| --- Log opened Mon Apr 09 00:00:24 2018 |
| 00:19 | | Kindamoody is now known as Kindamoody[zZz] |
| 00:36 | | Jessikat` [Jessikat@Nightstar-emhftl.dab.02.net] has joined #code |
| 00:39 | | Jessikat [Jessikat@Nightstar-8oi8od.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 01:01 | <&McMartin> | Man, old habits die hard |
| 01:01 | <&McMartin> | I'm sitting here fretting about maybe wasting an extra dozen bytes in my implementations of things |
| 01:01 | <&McMartin> | Meanwhile, the absolute minimum size this ROM can be is 64KB and there is ~15KB of empty space there right now |
| 01:01 | <&McMartin> | Total current actual program size: 1,728 bytes |
| 01:02 | <&McMartin> | (Everything else is bulk-converted data or enforced alignment restrictions) |
| 01:46 | <@Reiv> | This is a matter of changing chipsets and having gotten used to vastly tighter constraints, yes? |
| 01:51 | <&McMartin> | It's less the CPU than what the CPU is hooked gto |
| 01:51 | <&McMartin> | *to |
| 01:52 | <&McMartin> | CPUs, really, I guess |
| 01:52 | <&McMartin> | But yeah, this is a system where the intended size of a cartridge is 4MB, and emulator/flashcart formats only acknowledge steps of 16KB... |
| 01:53 | <&McMartin> | ... and when divvying up chunks of address space to the CPUs, it does so in 32KB chunks, which it turns out Matters for some of the code I have |
| 01:53 | <&McMartin> | So of the 64KB that I have right now, I'm using about 48KB of it, but I have to use 64 because the 32KB chunking thing means that I need Two Chunks |
| 01:54 | <&McMartin> | And of that ~48KB that I'm using, about 46KB of it is just sound/gfx data dumped from format converters. |
| 01:54 | <&McMartin> | Meanwhile, yeah, when it comes to, say, the C64 |
| 01:55 | <&McMartin> | Disk space is allocated in 254-byte chunks, and since the program itself lives in RAM, every byte counts, always. |
| 01:55 | <&McMartin> | In the modern world, the equivalent would be "you want your inner loop and the data it works on to live entirely within the L1 cache" |
| 01:55 | <&McMartin> | But once that happens, s/L1/L2/ and things don't get appreciably worse until you run out of L2 cache |
| 01:56 | <&McMartin> | ... and then until you run out of physical RAM |
| 01:56 | <&McMartin> | then you're thrashing and you generally just give up~ |
| 01:57 | <&ToxicFrog> | Also, modern L2 caches tend to be in the 256-512KB range :P |
| 01:57 | <&McMartin> | Speaking of "new CPU" though, man it is *so nice* to be able to fit an entire pointer in one register, and I still need to remind myself that sometimes I don't need to bother with putting addresses *in* registers because I can jam all the information I need directly into instructions without touching registers at all. |
| 01:57 | <&McMartin> | TF: That "and the data you're working on" part makes that pretty constrained, I'd expect. |
| 01:57 | <&ToxicFrog> | Mmm, true |
| 01:58 | <&ToxicFrog> | Just saying that your entire ROM could fit in L2 cache several times over with room to space |
| 01:58 | <&ToxicFrog> | *spare |
| 01:58 | <&McMartin> | The largest reasonable problem I can imagine fitting in there is JPEG decoding or encoding and even there I'm not 100% on how big an image that even gives you |
| 01:58 | <&McMartin> | That's not enough to hold a 1080p frame decompressed. |
| 01:58 | <&McMartin> | Oh sure. |
| 01:58 | <&McMartin> | This chip doesn't have any cache. |
| 01:59 | <&McMartin> | It's got 136KB of RAM split up amongst three chips. |
| 01:59 | <&McMartin> | So also part of the trick is making sure your buses are not fite |
| 02:01 | <&McMartin> | On the plus side, there is an ~~~= actual bus arbiter =~~~ on the Genesis board which means that I can actually do significant processing outside of VBLANK |
| 02:01 | <&McMartin> | But I'm not yet sure whether or not that will introduce visibly awful tearing |
| 02:02 | <&McMartin> | Fortunately, we already solved this problem on the Game Boy, so~ |
| 02:04 | <&McMartin> | And, you know, even without that, just not going nuts means I can get very nice code sizes out |
