Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Posted on 19-02-16, 04:52 in BSNES Beta question
Full mod

Post: #121 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Previous versions of higan expected the ROMs to be named "dsp1.program.rom" and "dsp1.data.rom", but between the last release of higan, and the first beta of bsnes, the expected names were changed to "upd7725.program.rom" and "upd7725.data.rom". Just rename the files (or copy them, if you want the same game to work for both higan 106 and bsnes), and it should work fine.

The ending of the words is ALMSIVI.
Posted on 19-02-17, 07:56 in Minor complaint with higan on linux..
Full mod

Post: #122 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
How very Aztec.

The ending of the words is ALMSIVI.
Posted on 19-02-17, 22:39 in Mozilla, *sigh*
Full mod

Post: #123 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Posted by wareya
They didn't actually backtrack on anything, they're still going to kill ublock.

Do you have a source for that?

From the actual announcement quoted in the article, it sounds like they haven't completely rolled over and promised to give ad-blockers everything they want, but it sounds like they *are* addressing the most common objections to the original proposal:

- "webRequest" was going to be removed, now it's just read-only (for extensions other than ad-blockers)
- the replacement "declarativeNetRequest" will be taught to add and remove rules at runtime, so users can edit their ad-blocker lists
- "declarativeNetRequest" originally had a maximum rule count that was too small for the current EasyLIst, never mind future expansion. Google will raise the limit by some amount, but still wants *some* limit, and suggests 90% of EasyList is unused.
- "declarativeNetRequest" originally had a small number of ways to handle a request, and very simple tests to decide how to handle a given request, while the AdBlock Plus/uBlock rule syntax is quite sophisticated. Google will add fancier tests, and fancier ways to handle requests.

Obviously there's no *details* in the announcement, so it's impossible to say whether ad-blockers in their current form will be possible with the new API, but it sounds a lot better than it did a few weeks ago.



The ending of the words is ALMSIVI.
Full mod

Post: #124 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Now, how long will it be until BBoard2 is no longer the first site Firefox recommends when I open a new tab?

The ending of the words is ALMSIVI.
Posted on 19-02-19, 04:07 in Mozilla, *sigh*
Full mod

Post: #125 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
As I understand it, AdBlock Plus and uBlock Origin both support block-lists in the same format, but uBO is smarter about storing the filter list in memory and more efficiently testing it against incoming requests. It sounds like Google want to supply their own optimised implementation of that code.

In some ways, that's a good thing, since it means carelessly written ad-blockers can't be inefficient and bog down the browser. uBlock Origin will probably get even lighter-weight and faster, once C++ is doing all the heavy lifting, and AdBlock Plus gets to match uBO's performance without having to write a whole new matching engine from scratch.

In some ways, it's a bad thing, since it means even carefully-written ad-blockers can't be faster or smarter than their competition.

Fundamentally, it's difficult to give people the power to help each other without also giving them the power to screw each other over.

The ending of the words is ALMSIVI.
Posted on 19-02-22, 00:45 in Board feature requests/suggestions (revision 1)
Full mod

Post: #126 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
While I can browse the site with Firefox, my RSS reader (https://www.newsblur.com/) can no longer fetch the feed.

It turns out I can't fetch the page with curl, either:
$ curl -v https://helmet.kafuka.org/bboard/
[...]
* Trying 151.236.14.55...
* TCP_NODELAY set
* Expire in 200 ms for 4 (transfer 0x55d99e3edbd0)
* Connected to helmet.kafuka.org (151.236.14.55) port 443 (#0)
* ALPN, offering h2
* ALPN, offering http/1.1
* successfully set certificate verify locations:
* CAfile: none
CApath: /etc/ssl/certs
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
* TLSv1.3 (IN), TLS handshake, Server hello (2):
* TLSv1.3 (OUT), TLS alert, protocol version (582):
* error:1425F102:SSL routines:ssl_choose_client_version:unsupported protocol
* Closing connection 0
curl: (35) error:1425F102:SSL routines:ssl_choose_client_version:unsupported protocol


The SSL Server Test reports that the webserver doesn't support anything newer than TLS 1.0, which apparently is a Bad Thing™ and might explain why NewsBlur is having issues.

The ending of the words is ALMSIVI.
Posted on 19-02-22, 02:17 in Board feature requests/suggestions
Full mod

Post: #127 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
That should work, yes.

The ending of the words is ALMSIVI.
Posted on 19-02-22, 06:41 in bsnes v107 released
Full mod

Post: #128 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
The release announcement and changelog is on the front page of byuu's site, but the permanent download link is on the official bsnes page.

Posted by "Change log"
Today I am posting the first release of the new bsnes emulator.

bsnes is designed to be a revival of the classic bsnes design, focusing specifically on performance and ease of use for SNES emulation.

