#jogamp @ irc.freenode.net - 20140519 05:05:15 (UTC)
20140519 05:05:15 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20140518050515.html
20140519 05:05:15 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20140519050515.html
20140519 05:07:37 * hija (~hija@anon) Quit (Ping timeout: 265 seconds)
20140519 05:07:47 * hija_ (~hija@anon) has joined #jogamp
20140519 06:05:05 * guardianp (~gp@anon) has joined #jogamp
20140519 06:30:58 * monsieur_max (~maxime@anon) has joined #jogamp
20140519 06:34:13 * zzuegg (~zzuegg@anon) has joined #jogamp
20140519 06:38:40 * guardianp is now known as petmonster
20140519 07:02:10 * petmonster (~gp@anon) Quit (Ping timeout: 265 seconds)
20140519 07:53:45 * hija_ (~hija@anon) Quit (Quit: hija_)
20140519 08:17:39 * guardianp (~gp@anon) has joined #jogamp
20140519 08:35:25 * hija (~hija@anon) has joined #jogamp
20140519 09:27:22 * guardianp is now known as pet
20140519 12:49:45 <xranby> good day all, drawing glyphs for chinese, japanese and korean typefaces in small sizes is a huge real word problem for mobile.. this alone would make siggraph asia an eyeopener
20140519 12:49:51 <xranby> since it can be solved by graph!
20140519 12:51:00 <xranby> http://www.reddit.com/r/programming/comments/25vca1/androids_font_renderer_efficient_text_rendering/chlhfd9 <- tizen and android debate all focused on solving it using textures
20140519 12:56:37 <xranby> http://www.zerohedge.com/news/2014-05-16/where-worlds-unsold-cars-go-die - someone sit on waaaay to many cars
20140519 13:54:30 <hija> @xranby so how does font-rendering in graph compare performance-wise compared to the texture solutions?
20140519 13:55:43 <hija> there is an obvious gain I guess that you can have any resolution glyph with vectors
20140519 13:56:11 <bbbruce> but then you have antialiasing issues
20140519 13:56:23 <hija> with the vectors?
20140519 13:56:39 <bbbruce> yes
20140519 13:56:49 <hija> i guess
20140519 13:58:14 <hija> does the effect get any better with high-ppi displays?
20140519 13:58:57 <hija> any idea bbbruce?
20140519 14:05:57 <sgothel> on modern higher dpi devices there are no AA issues, as tested on a 10" 1080p display
20140519 14:07:01 <sgothel> on lower dpi devices our brute-force AA is ok'ish (FBO/AA) by now
20140519 14:08:00 <sgothel> (4-8 'oversampling')
20140519 14:11:30 <xranby> i can start looking into font hinting
20140519 14:11:56 <xranby> see what is required to make fonts with added font hinting work
20140519 14:12:04 <xranby> the small tweaks to make it look really good
20140519 14:12:21 <sgothel> whether a 'line' shall be in the raster ?
20140519 14:12:43 <sgothel> this won't work in our framework right now, since we cache the outlines
20140519 14:13:06 <xranby> hinting is as i understood crucial for asian fonts.. yes http://www.freetype.org/autohinting/hinter.html
20140519 14:13:09 <sgothel> we operate more like Apple's TTF rendering w/o hinting
20140519 14:13:37 <sgothel> where MS is grid-fitting .. and hence the width of text varies according to output device
20140519 14:14:52 <xranby> i will look at how freetype does it in their experimental branch
20140519 14:14:59 <xranby> and try make a conclusion
20140519 14:15:12 <sgothel> freetype does both .. grid-fitting + subpixel .. depends on setting
20140519 14:16:28 <sgothel> we produce an Outline[Shape] from each glyph .. then latter translate (move) and transform (zoom) it for each 'string' to save performance
20140519 14:16:37 <xranby> as a first step i will try render som asian fonts and see how it looks
20140519 14:17:00 <xranby> especially when scaled down to small sizes
20140519 14:17:07 <sgothel> that is a good idea .. add an asian font ttf file to demos + unit test
20140519 14:17:25 <sgothel> hopefully there is one avail w/ a good license
20140519 14:17:29 <xranby> maybe the ubuntu font includes it already?
20140519 14:17:32 <xranby> hmm lets see
20140519 14:17:33 <sgothel> nope
20140519 14:17:35 <xranby> oh
20140519 14:17:59 <sgothel> CMYK .. special ttf files AFAIK
20140519 14:18:16 <sgothel> err .. CJ.. well, you know :)
20140519 14:18:57 <xranby> CJK
20140519 14:19:06 <xranby> chinese, japanese, korean
20140519 14:19:16 <sgothel> the big ones .. right :)
20140519 14:20:40 <xranby> i will try this testcase for a starter 黑幣
20140519 14:21:22 <xranby> "black currency"/"black dollar"
20140519 14:22:26 <xranby> ask qun if i got it right
20140519 14:22:50 <xranby> the idea is to name currency in a way that do not instantly become money laundering or black market ;)
20140519 14:25:17 <sgothel> I see where this requires grid fitting now .. *putting my glasses on* :)
20140519 14:27:24 <xranby> many strokes much compact :)
20140519 14:39:30 <xranby> ok the 黑幣 world wide adoption program has started http://www.businesswire.com/news/home/20140519005859/en/Max-Borges-Agency-Public-Relations-Firm-Accept#.U3oXiobcbh1
20140519 14:42:42 <xranby> me and my wifes yoga school has started to accept blackcoin and that why i am inclined towards it! http://www.gongiversum.com/en/blackcoin/accepted/
20140519 14:43:34 <xranby> and while doing it i am exploring how to broad the use thus mobile applications are needed with asian text hinting working
20140519 14:43:41 <xranby> thus graph!
