Děkujeme všem, kteří byli součástí OSCloud v roce 2025 — uživatelům, podporovatelům, přispěvatelům i celé komunitě kolem.
Do roku 2026 přejeme hlavně zdraví, klid, stabilní systémy, funkční zálohy a komunitu, která drží při sobě.
Honestly as much as I like what cloudflare offers, I just saw a page which had a thunnel error again and am glad that for my services, I actually choose to no longer relly on CF thunnels. Yes the reverse proxi goes through Tailscale right now, which is similar, and you do always have to depend on some sort of provider in the one or other way, but once I have headscale at some point, this setup will be quite solid as far as I can see.
So I'd like to start archieving websites with tutorials or other content I would like to keep, you never know when something might go down and not back up, or people decide to move on with things. So the question, which applications, either for running on a server in docker, or on Windows accessible for screenreaders, would be suitable for this purpose? Edit: archivebox seems quite good. The interface is a bit cluttered, but I can always access the data manually if I really need to. I guess that works for now, need to think about what exactly could be improved to maybe create an issue.
If you're looking for an interesting read, my top enjoyments of 2025 were: Dreambound, by Dan Frey. Of Monsters and Mainframes, by Barbara Truelove. When the Moon Hits Your Eye, by John Scalzi. Last Lesson, by James Goodhand. God's Junk Drawer, by Peter Clines.
I'll be emailing out monthly updates throughout 2026 with all my newly-read material if you want to be on the list or just read more about what I enjoyed this year, the post is up at buttondown.com/SeansShelf/arch…
Well hello, dear readers about reading, and welcome to the start of a new year. As you might have noticed, the last few weeks of the year truly fell by the...
Not just hashtags. I feel it's fairly obvious that comments on a post also need to be sorted by 'relevance'. At least if a post gets more than a handful of comments.
Tak dneska je ten slavný den, kdy si dobrovolně hodíme do přírody pár tun těžkých kovů. Jsme na kopci, kde fouká, lidi v centrech na tom budou mnohem "lépe" (haha). Zákaz samozřejmě nebude mít žádný vliv, znáte to rčení o vlku a koze. Alespoň budou data pro článek.
byl jsem překvapený, že u nás celkem na pohodu. Po půlnoci to na chvíli vystřelilo na 30 mikrogramů na m3 a pak to spadlo do normálních hodnot. Nikdo v naší blízkosti nic neodpaloval a asi pomohly i dobré rozptylové podmínky.
@martenson u nás bylo letos citelně klidněji. Sice to je těžko vymahatelný zákaz, ale myslím, že to funguje už jenom jako jasný signál, že ohňostroje a petardy jsou společensky mírně řečeno nevítané. Je to jako s 1,5m u předjíždění cyklisty. To je taky těžko vymahatelné a policajti kvůli tomu s metrem po cestě neběhají, ale stejně pozoruji, že mě auta objíždí ve větší vzdálenosti. Hodně řidičů si kvůli tomu asi uvědomilo, že objíždění cyklisty ve vzdálenosti půl metru není OK.
Let it not be forgotten that 2026 started on a #Thursday. About 20 minutes after midnight, I tripped on a tree route and scraped my arms on nearby rocks trying to catch my fall. I was unsuccessful, by the way.
Happy new year to all the beings of the eastern time zone! Hoping it's a good one. Someone over there should tell me something about what the future is like.
Happy new year everyone! To begin the celebrations, I have just open sourced my nvgt include in my games to allow one to easily manage many effects in their audio games. Check it out here! github.com/IMS-Productions/nvg…
Hi everyone, happy newyears! I've decided to do something I said I'd do a long time ago and release some of my older games open source, such as Bloodshed, Iminent Death, Sammy Center and a couple others. I don't have binary builds of these right now so you'll need NVGT to play and mod them, but regardless I hope you all have some fun with these and that they bring back some good memories! github.com/samtupy/legacy_game…
Honestly I love these old sounds/logos and selfvoicing menus more than looking at the code. And I've not even known audio games when these have been created, so it's not nostalgic, but still super fun to just look at these old stuff people started out with.
