| + reformer (~reformer@softboy.mntmn.com) | 17:14 | |
| + reformer (~reformer@softboy.mntmn.com) | 17:20 | |
| * ChanServ changed mode (+v, reformer) | 17:20 | |
| + trixster2 (trixster40@host81-153-87-71.range81-153.btcentralplus.com) | 17:20 | |
| * ChanServ changed mode (+o, mntmn) | 17:20 | |
| mntmn | logs are back, sorry for the service interruption | 17:21 |
|---|---|---|
| trixster2 | good afternoon all | 17:31 |
| mntmn | hello | 17:32 |
| trixster2 | feeling a bit poorly today, sojust chilling withthe amiga andsome tunes and the football whilst the wife and dog are out | 17:32 |
| mntmn | i am watching shadowrun SNES longplay in the background and wondering why there was no amiga version | 17:33 |
| trixster2 | i was watching a longplay of the amazing BBC Micro game Exile this morning (it had a superb port to the Amiga too, by Audiogenic) | 17:34 |
| mntmn | oh neat, never really checked it out | 17:35 |
| trixster2 | such a clever game, with superb physics. it was about the pinnacle for BBC Micro games programming, and came out late for the system in 1989 | 17:35 |
| mntmn | ah there was also a c64 version | 17:36 |
| trixster2 | and the amiga conversion was amazing too, a really tastfully done overhaul for graphics and sound but lost nothing of the original. 8 to 16bit done properly | 17:36 |
| trixster2 | yes, the c64 version was excellent too | 17:36 |
| trixster2 | the BBClongplay is on youtube by a guy called Lord Triax. it#s fully annotated so its very interesting to watch | 17:37 |
| trixster2 | sorry about the typing, im on the A4000 and the keyboard is a bit rough :D | 17:37 |
| mntmn | haha | 17:39 |
| mntmn | no problem | 17:39 |
| trixster | the 4000 has been neglected recently due to the hot weather and me showing the A1200 some attention thanks to it's new TF1260 :) | 17:40 |
| mntmn | oh neat how is the tf1260? | 17:41 |
| trixster | superb now that the beta fw has been released. ram access speed read/write is amazing now, faster than the Blizzard 1260 and Apollo 1260 clock for clock by some margin | 17:42 |
| mntmn | noice | 17:42 |
| mntmn | i don't quite remember, was there an adapter for a3000/4000? | 17:42 |
| trixster | i dont believe so. he has made a TF4060 though, that's with a few people for alpha testing | 17:43 |
| trixster | my Edu Arana made an adapter, I lose track! | 17:43 |
| trixster | *maybe | 17:43 |
| mntmn | oh, that sounds like a product for me | 17:43 |
| mntmn | i need a replacement for my unstable warpengine | 17:44 |
| mntmn | where i could tinker with the firmware myself | 17:44 |
| trixster | i think i'll definitely be getting either a TF4060 or Matze's 68060-TK for my 3000 | 17:44 |
| trixster | it will be interesting to see the price of these two boards (without CPU), pre-assembled | 17:45 |
| + gdonner (~gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 17:46 | |
| mntmn | wow exile looks really good (watching the amiga version) | 17:50 |
| trixster | there was an AGA version as well, but they made the mistake of making the main sprite too large and the view too zoomed in. So the OCS version is the better version | 17:51 |
| mntmn | oh ok! | 17:52 |
| trixster | it's an incredibly difficult game, hence why i watch the longplay as i could never get more than about 2 hours into it! | 17:52 |
| _Bnu | You can't really tinker with the firmware for the TF cards, no source is available. | 17:55 |
