|
Angryhead posted:https://gfycat.com/WelltodoFelineBooby Makes sense, the "corpse" is the debris left behind and the loot is the tasty minerals that were contained the wall dwarf fortress style. The blood you should see your doctor about.
|
#
?
May 12, 2017 18:09
|
|
|
|
| # ? Apr 28, 2024 20:18 |
|
|
Scaramouche posted:Makes sense, the "corpse" is the debris left behind and the loot is the tasty minerals that were contained the wall dwarf fortress style. The blood you should see your doctor about. It's the blood of Gaia you industrialist monsters
|
|
#
?
May 12, 2017 18:47
|
|
|
well that's not right https://twitter.com/LuigiThirty/status/873990598264541184 (I wish I could capture the video some way other than my phone...)
|
|
#
?
Jun 11, 2017 20:50
|
|
|
Luigi Thirty posted:well that's not right You've gotten this far, surely you can figure that part out. 
|
|
#
?
Jun 12, 2017 15:17
|
|
|
Luigi Thirty posted:well that's not right That's pretty cool. I should break out my old Macs and try building something on them. I think I still have the LaMothe software rendering book somewhere around here, maybe a dithered raytracer instead?
|
|
#
?
Jun 14, 2017 16:23
|
|
|
I did eventually get it working  All you need is a copy of CodeWarrior. I'm pretty sure it runs on any PPC Mac.
|
|
#
?
Jun 15, 2017 06:35
|
|
|
I like libGDX. I really do, but the fact that the default behavior of their Vector and Matrix library is _IN PLACE MODIFICATION_ drives me loving insane. bez = foo.scl(0.1) means both bez AND foo are now 0.1*foo.   
|
|
#
?
Jun 15, 2017 07:05
|
|
|
Jo posted:I like libGDX. I really do, but the fact that the default behavior of their Vector and Matrix library is _IN PLACE MODIFICATION_ drives me loving insane. That's the worst when you're working on something for a long time, and you don't know it's affecting something else the entire time just because of your viewing angle when you test it.
|
|
#
?
Jun 15, 2017 17:06
|
|
|
I resurrected my NES emulator project and started implementing the audio unit. The first attempt at playing sound was less than successful https://www.youtube.com/watch?v=3q3o_x0ro04
|
|
#
?
Jun 15, 2017 18:54
|
|
|
I don't know. That sounds like sound to me.
|
|
#
?
Jun 15, 2017 20:35
|
|
|
That sounds awesome. Wrong, but awesome.
|
|
#
?
Jun 15, 2017 23:07
|
|
|
Valtis posted:I resurrected my NES emulator project and started implementing the audio unit. The first attempt at playing sound was less than successful Do you know the cause yet? It sounds like the sound is getting reversed in little chunks at a time but maybe that's just my brain trying to make sense of it.
|
|
#
?
Jun 16, 2017 00:51
|
|
|
Corla Plankun posted:Do you know the cause yet? It sounds like the sound is getting reversed in little chunks at a time but maybe that's just my brain trying to make sense of it. It was a combination of various bugs and issues; incorrect sampling rate, small audio buffer size and some bugs in the pulse channel implementation. The CPU timing code was broken as well, which I hadn't noticed before. The system was running ~5% slower than it was supposed to, which caused further issues with audio timing.
|
|
#
?
Jun 16, 2017 11:03
|
|
|
I really hope you put up a few other samples of broken songs. There's something neat about the distortion.
|
|
#
?
Jun 16, 2017 11:44
|
|
|
SystemLogoff posted:I really hope you put up a few other samples of broken songs. Unfortunately I already fixed the bugs and I tend to ignore all the commit best practices when I work with my personal projects, so the broken version no longer exists.
|
|
#
?
Jun 16, 2017 12:42
|
|
|
I think this is more of a GPU featurette
|
|
#
?
Sep 27, 2017 23:35
|
|
|
Found this right after submitted to MAGFest. https://twitter.com/GhostCrabGames/status/917470473226682368
|
|
#
?
Oct 9, 2017 20:26
|
|
|
Messing with a timing value for the falling boulder trap.. I'm not touching you!
|
|
#
?
Oct 11, 2017 22:25
|
|
|
Ahhh is that a tiny bird knight???
|
|
#
?
Oct 11, 2017 22:35
|
|
|
Wheeee, Philips rang the closing bell today and none of the floor terminals could get the time when their stock opened and ended up showing a blank space on the screen instead. Not terribly exciting but  Literally the only stock out of 3,000 they could find that was broken was the one ringing the bell   I literally debug my code using broadcast TV 
MrMoo fucked around with this message at 22:26 on Nov 2, 2017 |
|
#
?
Nov 2, 2017 22:23
|
|
|
Was there something special about philips that caused it?
|
|
#
?
Nov 2, 2017 23:14
|
|
|
Just finished a round of debugging with support in California (Interactive Data Corp) and they found a handful of East coast servers that are sending the wrong data but they have to escalate up through their development to find more. Hopefully they will tell me what's wrong but usually they are so embarrassed they keep quiet and I have to find out internally via high level NYSE staff. It's been randomly on and off for the last month so  I just love that this is like a 20 year old product and I'm the first person to keep on hitting different issues like this. Also awesome having to write all your own tools because they don't ship any. The support staff were lucky they just wrote some new code this week and needed that today to confirm the issue. MrMoo fucked around with this message at 23:31 on Nov 2, 2017 |
|
#
?
