Views: 6,899,311 | Homepage | Main | Rules/FAQ | Memberlist | Active users | Last posts | Calendar | Stats | Online users | Search | 04-23-24 08:38 PM |
Guest: |
Main - Posts by squeegily |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 1/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Title says it all
When I try to Run anything after melonDS was launched in sight of the romlist, I get melonDS 0.8.3
http://melonds.kuribo64.net/ OpenGL: renderer: AMD TURKS (DRM 2.50.0 / 5.5.5-200.fc31.x86_64, LLVM 9.0.0) OpenGL: version: 3.3 (Core Profile) Mesa 19.2.8 Audio output frequency: 48000 Hz ARM9 BIOS loaded ARM7 BIOS loaded Segmentation fault (core dumped) This was tested with both the built-in "DS Menu" firmware, and ACWW#0223. Both crashed if the romlist was present. However, if melonDS was launched without access to the romlist, it complained on startup, but at least begun to run both without issue. (ft. filesize and crc32 where applicable to clear up possible ambiguities:)
I have not yet tested to see to what degree saving is broken (on the games I care about) without the romlist, but I'd assume this is something to be fixed, rather than "worked around" thus. If you need any more data, please ask and I'll do my best to oblige. |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 2/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Well. It seems my problem is: I'd put all the "system files" in the "system folder" (that users can't change) where the melonDS binary stood.
little did I realize that it requires to make changes to firmware.bin for basic operation (although it politely backs up the original copy first), and will die if it cannot. I had fooled myself into thinking the presence of the romlist was the problem—when, in fact, that was an artifact of my moving between the (write-protected) folder containing the romlist, and the user's folder without it… |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 3/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Posted by Arisotura "Ultimately", this should actually become irrelevant once an explicit "data folder" for storing the ini and [at least changes to] the firmware is chosen—as other robust programs (such as Minecraft and Firefox) do with their "internal configuration" data in fact, "not knowing anything else", I'd prefer to be warned / notified / see visible problems when, for whatever reason DS system settings (such as system nickname, and MAC) are unable to be saved. |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 4/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
I am trying to get WFC (Wiimmfi, via DS Wi-Fi config DNS override to spoofing server 164.132.44.106 ) working with melonDS 0.8.3 on Fedora 31.
I am getting connection errors after a suspiciously long attempt to connect: animal crossing softlocks; mario kart is kind enough to yield error 20100. I have, at the recommendation of various places online, temporarily changed to a wired connection (USB tethering via my smartphone). Here are some preliminary packet dissections, all from single-run attempts at requesting Mario Kart friend code issuance over the wired connection:
cap_net_admin,cap_net_raw+eip
Note that WFC Setup "Test Connection" claims to succeed under:
(Both WFC Setup "Test Connection" and Mario Kart yield error 52000 when using direct mode on a wireless interface.) Under no circumstances does Mario Kart claim to succeed. Do you have any further recommendations, or requests for more tests? |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 5/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Posted by Blockboy What is unique about that copy? Is it just a known-good version, or is it a build with particular patches/configurations that increase its power as far as melonDS is concerned? |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 6/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Things to try (as of v0.8.3):
that said, yea, a 1.xGHz GPU for an emulator with specialized audio requirements…I'm honestly surprised it's working as well as you described |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 7/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Posted by Sorer You might want to observe 262#1420 It seems that there is a build of pcap floating around that works around whatever's insufficient about Wi-Fi adapters. highly sketch but it allegedly works (for Windows, not Linux, interestingly) Posted by Arisotura I wonder if it'd be reasonable/logical (when using Direct Mode) to override the virtual DS's MAC Address to be equal to that of the host's card? [ETA: that said, I'd love to see indirect mode be regarded as "the norm" / "the goal" / "the default" / "the most-supported": for DS game traffic to be emitted and recieved just exactly like other game/application traffic sounds a lot more robust+sustainable than emitting and receiving highly unusual packets in a way that requires administrative networking privileges and increases inter-OS divergence] |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 8/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
“DMZ” is just “the default IP that incoming connections go to”
if you and your friend have the same (“our”) public IP, then only one of you is in the DMZ. You can't both have all the ports forwarded to you! Try turning off the DMZ, double-checking that UPnP is on, and, while you're at it—I think that game no longer needs patches, anyway, so—try running an unpatched ROM and just using a DNS spoofing server (e.g.: RC24 for Wiimmfi; PCN for AltWFC). I assume the context here is two laptops (or a laptop and a desktop), both running Windows 10? |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 9/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
If the melonDS instances are on the same LAN, that's the "virtual equivalent" of two DSs being physically-near enough to send the DS-to-DS packets.