In addition to all of the features of higan, bsnes supports the following features:

300% faster (than higan) scanline-based, multi-threaded graphics renderer
option to disable sprite limits in games
option to enable hires mode 7 graphics
option to enable more accurate pixel-based graphics renderer
option to overclock SuperFX games by up to 800%
periodic auto-saving of game save RAM
save state manager with state screenshots
several new save state hotkeys such as increment/decrement slot#
option to auto-save states when unloading a game or closing the emulator
option to auto-load aforementioned states when loading games
save state undo and redo support (with associated hotkeys)
speed override modes (50%, 75%, 100%, 150%, 200%)
recent games list
frame advance mode
screenshot hotkey
path selection for games, patches, saves, cheats, states, and screenshots
dynamic video, audio, input driver changes
direct loading and playing of games without the use of the higan library
ZIP archive and multiple file extension support for games
firmware folder for unappended coprocessor firmware (see documentation for more)
compatibility with sd2snes and Snes9X MSU1 game file naming
compatibility with higan gamepaks (game folders)
soft-patching support for both BPS and IPS patches
menubar that does not pause emulation when entered
video pixel shaders (requires OpenGL 3.2)
built-in game database with over 1,200 games to ensure perfect memory mapping
(Linux, BSD only:) audio dynamic rate control to eliminate stuttering
and much more!

The one feature I regret not being able to support in this release is Windows dynamic rate control. I put in my best attempt, but XAudio2's API is simply not fine-grained enough, and the WASAPI driver is not mature enough. I hope that DRC support can be added to the Windows port in the near future, and I would like to offer a large cash bounty to anyone who can help me make this happen.


The ending of the words is ALMSIVI.
Posted on 19-02-22, 06:57 in bsnes v107 released
Full mod

Post: #129 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Because this isn't an official higan release, I've pushed it to a separate branch of the git repo. I don't expect any further releases in the v107 family, but if there are any, I'll put them also on the branch.

For technical reasons the higan documentation site *does* provide docs for "higan v107" but that's for the hypothetical version of higan one might get by building higan from the bsnes source tarball. Most of it's not very relevant for bsnes, because the UI is quite different, but where the changelog says "firmware folder for unappended coprocessor firmware (see documentation for more)" you probably want to look at the "games with co-processor firmware" section of the higan v107 docs.

The ending of the words is ALMSIVI.
Full mod

Post: #130 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
The problem is that the Internet currently runs on advertising money. Firefox isn't really big enough to destroy the entire industry with one punch, and do you *really* want to deal a non-fatal blow to an industry that specialises in shaping people's opinions? It's like the old saying: "Do not meddle in the affairs of bards, for your name sounds funny and fits the meter of many popular songs".

Even if Firefox *could* end the advertising industry, that might not be a wise plan. Like I said, the Internet currently runs on advertising money so cutting off that revenue would kill off a lot of the internet. Sure, the parts you or I care about wouldn't be the worst hit, but we enjoy high-speed, high-quality Internet access because it was bankrolled by large companies who wanted high-speed, high-quality Internet access for themselves and their customers. If the Internet suddenly became "business hostile", maybe we'd all have to go back to 28.8k modems or Fidonet or something.

Mozilla's plan is more gentle because it gives people time to adjust - if advertising revenue decreases by 5%, that's not an existential crisis, but it encourages people to find alternative revenue streams. And later, when Mozilla does something else that decreases advertising revenue by another 5%, people can invest more into those alternative revenue streams, and eventually all the generally useful companies are weaned from advertising money, and all the parasite companies are dead.

What makes this approach particularly interesting is that for a long time Mozilla has been trying to wean itself off advertising money, build up alternative revenue streams, and generally demonstrate that it's possible to build a sustainable, ethical business online. And every time they try something, Slashdot and Reddit erupt with indignant fury at the idea of supporting Mozilla more directly than by letting Google track them across the Internet.

The ending of the words is ALMSIVI.
Posted on 19-02-23, 01:31 in BSNES Beta question
Full mod

Post: #131 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Posted by jarsonic
Ok, here's a somewhat weird one, Screwtape:

With bsnes v107 on Windows 10, I dropped the "dsp1.program.rom" and "dsp1.data.rom" files into /firmware and renamed them to "upd7725.program.rom" and "upd7725.data.rom", respectively. Super Mario Kart now works great with no issues.

However, when I load Pilotwings (USA).sfc, I get the "Error: missing required data udp7725.program.rom" message. Is this a bug, or am I missing something?

Much like higan is a program that runs on an x86_64 processor, "DSP1" is a program that runs on a "µPD7725" processor, and so are "DSP1B", "DSP2", "DSP3" and "DSP4". When we're talking about Super Mario Kart specifically, it expects the µPD7725 to be running the DSP1 software, but you might want to replace it with one of the other DSP programs just to see what happens, or even with a homebrew program that doesn't infringe Nintendo's copyright. If the program running on the µPD7725 isn't exactly the DSP1 software, it doesn't make sense to put "dsp1" in the filename, so higan uses "upd7725" in the filename for all DSP-like chips.

Of course, you can't put all the different DSP files into the firmware directory all named "upd7725.program.rom", or they'd overwrite each other. You can call them whatever you like in the firmware directory, bsnes should identify them by hash.

The specific problem that you're hitting is that there's actually three variants of the DSP1 chip. The DSP1 was the original, the DSP1A was the same software in a physically smaller chip, and the DSP1B included various minor bug-fixes in the same physical format as the DSP1A. While most games that use the DSP1 (including Super Mario Kart) are happy with either the DSP1/1A software or the DSP1B, Pilotwings depends on the bugs present in the DSP1 version.