Nov 2, 2017 23:26
|
|
|
Apparently when there is data loss inside their data distribution network the resync message only contain a subset of the fields in each record and someone decided that opening timestamp was not important and shouldn't be there 
|
|
#
?
Nov 3, 2017 21:56
|
|
|
Wow this is really cool. Your code is literally on national tv.
|
|
#
?
Nov 4, 2017 19:41
|
|
|
I'm enjoying the challenge, but it is very much   10 monitors connected to a single server. Cannot be separate instances or windows because Chrome will reduce the clock priority of itself in the background.
|
|
#
?
Nov 4, 2017 22:44
|
|
|
So it’s running 10 screens in a single Chrome instance on one window? If it’s not too complicated to explain, how does it split off 10 of those from one window to different monitors? I’m guessing it’s a program to specify each bounds of a screen on the window.
|
|
#
?
Nov 5, 2017 01:01
|
|
|
In Linux it's just basic XRandR configuration, in Windows you would just setup the 10 monitors arrange in the 2×5 configuration like the screenshot above. The webpage is just designed to split things up to match the monitor boundaries. Bezel correction is all done in the video hardware. HTML code:
|
|
#
?
Nov 5, 2017 02:37
|
|
|
MrMoo posted:In Linux it's just basic XRandR configuration, in Windows you would just setup the 10 monitors arrange in the 2×5 configuration like the screenshot above. So what's the benefit of using HTML/JS to do this? Once you get into stuff like dealing with multiple monitors, I would think it would be easier to just write a native app...
|
|
#
?
Nov 5, 2017 03:44
|
|
|
they bought that 64GB RAM box, gotta prove they paid good money for it!
|
|
#
?
Nov 5, 2017 04:16
|
|
|
I think they're only 8GB boxes, I have the usage under 200MB now though. There are two per post, originally one per display. There is not a lot of room for so much hardware on the floor, they're packed full of all the traders gear. Behind each monitor is a half-rack of servers. Every monitor is also connected to two different systems so that they can switch over at any time and play video and whatever else, typically for events after hours. An old pic but you can see some junk behind. It's more about avoiding vendor lock-in. The Reuters implementation was originally scoped to support MSIE 6, Firefox, and Chrome. The news tickers went Chrome only and was close to being a native app although the timeline would be horrendous. To note I replaced the entire Reuters system of however many servers (originally 200+), managed by a team, from scratch with no requirements or specifications in 3 months, using only 2 servers. ICE is a huge company, 3 months is their typical time for IT just to roll out a piece of software, add on additional time for hardware and networking. It's a smart business choice, if a little over progressive. The previous version is 6 years old already. The Thomson Reuters desktop app today is literally a Chrome Embedded Framework×Trident container with support for Java & .net. Bloomberg is a lot behind in comparison. The plan however is to integrate all this stuff into their corporate network and graphics department so they can start showing it all over the company. They have far too much screen scraping and rebroadcasting today. MrMoo fucked around with this message at 04:50 on Nov 5, 2017 |
|
#
?
Nov 5, 2017 04:32
|
|
|
That's actually pretty neat. Is there a thread on here for cool project stories?
|
|
#
?
Nov 5, 2017 18:41
|
|
|
MrMoo posted:Reuters system of however many servers (originally 200+)
|
|
#
?
Nov 5, 2017 19:00
|
|
|
flatluigi posted:That's actually pretty neat. Is there a thread on here for cool project stories? There's the project.log sub-forum, I think there was a thread but most people keep quiet because of their NDAs.
|
|
#
?
Nov 5, 2017 21:11
|
|
|
Woo! I got perspective projection working! Time to celebrate by rotating my model around the Y axis. oh dear
|
|
#
?
Nov 29, 2017 02:20
|
|
|
Luigi Thirty posted:Woo! I got perspective projection working! Time to celebrate by rotating my model around the Y axis. Looks like you accidentally created a spin-1/2 cube.  That's fine though, fermions happen all the time, and to everyone!
|
|
#
?
Nov 29, 2017 05:55
|
|
|
Luigi Thirty posted:Woo! I got perspective projection working! Time to celebrate by rotating my model around the Y axis. Until it glitches at the end, it's pretty much what I always imagined Anne saw when Valentine Michael Smith disappeared stuff.
|
|
#
?
Dec 1, 2017 04:53
|
|
|
I'm making an "emulator" for a made up retro computer and I'm trying to get tinybasic working with varied results
|
|
#
?
Dec 2, 2017 17:17
|
|
|
If the cube slides off the right side of the screen, the polygons are clipped properly. If it slides off the left side, well... https://twitter.com/LuigiThirty/status/938266196151062528
|
|
#
?
Dec 6, 2017 07:02
|
|
|
netcat posted:I'm making an "emulator" for a made up retro computer and I'm trying to get tinybasic working with varied results Is this clickteam fusion 2.5? I love that shader
|
|
#
?
Dec 6, 2017 11:30
|
|
goku i won't do what u tell me
|
|
| # ? Apr 28, 2024 20:18 |
|
|
KRILLIN IN THE NAME posted:Is this clickteam fusion 2.5? I love that shader Nah I'm making this in C++. I found the shader here, I think it's the "CRT-Lottes" one
|
|
#
?
Dec 6, 2017 23:12
|
|