| --- Log opened Sun Feb 25 00:00:20 2018 |
| 00:03 | <&McMartin> | woot |
| 00:04 | <&McMartin> | You have actually recreated a technique very similar to what yacc, bison and other first-gen parser generators do; maintaining a stack of tokens and then evaluating what to do based on what's coming next |
| 00:05 | <&[R]> | Took me a while to figgure it out, there's still some bugs though |
| 00:05 | <&McMartin> | The other major technique, often more popular for hand-written parsers but formally somewhat more constrained unless you do some trickery, is called "recursive descent" |
| 00:05 | <&McMartin> | Where you have one function for each kind of thing you can parse out, and they call each other as needed based on what's coming up next. |
| 00:06 | <&McMartin> | So the trick there is that you can make precedence fall straight out of it |
| 00:06 | <&McMartin> | expr = term + expr |
| 00:06 | <&McMartin> | term = term * factor |
| 00:06 | <&McMartin> | factor = dieroll | str | number | - factor | ( expr ) |
| 00:07 | <&McMartin> | In the stack based system you either shift or reduce based on what you have and what you see. |
| 00:08 | | Degi [Degi@Nightstar-orn9ji.dyn.telefonica.de] has joined #code |
| 00:08 | <&McMartin> | er |
| 00:08 | <&McMartin> | term = factor * term up there |
| 00:08 | <&McMartin> | One of the things recursive descent is real bad at is handling cases where the first item in the formula is something that can make a loop |
| 00:09 | <&McMartin> | (This can be fixed by means of a mechanical translation, which is good, becuase which one is the recursive term *also* ends up determining the associativity of the operator, if it matters.) |
| 00:10 | | * [R] barely follows |
| 00:11 | <&McMartin> | Yeah, I, uh, kind of just went through half of a third-year university course in two screens there >_> |
| 00:12 | <&McMartin> | Let me also then link some code where I used that alternate technique, albeit in Python and with a medium-strong C accent. https://github.com/michaelcmartin/Ophis/blob/master/src/Ophis/Frontend.py |
| 00:13 | <&[R]> | I'm having trouble with the (...)*num case |
| 00:14 | <&[R]> | It's coming out flat, so the paren operation is in the same group as the following multiplications, which is going to make evaluation more complicated |
| 00:15 | <&McMartin> | OK, so that's part of an issue right there |
| 00:15 | <&McMartin> | Paren has higher precedence than mult. |
| 00:17 | <&McMartin> | Or rather, it should have higher precedence |
| 00:23 | <&McMartin> | . |
| 00:23 | <&McMartin> | "-funsafe-math-optimizations allows math optimizations that assume valid arguments and can violate ANSI and IEEE standards (caution, not actually fun and safe)" |
| 00:30 | <&ToxicFrog> | I have seen some really neat bugs related to -funsafe-math-optimizations. |
| 00:30 | <&ToxicFrog> | I think my favourite was a 3d cityscape renderer where compiling with that flag, on certain compiler version/runtime architecture combinations, resulted in ~1% of the city melting. |
| 00:56 | <&[R]> | Okay, got the parens issue fixed! |
| 01:02 | <&[R]> | I wasn't consistient about how I was handling the current stack state. |
| 01:43 | | Kindamoody is now known as Kindamoody[zZz] |
| 01:57 | | Derakon[AFK] is now known as Derakon |
| 01:57 | | VirusJTG [VirusJTG@Nightstar-42s.jso.104.208.IP] has quit [Connection closed] |
| 02:08 | | VirusJTG [VirusJTG@Nightstar-42s.jso.104.208.IP] has joined #code |
| 02:08 | | mode/#code [+ao VirusJTG VirusJTG] by ChanServ |
| 02:09 | <&McMartin> | Oh my goodness |
| 02:10 | <&McMartin> | ToxicFrog: Incidentally, not only does ImageMagick's convert utility output XPM, it seems that my stock copy of emacs will *render the graphic for you* if you load it up, with C-c C-c flipping between "look at it" and "text buffer" |
| 02:15 | <&ToxicFrog> | That is pretty cool. |
| 02:18 | <&McMartin> | That will be really nice if I decide to be hardcore and manually clean up scaling artifacts after colordepth drops. |
| 02:25 | | * McMartin dumps out 16 KB of highly redundant tile data. |
| 02:29 | | Degi [Degi@Nightstar-orn9ji.dyn.telefonica.de] has quit [Connection closed] |
