforge guide is a really good start: http://designwithfontforge.com/en-US/index.html

Also, there are some great tutorials on the Glyphs website: https://glyphsapp.com/tutorials/drawing-good-paths

Sometimes I've asked for advice on type design forums such as http://typedrawers.com/ .. be prepared for possibly blunt but honest and useful feedback. ;-)

I tried fontforge once. Frankly, I couldn't figure it out how to get it to work after playing with it for a few hours. My dreams of building custom programming-oriented ligatures for a proportional font are still unrealized.

Some time ago in gamedev subredit someone asked for font with ligatures, I made minimal working sfd file (internal format of fontforge) from which you can generate ttf font:

https://www.reddit.com/r/gamedev/comments/6nrpvf/how_would_i...

the sfd is at the bottom of the thread (in the middle of thread is how to make custom ligatures). But I agree that FontForge is horrendous (and that comes from someone who made moderately sucessfull (50k downloads) font with it)

You get what you pay for. I haven't worked with it for a few years now but I found it to be quirky and unstable. But it was free, which was better than the $650 or so that you needed for FontLab, which was the main piece of software for font creation out tehre at the time, and it seemed to have a more features than I needed for what I was working on. Cross-platform was a nice feature as well.

No one has really documented the OSS process for modifying and deploying a font yet, though supposedly it exists.

As an aside, the fragment of Shakespeare quoted at the start is a famously dirty joke. (The fact that the two characters quoted both seem blithely unaware of what they are saying is a further joke; both are being laughed at.)

I briefly hoped that this would be a link to

https://github.com/search?utf8=%E2%9C%93&q=%22what+on+earth+...

Title gore

Yes, and don't get George srarted about Chain Ringing! It may be the only thing he loves more than fonts... or lemurs.

In all seriousness, he's done great work on open source projects he loves... as a truly great software engineer. All paid for by a the risk he took working for a Start-up that was acquired by AOL before the big boom.

Do you mean Change Ringing ?

Yes, with bells! Didn't catch the mobile autocorrect :( https://en.m.wikipedia.org/wiki/Change_ringing

> Within a month I had received my first bug report, and presumably had my first user.

This isn't really discovering Agile before it's time, but maybe some people think it is.

Reply on news.ycombinator.com