@TomGrant91 I really do appreciate your awesome work, and I’m thankful that you are deciding to open source these masterpieces for us to take a look at. You are amazing sam. Thank you. Freedom is one great thing. I'll see about playing these on mac and linux and sharing my results and of course crediting you because that's important, and I like you, this that and the other. Thanks Tomlet for reposting this one.
We recorded Friday's episode of #StroongeCast yesterday, since Kirsten's out today, so if you want it early, please consider becoming a youtube member and you can listen there. Same time same place on friday as usual though. youtube.com/channel/UCqtMRZeGS…
Hi, how are you doing? Thanks for visiting my channel, firstly.
A bit about me then.
I'm Andre Louis (AKA Onj), a visually impaired musician from London.
What will you find here?
Videos on this channel fall into a few categories.
If you run an instance, don’t forget to register a DMCA agent with the US Copyright Office. If any of your users posts infringing materials, you can be held liable even if you had nothing to do with it; but, if you have a registered agent (and follow the other steps in §512) you are immune.
In response to multiple threats by the United States to turn Canada into the 51st state, Canadian Prime Minister Mark Carney surprised Canadians and Washingtonians by declaring Washington Canada’s 11th Province: theneedling.com/2025/03/22/can…
After weeks of trying to strike up friendly conversations with people about their neighbors, lonely and distraught local ICE Agent Martin Briggs says no one is talking to him here because of the Seattle Freeze: theneedling.com/2025/02/04/ice…
“Bigg’s orcas usually aren’t as picky about what they eat, but this little rudder-snapper won’t eat a thing his mom puts in front of him unless she cuts it into little yacht shapes." theneedling.com/2025/11/27/pic…
Local marine biologists have discovered one picky orca calf won’t eat their rich unless their mom cuts them into little yacht shapes. “Bigg’s orcas usually aren’t as picky about what they eat…
Today a local conservative and ICE agent who cried bloody murder when asked to wear a mask during the height of the COVID-19 pandemic clarified he’s actually totally fine with wearing masks as long as it’s to ruin lives, not save them:theneedling.com/2025/03/31/loc…
Today a local conservative and ICE agent who cried bloody murder when asked to wear a mask during the height of the COVID-19 pandemic clarified he’s actually totally fine with wearing masks as long…
“You know how transplants are: Even when they’re cats – didn’t skip a beat,” said Seattle Humane spokesman Landon Furcz as he finished his Taco Time:theneedling.com/2025/01/13/res…
Within 24 hours of landing in Seattle, animal shelter workers are confirming that all 19 rescue cats flown in from California are already complaining about the Mexican food here. “You know how tran…
“We tried remote, we tried hybrid, but after a lot of thought and investment in AI, we really think a full return to work somewhere else policy for everyone is the way to go,” said CEO Satya Nadella: theneedling.com/2025/07/18/mic…
Hi! This is a test post from the NVDA remote client developed by @Bri It seems to have a few quirks, but over all it seems to work decently. The only difficulty I've had so far is that it's a bit difficult/seems to have odd trouble going between the modifiers and the common keys, making things like shift tab a bit difficult. But other than that, it's working well. Oh, and it's very weird to be able to BSI to my computer. Lol
It seems to be a bit iffy if you navigate by heading to get between the sections. Sometimes if you navigate quickly the buttons don't pop up/into focus until you swipe around a bit. Can make a recording if it'd help you replicate/see the behavior
I recorded from a minute before midnight until about 12 after with my Zoom H5 Studio and it's included X/Y microphones on a tripod out in the back yard, while I was sitting on the back porch with a hand-cranked mechanical siren, which can be heard way off to the right in this recording. Given all the guns, the siren seems appropriate somehow.
This was recorded at an extremely conservative level, and still got turned down by 20dB when normalized because of the transients from particularly large rockets. I squished the dynamics to make it more listenable.