| 02:37 | | Jessikat` [Jessikat@Nightstar-p8o959.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 02:43 | <&[R]> | Which one is XPM again? The X11 image format that's actually an C header, or the image format DOS games used to use but nothing uses anymore? |
| 02:43 | <&McMartin> | The grandchild of the former. |
| 02:44 | <&[R]> | Ah |
| 02:44 | <&[R]> | What am I thinking about with the latter? |
| 02:46 | <&[R]> | PCX I think |
| 02:46 | <&McMartin> | Could be. It was certainly that thing. |
| 02:48 | <&McMartin> | Not a terrible format, for what it needed to be, but it ended up getting expanded further than it could comfortably go, and as a result ended up getting subsumed by BMP. |
| 03:05 | <&ToxicFrog> | [R]: XPMv1 is C header, XPMv2 is plain text file containing the same data |
| 03:05 | <&ToxicFrog> | (there's also the older XBM, which, as the name implies, does black and white only) |
| 04:29 | | Pinkhair [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 04:31 | | Pink` [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 04:31 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 04:33 | | Pinkhair [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 04:55 | | Vornlicious [Vorn@Nightstar-5t04kl.sub-174-211-5.myvzw.com] has joined #code |
| 04:56 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 04:57 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 05:15 | | Derakon is now known as Derakon[AFK] |
| 05:19 | <&McMartin> | ImageMagick actually produces XPMv3, which is C header again but in a different format. |
| 05:20 | | * McMartin meanwhile has cleaned up this image as much as it will clean. |
| 05:20 | <&McMartin> | I really need a new Dignified Umbrella In Snow logo that survives drops in resolution or color depth without making the handle vanish. |
| 05:24 | <&McMartin> | https://i2.wp.com/bumbershootsoft.files.wordpress.com/2018/02/smd_orin_logo.png |
| 05:31 | < Vornlicious> | Wow, yeah |
| 05:34 | <&McMartin> | And that result is after a great deal of manual cleanup to prevent the text from becoming an illegible mess, but, well, yeah, don't shrink then dither your text, kids |
| 05:34 | <&McMartin> | Also man that logo is like 15 years old now |
| 05:35 | <&McMartin> | Also now that I have roughly 16KB of memory that needs to be blitted unchanged to VRAM... |
| 05:35 | <&McMartin> | ... I finally have an excuse to deploy BLAST PROCESSING |
| 05:35 | <&McMartin> | Because I totally can't actually post an I-did-a-thing for the Genesis unless it includes BLAST PROCESSING. |
| 05:36 | <&McMartin> | (Caveat: BLAST PROCESSING happens during system initialization when it's resetting the graphics chip into a known state. This does not count.) |
| 05:47 | < Vornlicious> | Blast |
| 05:48 | <&McMartin> | Also yes this is what you get when you give a dumb marketing name to an extremely ubiquitous hardware feature. |
| 05:48 | <&McMartin> | (BLAST PROCESSING was programmable DMA) |
| 05:50 | < Vornlicious> | MC Programmable DMA is my rapper name |
| 05:50 | <&McMartin> | And this case you're even backed up by MC68000 |
| 05:52 | < Vornlicious> | Nice |
| 06:18 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 06:18 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 06:32 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 07:15 | | * McMartin gets his DMA working. |
| 07:25 | | Jessikat [Jessikat@Nightstar-o84ov0.dab.02.net] has joined #code |
| 08:00 | | * McMartin greets Jessikat with BLAST PROCESSING and half-competently downgraded images. https://i2.wp.com/bumbershootsoft.files.wordpress.com/2018/02/smd_orin_logo.png |
| 08:00 | < Jessikat> | Invisible umbrella stem, I like it |
| 08:02 | <&McMartin> | Yeah, this source image from 15 years ago is seriously vulnerable to that =( |
| 08:03 | < Jessikat> | :< |
| 08:04 | <&McMartin> | The original source image is like 512 tall and even if you scale down with full 24-bit color the stem keeps vanishing |
| 08:04 | <&McMartin> | But! I got data import into the Mega Drive working, plus DMA transfer from ROM to VRAM |