| 02:05 | <&McMartin> | I need to turn a 128x128 array of bytes into an interleaved pair of 8K graphics tiles, and that comes out to 58 bytes |
| 02:06 | <&McMartin> | Which is in part because of *16-bit indexed addressing modes holy shit* though anyone who's primarily been writing x86 code, especially IA32 code, would not be so immensely relieved by this |
| 02:42 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 02:42 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 04:26 | <&McMartin> | ...... |
| 04:26 | <&McMartin> | Well |
| 04:26 | <&McMartin> | I've finished the weekend's Genesis hacking |
| 04:26 | <&McMartin> | Total change in actual code size (not ROM or data size): negative 14 bytes. |
| 04:28 | <&McMartin> | That doesn't really sound right. |
| 04:29 | <&McMartin> | OK, there we go. That makes more sense |
| 04:30 | <&McMartin> | From $05a0 to $06b4, so that's 276 net bytes of code. |
| 04:31 | <&McMartin> | Rather more gross, since I was also stripping out a bunch of stopgap code that no longer needs to be there. |
| 04:44 | | Degi [Degi@Nightstar-vrmak5.dyn.telefonica.de] has quit [Connection closed] |
| 05:10 | | Derakon is now known as Derakon[AFK] |
| 05:16 | | Vorntastic [Vorn@Nightstar-q5v2ai.sub-97-33-192.myvzw.com] has joined #code |
| 05:17 | | Vornlicious [Vorn@Nightstar-t9ha10.sub-174-211-6.myvzw.com] has quit [Ping timeout: 121 seconds] |
| 05:18 | | Vornlicious [Vorn@Nightstar-t9ha10.sub-174-211-6.myvzw.com] has joined #code |
| 05:20 | | Vorntastic [Vorn@Nightstar-q5v2ai.sub-97-33-192.myvzw.com] has quit [Ping timeout: 121 seconds] |
| 05:56 | | Kindamoody[zZz] is now known as Kindamoody|out |
| 06:24 | | Jessikat` is now known as Jessikat |
| 06:41 | | Kindamoody|out [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Ping timeout: 121 seconds] |
| 06:45 | | Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code |
| 06:45 | | mode/#code [+o Kindamoody|autojoin] by ChanServ |
| 07:06 | | Jessikat` [Jessikat@Nightstar-q8bgkr.dab.02.net] has joined #code |
| 07:09 | | Jessikat [Jessikat@Nightstar-emhftl.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 07:14 | | celticminstrel [celticminst@Nightstar-m1ak9t.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 07:34 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 07:37 | | Vornlicious [Vorn@Nightstar-t9ha10.sub-174-211-6.myvzw.com] has quit [Ping timeout: 121 seconds] |
| 10:41 | | Jessikat` [Jessikat@Nightstar-q8bgkr.dab.02.net] has quit [The TLS connection was non-properly terminated.] |
| 10:41 | | Jessikat [Jessikat@Nightstar-q8bgkr.dab.02.net] has joined #code |
| 10:46 | | Jessikat [Jessikat@Nightstar-q8bgkr.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 11:06 | | Degi [Degi@Nightstar-vrmak5.dyn.telefonica.de] has joined #code |
| 12:08 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 13:57 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 14:52 | | gnolam_ [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 14:55 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has quit [Ping timeout: 121 seconds] |
| 15:16 | | gnolam_ is now known as gnolamn |
| 15:16 | | gnolamn is now known as gnolam |
| 15:16 | | mode/#code [+o gnolam] by ChanServ |
| 15:25 | | Degi [Degi@Nightstar-vrmak5.dyn.telefonica.de] has quit [Connection closed] |
| 15:43 | <&McMartin> | These are probably two memes that didn't need to be mixed, but the Internet is extremely insistent. https://pbs.twimg.com/media/DaPaDezVwAEM2M0.jpg |
| 15:49 | <~Vornicus> | ....silliness |
| 15:53 | <@TheWatcher> | ... wat |
| 15:54 | <&McMartin> | This is apparently that "American Chopper" meme that people keep meming about |
| 15:54 | <&McMartin> | Combined with the source and context for the "A monad is just a monoid in the category of endofunctors, what's the big deal" gag |
| 15:55 | < [> | are they both agreeing though? the meme is supposed to be an argument |
| 15:55 | <@TheWatcher> | Yeeah, just going to stay under my rock, I think~ |
| 15:55 | <&McMartin> | Part of what made me actualy smirk at this was the idea that this sequence of text *was* an argument |