Mostly, a lot of gunfire from semi-automatics, with some fireworks is heard.
#HappyNewYear U.S. Central, Canada- Winnipeg, Regina, more of Nunavut; Mexico- Mexico City, Acapulco; Guatemala, Honduras, El Salvador, Nicaragua, and lots other places...
Happy mew year! Ever wanted to NVDA Remote into a computer on your iOS device? But whoops, forgot your keyboard? Or have a keyboard? Either way, NVDARC has got you covered! NVDA Remote Controller is a new app for iOS and Mac (as soon as it finishes reviewing) that allows you to control computers over the NVDA Remote protocol. It also offers an onscreen keyboard, as well as page swiping gestures and magic tap, for sending keys with the touch screen. Useful for when you have things you want to do on the computer and are away from it. Try out an early version here! testflight.apple.com/join/edg8…
Oh and last spam I promise... Was thinking about having the headings be expand/collapse items. I never, ever want to see the numpad for example so if it was a heading and a button, I could just shut it down and keep it out the way and off-screen.
I didn't realise that it wasn't using US/UK keyboard layout, so my apostrophe was sending some random crap. Made typing hard. Haven't tested yours with a keyboard yet but I know I won't have to worry about that.
Trying it now and boy am I super happy about something actually good, rather than this thing that uses the totally wrong keyboard layout and frankly was quite an abandoned project. Marvelously done for a start! Will there be any way to detect whether a list of utterances has been sent to the synth (output queueing) and therefore let each utterance run in turn (certainly important when reading with Say All)? I know VoiceOver is super horrible at this unless done right also. I've used both AVSpeech and VoiceOver and do see some utterance flow improvements here, but other than that, this is probably the best remote app I could have used other than an app delivering audio like Jump. I love that you can hear tones too! I hope in the future there can be a way to play from NVWave,, but that sounds very difficult to implement given how somewhat limited this protocol is (at least from my observations). Again, absolutely a fantastic start here and look forward to much more of this app to come!
@ChrisDuffley AVSpeech from my testing works perfectly with utterances, VoiceOver though is a sad state of affairs, and I haven't found a way to actually cue speech utterances yet so I recommend sticking with AVSpeech for best results.
I'm using the ESpeak synth on the iPhone with AVSpeech, and using Say All and also stuff such as a list name and then the contents of the list item, the first part of it gets interrupted so I have to look at the speech log for it. The same would have gone for Say All. I'll try with other voices and give you my results on it; I find it fascinating that you have no issues with utterance queues there though.
@ChrisDuffley for playing sounds I just recommend putting them into the app executable as the server does send the sound file names to play on specific events
@techsinger Say all with NVDARemote only tends to work well if you use the same synthesizer with exactly the same parameters on both sides. I wouldn't expect that to be particularly smooth here for that reason.
Seems you were right. My apostrophe is the # instead, even though UK and US have it in the same place. No apostrophe on my keyboard anywhere that I can find.
OK that's interesting. In the UK layout, not sure you know, but we have an upside down L enter key. As a result, we have one extra key to the right of apostrophe which in this latest build says '3', but when I press that with control, becomes backslash. Apostrophe is still labeled as '#' which is actually what that key should be. to make it easier to understand, our home row goes like this: jkl;'#
To the left of z where you have a long shift, that's our backslash which is actually correct. This means we have two backslashes at-present.
I just made an interesting discovery. If I press control plus my actual backslash to the left of z, it says Control plus Tick, but when I press it by itself, it says backslash.
Aah OK. Well don't stress about this, sleep soon, it's not urgent. I'm sure I can work around it by using a combination of on-screen plus BT if needed.
@tspivey Interesting. I don't have an iOS 18 device to try. Did you try changing your speech settings? Do any voices show up in the voices list? Does preview work?
@tspivey I'm on IOS 18 and haven't seen this, speaking works fine for me. Although it sometimes seems to randomly skip things here and there but from my tests it's nothing horrible.
When choosing a TTS voice, it isn't apparent which language each one belongs to. So for example, I have multiple entries for the Eloquence voices and have to manually select and sample each one.