| 08:04 | <&McMartin> | I really only have a couple more things to do in isolation before I'm actually happy with my command of the hardware. |
| 08:05 | <&McMartin> | (Input scan and the other three sound modes, basically.) |
| 08:06 | | celticminstrel is now known as celmin|sleep |
| 09:00 | | Kindamoody[zZz] is now known as Kindamoody |
| 09:04 | | Jessikat` [Jessikat@Nightstar-r7d8vk.dab.02.net] has joined #code |
| 09:07 | | Jessikat [Jessikat@Nightstar-o84ov0.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 09:09 | | Kindamoody is now known as Kindamoody|afk |
| 09:22 | | Jessikat` is now known as Jessikat |
| 09:35 | | Jessikat [Jessikat@Nightstar-r7d8vk.dab.02.net] has quit [[NS] Quit: Bye] |
| 10:34 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has quit [[NS] Quit: Gone] |
| 10:55 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 10:55 | | mode/#code [+o gnolam] by ChanServ |
| 11:48 | | Jessikat [Jessikat@Nightstar-fuqq53.dab.02.net] has joined #code |
| 11:57 | | Jessikat [Jessikat@Nightstar-fuqq53.dab.02.net] has quit [[NS] Quit: Bye] |
| 12:14 | | * Vornlicious frustrates at web compatibility tables |
| 12:29 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 12:30 | | Pink` [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 12:37 | | Pink is now known as ASCII |
| 12:46 | | Degi [Degi@Nightstar-mijjeu.dyn.telefonica.de] has joined #code |
| 14:23 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has quit [[NS] Quit: This year can fuck off and die already] |
| 15:10 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 15:10 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 15:57 | | Degi is now known as Degi[afk] |
| 16:14 | | Vornlicious [Vorn@Nightstar-5t04kl.sub-174-211-5.myvzw.com] has quit [[NS] Quit: Bye] |
| 16:14 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 16:14 | | gnolam [quassel@Nightstar-f22.ckv.119.62.IP] has joined #code |
| 16:14 | | mode/#code [+o gnolam] by ChanServ |
| 16:27 | | Jessikat [Jessikat@Nightstar-ioe4tq.dab.02.net] has joined #code |
| 16:51 | | celmin|sleep is now known as celticminstrel |
| 16:59 | | Degi[afk] is now known as Degi |
| 17:14 | | Vornlicious [Vorn@Nightstar-ils7ka.sub-174-211-19.myvzw.com] has joined #code |
| 17:18 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 17:26 | | Kindamoody|afk is now known as Kindamoody |
| 18:24 | | gnolam [quassel@Nightstar-f22.ckv.119.62.IP] has quit [[NS] Quit: Blergh] |
| 18:32 | | Derakon[AFK] is now known as Derakon |
| 18:46 | | Jessikat [Jessikat@Nightstar-ioe4tq.dab.02.net] has quit [[NS] Quit: Bye] |
| 18:47 | | Jessikat [Jessikat@Nightstar-ioe4tq.dab.02.net] has joined #code |
| 18:47 | | Jessikat [Jessikat@Nightstar-ioe4tq.dab.02.net] has quit [The TLS connection was non-properly terminated.] |
| 21:40 | | Jessikat [Jessikat@Nightstar-ioe4tq.dab.02.net] has joined #code |
| 22:24 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 22:36 | | Jessikat` [Jessikat@Nightstar-unn.equ.132.82.IP] has joined #code |
| 22:39 | | Jessikat [Jessikat@Nightstar-ioe4tq.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 22:41 | | gnolam [quassel@Nightstar-hsn6u0.cust.bahnhof.se] has joined #code |
| 22:41 | | mode/#code [+o gnolam] by ChanServ |
| 23:13 | | ASCII [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Connection closed] |
| 23:14 | | ASCII [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 23:18 | | Degi [Degi@Nightstar-mijjeu.dyn.telefonica.de] has quit [Connection closed] |
| 23:21 | | Jessikat` [Jessikat@Nightstar-unn.equ.132.82.IP] has quit [Ping timeout: 121 seconds] |
| 23:23 | | gnolam [quassel@Nightstar-hsn6u0.cust.bahnhof.se] has quit [[NS] Quit: Z?] |
| 23:26 | | McMartin [mcmartin@Nightstar-rpcdbf.sntcca.sbcglobal.net] has quit [[NS] Quit: Time for the weekly kernel upgrade] |
| 23:47 | | Pinkhair [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 23:48 | | McMartin [mcmartin@Nightstar-rpcdbf.sntcca.sbcglobal.net] has joined #code |
| 23:48 | | mode/#code [+ao McMartin McMartin] by ChanServ |
| 23:50 | | ASCII [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| --- Log closed Mon Feb 26 00:00:22 2018 |