| _Bnu | (Well, I guess you could maybe talk him into letting you have the source code, but I don't think you'd be allowed to make anything public.) | 17:56 |
| mntmn | _Bnu: oh ok | 17:56 |
| mntmn | maybe should rather check out matzes stuff then | 17:56 |
| trixster2 | matze is a good guy :) | 17:58 |
| mntmn | that is so | 17:59 |
| _Bnu | What if he's actually so evil that it overflowed into good... | 18:00 |
| gdonner | Just noticed this (https://eab.abime.net/showpost.php?p=1505637&postcount=228)--if true, will add it to the ZZ9000 page so others are aware of it | 18:01 |
| gdonner | my list of bugs is probably way out of date :-| | 18:02 |
| Jope | that's an old bug, haven't seen it in ages | 18:02 |
| Jope | maybe it still exists? | 18:02 |
| gdonner | yeah, they didn't say what version of firmware/drivers they were using | 18:03 |
| _Bnu | I get the feeling he changed the order of the bitmaps at some point. | 18:03 |
| _Bnu | So you need at least P96 3.1.0 for it to work. | 18:03 |
| gdonner | he did say he added a workaround for some badly-behaving apps | 18:03 |
| _Bnu | But this P96 stuff is getting really weird, telling people they should rewrite their software because something broke in the newest version (3.2.0?) | 18:03 |
| _Bnu | I don't have an active license at the moment, so I have no idea what's up. | 18:04 |
| mntmn | there was at least one faulty p96 version | 18:04 |
| gdonner | hmm; no problems here with the lastest, except maybe a solid orange screen that write itself to the frontmost screen every once in a while | 18:05 |
| gdonner | *latest | 18:05 |
| gdonner | not sure where that orange screen comes from | 18:05 |
| _Bnu | Yeah I dunno, I don't keep track of it at the moment. | 18:05 |
| mntmn | thor also wrote me (and i did but didn't release yet as a .card) that we should remove some weird address masking in one function in the driver | 18:05 |
| _Bnu | What function? | 18:05 |
| mntmn | ah, thor is talking about the screen split bug there | 18:05 |
| _Bnu | I already removed the address masking for the pan function. | 18:06 |
| mntmn | that is a rare race condition i think | 18:06 |
| mntmn | yeah not pan, wait | 18:06 |
| mntmn | an ancient thing | 18:06 |
| trixster | i'm still delaying moving to P96 3.xx as some change made to rtg.library broke fullscreen PPC warp3d speed :( | 18:06 |
| trixster | niche case, but annoying! | 18:07 |
| _Bnu | Just tell them to rewrite their software...! | 18:07 |
| trixster | hah, yeah! | 18:07 |
| mntmn | oop i didn't even push that | 18:07 |
| Jope | :-) | 18:07 |
| gdonner | P96 3.1.1 seems the most stable for me; 3.1.2 is fine too except for that orange screen | 18:08 |
| _Bnu | Uhh. | 18:08 |
| _Bnu | You didn't pull my commit either. | 18:08 |
| _Bnu | So now you broke the driver flags again. | 18:08 |
| mntmn | _Bnu: https://source.mnt.re/amiga/zz9000-drivers/-/blob/wip-zz9000ax/rtg/mntgfx-gcc.c#L622 | 18:08 |
| mntmn | _Bnu: what did i break? i am in my own branch | 18:08 |
| Jope | I'm out of date again, I'm running 3.1.1 :-((( | 18:08 |
| gdonner | but then again, I am using firmware 1.9 (still beta obviously) | 18:09 |
| Jope | the fast paced world of amiga | 18:09 |
| _Bnu | b->Flags = BIF_INDISPLAYCHAIN | BIF_GRANTDIRECTACCESS | BIF_HARDWARESPRITE | BIF_FLICKERFIXER | BIF_VGASCREENSPLIT | BIF_PALETTESWITCH; | 18:09 |
| _Bnu | b->Flags = BIF_INDISPLAYCHAIN | BIF_GRANTDIRECTACCESS | BIF_HARDWARESPRITE | BIF_FLICKERFIXER | BIF_VGASCREENSPLIT | BIF_PALETTESWITCH | BIF_BLITTER; | 18:09 |
| mntmn | _Bnu: yeah thor wrote me to add that | 18:09 |