20140519 14:44:15 <sgothel> no objections Sir .. (== no killing machines :)
20140519 14:44:23 <xranby> awesome
20140519 14:46:07 <sgothel> one day .. maybe we can mix & match all programing languages easily .. like llvm+vmkit+.. so we don't need to care about barriers
20140519 14:46:40 <xranby> oh yes.. i have to read up in the logs if you have discussed the oracle vs google verdict
20140519 14:46:48 <xranby> and its implications
20140519 14:47:18 <sgothel> sure, discussed w/ Mark a bit
20140519 14:48:51 <xranby> soo.. is amd now out of busniess.. the x86 instructionset being an api copy of intel etc?
20140519 14:49:17 <sgothel> they pay .. and have an xchange w/ Intel (x86_64 etc)
20140519 14:50:03 <sgothel> again 'fair use' is not yet out of the loop, addressing public API etc ..
20140519 14:53:59 <xranby> how much of the api can one use before it is unfair? :)
20140519 14:54:48 <xranby> use all bytecodes except nop ?
20140519 14:55:05 <sgothel> Well, all EU citizen may vote against the EU end of May to safe freedom - i.e. vote against TTIP etc ..
20140519 14:56:00 <xranby> right!
20140519 14:56:12 <xranby> err vote P!
20140519 14:57:01 <sgothel> my problem here is, there is no easy party to vote for to reduce the EU powers, and bring back democracy
20140519 14:57:35 <sgothel> so I have to byte on my tongue .. and make a compromise
20140519 14:58:39 <sgothel> like UKIP/Farrage sort of parties .. while I like their sentiment to reduce EU powers, the rest they may proclaim is sort of .. difficult :)
20140519 14:59:15 <sgothel> then .. we don't even have such a party in Germany .. it's all 'great'
20140519 14:59:34 <sgothel> Pirates/Left/.. they all support the EU as-is
20140519 15:00:28 <sgothel> best: we all follow the Swiss model
20140519 15:10:49 <sgothel> afk .. (garden work) .. laters
20140519 15:51:23 * xranby (~xranby@anon) Quit (Ping timeout: 252 seconds)
20140519 16:12:12 * pet (~gp@anon) Quit (Quit: liteIRC for Android)
20140519 17:17:11 * rmk0 (~rmk0@anon) Quit (Ping timeout: 258 seconds)
20140519 17:28:08 * hija (~hija@anon) Quit (Quit: hija)
20140519 17:30:00 * monsieur_max1 (~maxime@anon) has joined #jogamp
20140519 17:34:19 * rmk0 (~rmk0@anon) has joined #jogamp
20140519 17:34:19 * rmk0 (~rmk0@anon) Quit (Changing host)
20140519 17:34:19 * rmk0 (~rmk0@anon) has joined #jogamp
20140519 17:36:53 * odin_ (~Odin@anon) Quit (Quit: Leaving)
20140519 18:24:47 * [Mike] (~Mike]@anon) Quit ()
20140519 19:33:29 * [Mike] (~Mike]@anon) has joined #jogamp
20140519 20:24:11 * [Mike] (~Mike]@anon) Quit (Ping timeout: 252 seconds)
20140519 21:15:40 * monsieur_max1 (~maxime@anon) Quit (Quit: Leaving.)
20140519 21:20:23 * [Mike] (~Mike]@anon) has joined #jogamp
20140519 21:59:31 * hija (~hija@anon) has joined #jogamp
20140519 22:37:10 * hija (~hija@anon) Quit (Quit: hija)
20140519 22:39:04 * hija (~hija@anon) has joined #jogamp
20140519 22:52:40 * rmk0 (~rmk0@anon) Quit (Ping timeout: 240 seconds)
20140520 01:42:07 * guardianp (~gp@anon) has joined #jogamp
20140520 01:44:07 * guardianp is now known as pet
20140520 02:02:46 * pet (~gp@anon) Quit (Quit: liteIRC for Android)
20140520 02:03:05 * guardianp (~gp@anon) has joined #jogamp
20140520 03:00:37 <sgothel> computing discrete logarithms: from O(c^n) -> O(n^c) /* exponential -> polynomial */ http://arxiv.org/pdf/1306.4244v2.pdf (public private key .. Diffie Hellman etc) .. interesting
20140520 03:25:55 * guardianp (~gp@anon) Quit (Quit: liteIRC for Android)
20140520 03:27:06 * kermyt (~kermyt@anon) Quit (Ping timeout: 240 seconds)
20140520 03:30:19 * kermyt (~kermyt@anon) has joined #jogamp
20140520 04:49:48 * hharrison (~chatzilla@anon) Quit (Quit: ChatZilla 0.9.90.1 [Firefox 29.0/20140428120048])
20140520 05:05:15 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20140520050515.html