I only seem to be able to adjust the speech rate in 10 percent increments, but would prefer it to be more granular.
When connected, I often have to press the "type text" button twice to make the keyboard actually appear. I get the feeling this most often happens when I've been swiping around to other buttons which maybe makes the "type text" button scroll out of view.
When typing text, pressing Return makes the keyboard disappear. I was trying to control an app that requires me to frequently press enter (often twice in a row), and this made it kind of impossible to use it efficiently.
When typing text, it might be helpful to have a sort of quick access toolbar above the keyboard with common keys like Tab, arrows, etc. maybe even let users customise what appears in that area.
Even with heading navigation, the keys for certain common actions are quite far apart (such as Alt+Tab). Maybe some preset combinations would help.
And lastly for now: this might just be me, but I keep looking for the Tab key under navigation rather than common.
@jscholes Re having commonly used functions above the keyboard, I have no idea how easy this would be to do but one of the coolest implementations of this I saw was in ish, where they had a single control for the arrow keys which VO read as an adjustable. Then swiping up and down would press the right/left arrow by default, but if you double-tapped it it would switch to pressing up/down instead. Having one for arrows and perhaps a second one for tab/shift-tab would be very intuitive I think.
@pitermach @jscholes I was thinking just now about having pickers that would change what the three finger swipes and magic tap gestures did on the fly.
@jscholes Aah, and keep VO on for that to work. I had to turn VO off if I wanted to use Control Windows and Alt combinations to control TweeseCake for example, as VO ate the keys.
This is really slick and I love how you did the 3-finger swipes for arrows. By other posts I've seen in the thread I'm guessing using AVSpeech is supposed to queue up speech properly? Because for me any new speech just interrupts whatever is being spoken, so if I tab into the desktop it just says the selected item but not desktop list for example, this is both with eloquence and samantha. I'm on the 26.3 beta which may be a factor.
@pitermach Yes, seriously, I'm sorry to be blunt, but this makes the older one look like a disaster by the contrast. This is a really usable program, it's what the older app could have been with a ton of development. Slick is definitely the word. I mean it's the usual thing one expects from @Bri it just gets out of the way and lets you work, like TC and BNT, but the fact that it's expected doesn't mean it's not great to see.
@Onj Up and down yeah, not left and right though. IDK if those are even bound by default. I myself mapped those to headings to not have to use the damn roter all the time but pretty sure they were unmapped by default.
@BorrisInABox I'm sorry to contradict, but this hasn't been my experience, though to be fair, I haven't tried with anything earlier than 2021. The NVDA remote companion works with say-all with a different synth of remote and local, and so does TeleNVDA. It has never been a problem for me. The problem has been a cuing issue with the remote companion, but that problem had nothing to do with the synth, you could change to anyone you wanted and would still have that issue before it was finally fixed a few days ago.
As for those nitpicks, first, if the program is talking and the screen locks, is it supposed to stop and disconnect? Is there a way to keep the screen on while in the program, or alternatively not to disconnect and go silent as the thing talks?
@techsinger iOS and background is weird. Apple has things they really don't want your app to do, and one of them is run in the background unless you're music or video etc. However, I can make the screen stay awake while you're connected, yes.
I will, if I may, put each issue in a new message so as not to get things confused. If you prefer another way of giving feedback, let me know and I'll use it.
There is a cuing issue. When I hit windows+m, for example, I hear, on the local NVDA "Desktop list" then "List top:" then "Recycle Bin 1 of 63". On the controller, all I hear is "Recycle Bin 1 of 63". The first two bits seem not to be spoken. This is the same issue I saw with the remote companion and which @Aryan fixed on that application. I am using the program's own speech, not VO announcement. I think someone has already reported skipping and this may be the same thing.
As you know, VO wants the capslock and alt+ctrl keys. Is it, or would it, be possible to remap, for example, right alt on an external keyboard to be an NVDA key? I'm doing that with Autohotkey on the server side, but that's not exactly great.