| mntmn | BIF_BLITTER | 18:09 |
| _Bnu | b->Flags |= BIF_GRANTDIRECTACCESS | BIF_HARDWARESPRITE | BIF_FLICKERFIXER | BIF_VGASCREENSPLIT | BIF_PALETTESWITCH; | 18:09 |
| _Bnu | Yeah I mean, I don't blame you for not remembering this. | 18:09 |
| _Bnu | But it's supposed to be |= | 18:09 |
| mntmn | ach so | 18:09 |
| apolkosnik[m] | The last one I got from August was making the screen go black | 18:09 |
| _Bnu | Because P96 will set flags from the tooltype and then never set them again. | 18:10 |
| mntmn | _Bnu: it has some default flags that need to be kept from... ah yeah | 18:10 |
| mntmn | i see | 18:10 |
| _Bnu | So using = there clears any flags that would be set from the tooltype. | 18:10 |
| mntmn | ok thanks | 18:10 |
| mntmn | will remember this now | 18:10 |
| mntmn | i did not have this saved in my brain | 18:10 |
| gdonner | @apolkosnik: 3.1.0 was fairly buggy; the later versions fixed a bunch of stuff | 18:10 |
| _Bnu | Anyway, let me delete that commit. | 18:11 |
| _Bnu | So you won't have any trouble if you want to merge it back to gcc-test later. | 18:11 |
| apolkosnik[m] | I meant the fw | 18:11 |
| gdonner | ah; sorry | 18:11 |
| _Bnu | There, removed it. | 18:12 |
| _Bnu | But you really shouldn't have BIF_INDISPLAYCHAIN in the flags there. | 18:12 |
| _Bnu | That means it'll always be set, so the multi monitor whatever stuff won't work. | 18:12 |
| mntmn | _Bnu: cool, thanks for bringing that up | 18:14 |
| mntmn | i am on reform in my home | 18:14 |
| mntmn | i shall fix it in the gitlab web IDE | 18:14 |
| mntmn | _Bnu: how do you like it now https://source.mnt.re/amiga/zz9000-drivers/-/commit/4db501471b987506c32ec66c2db2f580865e0ae0 | 18:16 |
| _Bnu | Looks good, but I'm still wondering if setting the blitter flag actually works? | 18:17 |
| _Bnu | It used to result in some horrible amount of rendering problems with menus and such. | 18:17 |
| gdonner | ah; one other thing I noticed in the P96 update notes (just curious): In case the board did not indicate the presence of the blitter, memory | 18:17 |
| gdonner | reorganization for screen dragging might not have worked at all, and thus | 18:17 |
| gdonner | screen dragging may have failed. This also hurt the ZZ9000 which, besides | 18:17 |
| gdonner | having a blitter, does not indicate its presence to P96. | 18:17 |
| gdonner | (but probably old news) | 18:18 |
| mntmn | _Bnu: i shall test it | 18:19 |
| mntmn | gdonner: that's precisely the thing yeah. i have not tested the effect of advertising the blitter yet though | 18:19 |
| _Bnu | I heard about it from Lukas, but I've tested it before and it didn't work properly. And I don't have my A4000 set up at the moment, 3.5"/5.25" drive bay is out because I wanted to check if I could do anything about the weird timing problems. | 18:19 |
| gdonner | ok, thanks! | 18:20 |
| gdonner | as always, thanks Lukas and _Bnu for all your work with this! | 18:21 |
| gdonner | (and anyone else I missed) | 18:21 |
| gdonner | OK; going AFK for a haircut :) | 18:23 |
| mntmn | enjoy! | 18:24 |
| _Bnu | I mean maybe that CalculateMemory thing was why it didn't work with the blitter "enabled", but I had no idea what that function was supposed to do anyway. | 18:34 |
| mntmn | yeah | 18:34 |
| _Bnu | On the PiStorm RTG, it just returns the address it was supplied with, and that seemed to work, lolb. | 18:38 |
| mntmn | yep | 18:39 |
| mntmn | this is a leftover from VA2000 | 18:39 |