| 15:56 | <&McMartin> | And yeah, to quote someone else on this network: "if nothing else it seems like American Chopper came onto the scene already operating on the 8th level of irony" |
| 15:58 | <&McMartin> | ... which was actually apparently part of a conversation mocking audiophiles, and includes the excellent line "In your choice of FLAC, ALAC, AIFF, or WAV, depending on whether or not you believe lossless compression is possible" |
| 17:29 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has joined #code |
| 17:29 | | Vornlicious [Vorn@Nightstar-sbl2mb.sub-174-211-9.myvzw.com] has joined #code |
| 17:31 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 17:32 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 17:32 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 17:33 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 18:11 | <@gnolam> | https://twitter.com/ncweaver/status/983005767723728896 |
| 18:16 | | * gnolam facepalms. |
| 18:19 | <@gnolam> | For like two weeks one customer has been complaining about an installation not working, and pls to fix. |
| 18:21 | <@gnolam> | From what we could see immediately, it wasn't our fault - whatever was going wrong was happening on the driver side. |
| 18:22 | <@gnolam> | But cue the usual back-and-forth blame game anyway. |
| 18:34 | <@gnolam> | After finally convincing them that yes, our stuff was in fact working fine, and it *must* be on the driver/hardware side, and then getting them on the phone directly with the hardware manufacturer, they got back to us with a "problem solved" message today. |
| 18:35 | <@gnolam> | "The guys left one of the hatches on the machine open." |
| 18:45 | <@gnolam> | So this *whole* time, not *one* person actually went and looked at it in person. |
| 18:51 | < Vornlicious> | When I worked in Utah there was an emergency situation with our software where it wasn't connecting to the on-site server |
| 18:52 | < Vornlicious> | "it's plugged in, right?" "Right, I see it plugged in and on" |
| 18:52 | < Vornlicious> | 4 hour drive later: "it's not plugged in, they were looking at the secretary's computer" |
| 18:56 | <&jeroud> | This is why I do not support other people's hardware. |
| 18:57 | <&jeroud> | Although a significant amount of my work stress is caused by the people who are supposed to support the hardware being utterly incompetent and occasionally maliciously dishonest. |
| 19:43 | | Vornlicious [Vorn@Nightstar-sbl2mb.sub-174-211-9.myvzw.com] has quit [[NS] Quit: Bye] |
| 19:43 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 20:30 | | Jessikat [Jessikat@Nightstar-bt5k4h.81.in-addr.arpa] has joined #code |
| 20:42 | | gnolam_ [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 20:43 | <&McMartin> | 11:41 < Roscoe> The sacrifices were stopped. The sun is silent. The dead now rise. |
| 20:43 | <&McMartin> | 11:41 < Roscoe> Defend the floating gardens. |
| 20:43 | <&McMartin> | Er |
| 20:43 | <&McMartin> | https://arstechnica.com/gadgets/2018/04/relive-the-1990s-on-windows-10-with-the-newly-open-sourced-file-manager/ |
| 20:43 | <&McMartin> | There we go |
| 20:44 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has quit [Ping timeout: 121 seconds] |
| 20:44 | <~Vornicus> | madness! |
| 20:45 | | gnolam_ is now known as gnolam |
| 20:45 | | mode/#code [+o gnolam] by ChanServ |
| 20:47 | | Degi [Degi@Nightstar-vrmak5.dyn.telefonica.de] has joined #code |
| 21:00 | | Kindamoody|autojoin is now known as Kindamoody |
| 21:21 | | macdjord [macdjord@Nightstar-ahbhn1.cable.rogers.com] has joined #code |
| 21:21 | | mode/#code [+o macdjord] by ChanServ |
| 21:24 | | mac [macdjord@Nightstar-ahbhn1.cable.rogers.com] has quit [Ping timeout: 121 seconds] |
| 21:57 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 23:05 | | Degi_ [Degi@Nightstar-vrmak5.dyn.telefonica.de] has joined #code |
| 23:06 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has quit [Ping timeout: 121 seconds] |
| 23:06 | | Degi_ [Degi@Nightstar-vrmak5.dyn.telefonica.de] has quit [[NS] Quit: Leaving] |
| 23:12 | | Kindamoody is now known as Kindamoody[zZz] |
| --- Log closed Tue Apr 10 00:00:26 2018 |