The crucial aspect is, I think, keeping the save files separate. (Currently, it looks like the only way to do that is to copy the ROM into two different folders, since there's no way to declare a savefile location yet.) So, copy your game into two folders (to create separate saves), open both "copies" of the game up in melonDS, and…start trading? I don't think there's "anything to it" per se. |
squeegily |
|
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 11/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Posted by Generic aka RSDuckdoes it look like it's making it into the release immediately succeeding 0.8.3? (I would assume so, based on “All these features are now integrated in the master branch of the melonDS git repository.”, but it's still ambiguous—there are tons of possible workflows) |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 12/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Posted by Generic aka RSDuck^ This post is short but deserves to be signal-boosted a ton (If there's a given anti-piracy measure still outstanding against melonAP, it should definitely be brought up, though) |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 13/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Posted by RizaiaThat is actually terrible news, if you're surely interpreting the situation correctly Would you mind running these commands in the Windows CMD and reporting back their output (in order)?:
dig , and prepend “@ ” to the DNS server where applicable
e.g.: dig nas.nintendowifi.net @164.132.44.106
**do include any extra "garbage" information the commands output—copy the whole thing (don't forget to wrap the output in [code] tags!)
I have known some routers to block+intercept+override requests to “unauthorized” DNS servers; you can usually disable this, but it'd be necessary to at least peruse the manual to confirm. Would you feel comfortable sharing the manufacturer and model # of your router with us? |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 14/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Yes: those DNS choices will both work on unpatched ROMs. (Just make sure you and your friend choose the same server and same ROM!)
Are you two using the same router/internet? Are you on the same computer? Did you verify with your routers that UPnP is enabled and that the DMZ is disabled? |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 15/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Posted by Generic aka RSDuck My question wasn't about the CI aspect of the workflow; it was about the staging: specifically, whether something's presence in master implies its presence in the next release |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 16/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
multiplayer isn't really "ready" yet for anything which requires rigorous timing on the network packets
even Pokemon trading is quite unreliable, from what I understand |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 17/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Here, I've attached full proper Wireshark logs of four different DNS servers, this is trying to get issued a WFC ID and Friend Code on MKDS [D47555BE] (I switched from ACWW to MKDS because it's much more agile about interfacing with the Wi-Fi settings, and doesn't softlock, ever):
I mean, look at this; I have no idea what to make of it: |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 18/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Sadly, all this even occurs on the latest [2020-03-30] Git build. |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 19/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Looking at packets 43 and 45 on that screenshot…why might the game be sending CEX twice?
It would seem reasonable, I suppose, for the server to balk at that. |
squeegily |
| ||
Half-eaten melon Positive Schlub Level: 10 Posts: 20/23 EXP: 4258 Next: 156 Since: 03-25-20 Last post: 1458 days ago Last view: 1444 days ago |
Run glxinfo in your terminal, and post the output here.
Also include lspci -nn | grep VGA
(don't forget to wrap the output in [code] tags!!) |
Main - Posts by squeegily |
Page rendered in 0.041 seconds. (2048KB of memory used) MySQL - queries: 22, rows: 109/109, time: 0.012 seconds. Acmlmboard 2.064 (2018-07-20) © 2005-2008 Acmlm, Xkeeper, blackhole89 et al. |