| mntmn | where i had to make sure that screens where opened at certain page aligned addresses | 18:39 |
| _Bnu | But on the PiStorm I just super ignored the screen split thing. Since the Pi 3 GPU can't even do two shader blits without the frame rate dropping, I don't even want to think about what it'd take to switch bitmaps at the right screen position, haha. | 18:40 |
| mntmn | because my sdram controller was kind of primitive at first | 18:40 |
| mntmn | _Bnu: crazy | 18:40 |
| mntmn | _Bnu: do you run everything in HD always or why is that? | 18:41 |
| _Bnu | No, it just can't do it. | 18:41 |
| _Bnu | Doesn't matter how small the bitmap is. | 18:41 |
| mntmn | huh. | 18:41 |
| _Bnu | It's because it like copies the texture data to the super tiny VRAM, then when it's done it copies it back. | 18:41 |
| _Bnu | So it's super slow. | 18:41 |
| _Bnu | That's one of the reasons why it can't even run Quake 3 at 60 FPS, haha. | 18:42 |
| mntmn | on imx8mq screen split could be done with a custom DCSS driver | 18:42 |
| _Bnu | It's like, gotta switch texture to render polygons from = instant death. | 18:42 |
| mntmn | because that has a "context loader" that can change the fb address at some x/y position | 18:42 |
| _Bnu | Yeah, I just wanted to use a shader for the CLUT screens and RGBA screens. | 18:43 |
| _Bnu | The GPU is supposed to be able to do RGB565BE, but it doesn't work. | 18:43 |
| mntmn | yeah makes sense | 18:43 |
| mntmn | i'm envious of the pi's GL version and vulkan driver but not of this stuff | 18:44 |
| _Bnu | And remapping the CLUT and AGBR to RGBA in software was even slower. | 18:44 |
| _Bnu | The Pi's GL version is fake, for what it's worth. | 18:44 |
| _Bnu | And Vulkan performance is pretty terrible... | 18:44 |
| mntmn | oop | 18:44 |
| mntmn | can it run firefox webrender? | 18:44 |
| _Bnu | You can probably have better GL (3 or so) on the Reform, if someone were to plop around with it. | 18:44 |
| _Bnu | The Reform GPU performs like ten times better with my NES emulator than the Pi 3. And that's probably an understatement, haha. | 18:45 |
| _Bnu | And no. | 18:45 |
| _Bnu | Web browsing on the Pi is absolutely miserable. | 18:45 |
| mntmn | mhm https://bugzilla.mozilla.org/show_bug.cgi?id=1663285 | 18:45 |
| mntmn | _Bnu: highly interesting | 18:45 |
| _Bnu | The Raspberry Pi 4, maybe. | 18:45 |
| _Bnu | It has ~3x the pixel pushing muscle of the Pi 3. | 18:46 |
| mntmn | so called PPM | 18:46 |
| _Bnu | But when I tried web browsing on it, it was crap. | 18:46 |
| mntmn | give me imx9 and rpi5... | 18:49 |
| mntmn | both of those will exist in the future, isn't that strange! | 18:49 |
| _Bnu | iMX2011 handshake initiated... | 19:01 |
| _Bnu | 768MB/sec is the highest speed imaginable by humans. | 19:21 |
| Jope | faster than the speed of light | 19:36 |
| - trixster2 (QUIT: Ping timeout: 252 seconds) (trixster40@host81-153-87-71.range81-153.btcentralplus.com) | 19:46 | |
| + my_pc_IS_amiga (~my_pc_IS_@2601:200:100:8fcb:31a2:b612:f83:bcee) | 20:18 | |
| - my_pc_IS_amiga (QUIT: Ping timeout: 256 seconds) (~my_pc_IS_@2601:200:100:8fcb:31a2:b612:f83:bcee) | 20:33 | |
| - gdonner (QUIT: ) (~gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 20:57 | |
| + trixster4000 (trixster40@host81-153-87-71.range81-153.btcentralplus.com) | 21:34 | |
| - trixster4000 (QUIT: Ping timeout: 252 seconds) (trixster40@host81-153-87-71.range81-153.btcentralplus.com) | 22:20 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!