It sounds like your "dsp1.{program,data}.rom" files are the DSP1B versions, and bsnes specifically doesn't want to use them with Pilotwings because of bugginess. You can manually copy the files to the Pilotwings game folder to see the bugginess for yourself (the plane in the attract mode demo crashes, I don't know if there are other differences), or you can track down a copy of the DSP1/1A firmware and drop that into the firmware folder before importing.

The ending of the words is ALMSIVI.
Full mod

Post: #132 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Yeah, I guess I should have said "wean itself off surveillance capitalism money". It's hard to invent wholly new revenue streams, we know advertising works on the Internet, and society has been dealing with non-surveillance advertising for centuries, so hopefully people are more likely to understand it.

The ending of the words is ALMSIVI.
Posted on 19-02-24, 11:50 in A bug in nall/primitives/integer.hpp?
Full mod

Post: #133 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Somebody just posted a comment on higan v106r80, pointing out that one of the methods of nall's Integer struct takes two parameters and only ever uses one of them, which smells like a bug.

Can anyone confirm?

The ending of the words is ALMSIVI.
Posted on 19-02-27, 10:26 in bsnes v107 released
Full mod

Post: #134 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
> I wonder whether it’s going to support HiDPI after 5 years of the 4K era.

You're in luck, byuu spent the past couple of days wrestling with Windows' HiDPI APIs to get things working. It won't be in v107, but it should be ready for v108 whenever that arrives.

The ending of the words is ALMSIVI.
Posted on 19-02-27, 13:22 in Mozilla, *sigh*
Full mod

Post: #135 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Today I wandered into about:config to look for a setting, and discovered it's been redesigned and rebranded, including a cute cartoon dinosaur like the one that appears on the "DNS error" page. At first I wondered why they'd put that much effort into a page that end-users generally won't see anyway, but apparently it was part of the "de-XUL all the things" push that's going on.

The ending of the words is ALMSIVI.
Posted on 19-02-27, 13:57 in Mozilla, *sigh*
Full mod

Post: #136 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
Double-clicking an entry does nothing, you have to click the button on the right.

One kind-of-useful feature that's now gone: previously, all the modified settings were sorted at the top of the list so you could find them easily. Now they're always in alphabetical order, like in very old versions of Firefox and Seamonkey.

The ending of the words is ALMSIVI.
Posted on 19-02-28, 03:30 in Mozilla, *sigh*
Full mod

Post: #137 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
> It should be trivial to patch Firefox to take extensions in dll/so format that interact with the internal API.

Anti-virus tools do this kind of thing all the time, and it's something like the #1 cause of Firefox crashes and security holes.

The ending of the words is ALMSIVI.
Full mod

Post: #138 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
In your "Megaman X2.sfc" folder, you only need "cx4.data.rom" (which should have the SHA256 hash ae8d4d1961b93421ff00b3caa1d0f0ce7783e749772a3369c36b3dbf0d37ef18), and "program.rom" which should have the SHA256 hash f3246755f608a1e1dc9c848b61da3b824c7853b29b3be40df6fc7f2793a887ed. Once those are in the same folder together, you should be able to load the game with higan.

You don't need to combine the two files into one to play the game, you only need to combine the files so that icarus can do all the setup and renaming for you. Since you've done that work yourself, you're fine, you don't need a combined file for icarus.

If you do need to import other games that use co-processor firmware, you'll need to do the work yourself, or you'll need to combine files. The instructions are under the "Games with co-processor firmware" section I linked in my previous message. If there's a part of those instructions you don't understand, let me know which part, so I can make it clearer.

The ending of the words is ALMSIVI.
Posted on 19-02-28, 23:55 in Mozilla, *sigh*
Full mod

Post: #139 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
> It should be trivial to patch Firefox to take extensions in dll/so format that interact with the internal API. That would solve the whole issue of 'sandboxing' and other crap, enabling you to run precisely whatever native code you wish at full performance

> Why can't they put them in a separate process like everything else?

"let's get rid of all the sandboxing for performance and flexibility, and then sandbox everything in separate processes for predictability and security".

> If you can't do it through a normal extension the user installs manually then you shouldn't do it at all.

But how can we justify charging people a $50 annual subscription fee if we don't visibly indulge our hubris?

The ending of the words is ALMSIVI.
Posted on 19-03-01, 02:41 in Mozilla, *sigh*
Full mod

Post: #140 of 443
Since: 10-30-18

Last post: 863 days
Last view: 59 days
> a poorly written extension wouldn't do it by accident.

Poorly written programs crash themselves by accident all the time. Heck, *well* written programs sometimes crash themselves, just because they're not 100% perfect 100% of the time, and that's when we're talking the same team or the same person writing the whole thing. If we're considering two separate teams working on two separate codebases that have to interact in the same memory space, you can bet there'll be fireworks more than occasionally.

The ending of the words is ALMSIVI.
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
    Main » Screwtape » List of posts
    Yes, it's an ad.