Project Seen(projectseen.com)
projectseen.com
Project Seen
http://projectseen.com/
23 コメント
This website does the worst job explaining what it's even here for. I have to spend ten minutes reading obfuscated to text to understand that the font shows different glyphs when I type a "dangerous" word? Why do this in a font? Why don't you build a spellcheck dictionary and misuse that technology instead of this one that's in front of me all day?
It was not that bad, sometimes I don't like information presented in A, B, C format and like to figure things out. I enjoyed it.
The 'a' glyph annoys me, the rest of the glyphs are rounded.
It would be nice if a few of the people upvoting this came into the comments section to explain what they liked about it. I personally find absolutely no value.
If you like this then check out the authors other work here: http://emilkozole.com/
"Tap the lines for terrorist activity.
Exfiltrate credentials using beach head and break the kill chain."
Nothing gets highlighted. Is this correct?
Exfiltrate credentials using beach head and break the kill chain."
Nothing gets highlighted. Is this correct?
> claims to protect against dangerous words
> fails to censor the name of that D&D-playing traitor Edward Snowden
0/10
> fails to censor the name of that D&D-playing traitor Edward Snowden
0/10
Changing the width is not a pleasant experience...
So it actually kind of highlight the words so they can be detected more easily?
Make sure to use strong SSL certs, otherwise how do I know its your website? There better be a little green lock in the upper left corner of my browser ;)
UK perspective: how come "NSA" is censored but "GCHQ" isn't?
The list is based on a publicly available list that was released by Homeland of National security in 2013, so the focus is on US list, with that being said it shouldn't be taken to literate since by recent reports NSA and GCHQ have identified more than 40.000 triggers. Project "Seen" wants to stimulate a conversation on this topic opposed to being an actual tool.
How can a font possibly affect automatic content scanning, is what I'm curious about.
Fonts can have ligatures, which represent several letters as a single glyph (https://en.wikipedia.org/wiki/Typographic_ligature).
There are several common ones for letter combinations like "oe" - perhaps this font simply registers longer ones for whole words?
Edit: if you're on a mac, download the .zip file and double-click on one of the .otf font files. It will open in Font Book: View->Show Font Info says it hs 451 glyphs, and View->Repertoire shows that there are indeed glyphs representing each of the censored words.
There are several common ones for letter combinations like "oe" - perhaps this font simply registers longer ones for whole words?
Edit: if you're on a mac, download the .zip file and double-click on one of the .otf font files. It will open in Font Book: View->Show Font Info says it hs 451 glyphs, and View->Repertoire shows that there are indeed glyphs representing each of the censored words.
Here is a nice article from a similar project that explains the trick in detail:
http://pixelambacht.nl/2015/sans-bullshit-sans/
http://pixelambacht.nl/2015/sans-bullshit-sans/
Exactly, so it can be used in any software that supports opentype format.
It cannot.
> This way it automatically highlights and points out the content prone to secret surveillance.
The font is designed to alert writer, not scanner systems. I think use of a font is a very clever idea: instead of supplying separate tool that is hard to integrate with other systems it reuses OS functionality. Utility value highly drops, but the concept is very cool.
Someone might get an idea for a much more sophisticated "font" system that would process the text behind the scenes. Imagine that WM provides html/markdown/pdf rendering capabilities and browser acts as a transport. In case of dirt simple WM (terminal) browsers could provide renderer plugins. The same browser core could be reused for pc/mobile/text, no need for complex sandboxing/multiprocess security and stuff.
> This way it automatically highlights and points out the content prone to secret surveillance.
The font is designed to alert writer, not scanner systems. I think use of a font is a very clever idea: instead of supplying separate tool that is hard to integrate with other systems it reuses OS functionality. Utility value highly drops, but the concept is very cool.
Someone might get an idea for a much more sophisticated "font" system that would process the text behind the scenes. Imagine that WM provides html/markdown/pdf rendering capabilities and browser acts as a transport. In case of dirt simple WM (terminal) browsers could provide renderer plugins. The same browser core could be reused for pc/mobile/text, no need for complex sandboxing/multiprocess security and stuff.
Cool! Should be treated more like art than a tool...