@Onj This would be nice, I'm using a bunch of units on the same subnet. It would be nice to have everything filled in so I can just change a few characters.
Yeah. I'm using all my machines on a locally run server via tailscale and therefore I don't ever remember the address, so I have to copy it from clipboard each time I want to set it up again. Very first-world problems of course, but remembering or duplication would be great, or a history of previously entered addresses, not sure.
BTW when I did copy the host address it added in a colon and allowed me to hit connect, but didn't throw an error. Maybe you can have it only allow standard things in that edit field, excluding colons? Sorry, don't want to make thousands of hours of extra work, just throwing out ideas.
@techsinger I can speed up or slow down the local or remote synthesizer by not very much, and it instantly breaks. Keep them exactly the same, and I've been able to read very long documents with no problem at all.
@techsinger I'm using the original add-on. Or, well, the one that is now built into NVDA, which is not actually the original add-on, but it did the same thing.
Yes, I get it :) I'm using TeleNvda, my machine with the core one isn't up. I'm using the IBMTTS driver. I set the server at a speed of 50 and the client at the speed of 30. Is that enough of a difference? I started say all on the server. It just kept reading. I have a book with fairly long chapters and it got through one of them without issues except that the client's speech was, in fact, slower. I stopped say all and nothing happened, I just went on as usual. What happens on your end? That is, how do you know it's broken?
@BorrisInABox You know, your speech log is making my life allot easier, troubleshooting this, or even trying, would be far more of a pain without it. Anyhow, I increased the speed of the voice from 60 to 80. Nothing changes. That same speech, desktop, list top, recycle bin 1 of 63, is sent to the log. What I hear is just "recycle bin 1 of 63". The speed did not change things on my end.
@techsinger I'm also using IBMTTS. 45% to 50% is enough to break it for me. If the remote is slower than local, it will eventually just pause for a while as the other machine catches up. If it's faster, then it eventually will jump to who knows where. It's just been like that for as long as I can remember, at least for me, so I've always kept them exactly the same when possible.
@BorrisInABox How long does it take for that to happen? Is 10K characters in multiple paragraphs enough? As I said, I went through a chapter with the speed 20% different and it didn't fail. I think the whole chapter was around 15000 characters. It was definitely significantly more than 10K. There were no skips and no jumps.
@BorrisInABox I'm totally stumped. The only thing I can suggest, if you want to bother with this, is to see whether you can manage to put in a telenvda server and client and see if you have the same issue. Oh, and one more thing while I think of it. I don't use a server. all my testing has been over direct connections between devices, whether on the LAN or over a VPN. I figured I had better mention it since I know many people use servers.
any clue if this plays nice with braille on either platform? If not, can it be made to? I've been controlling Windows devices from my mac using RIM and not having an braille output is that workflow's main weakness
Hi @BorrisInABox I'm listening to the Outside stream and I just heard a clock on there fairly close to the microphone. Did that just happen to occur or was that editted in to the stream? It's super cool to hear the fireworks on there too! Happy new year!
Who doesn't like #reverb on their tracks? I sure do. But what has frustrated me for the past 2 years of using #BitwigStudio is that the stock reverb device has felt not as smooth and silky, or easy to manipulate as I'd like. The Lexicon 224 sound is something to strive for, and those clever engineers found a way to make an incredible sound with not amazing spec digital audio (12-bit with a max frequency of 10,400Hz).
I think the secret sauce is the selective bands and delay times where you can have the the reverb swell into brightness or dwell into darkness with some simple tone controls.
I have found that if you stack two Bitwig Reverb devices in an FX-3 Multiband Device on the Low and Mid bands that you can get a really sweet sound. Here is my example with some Airwindows to round it out.
Debacle
in reply to Daniel Gultsch • • •Ærion
in reply to Daniel Gultsch • • •When will this be rolled out? My F-Droid release still reads 2014-2025 ☹️
Happy 🎊2️⃣0️⃣2️⃣6️⃣🎆