Exodus of IPv4 from War-torn Ukraine | Kentik Blog
kentik.com/blog/exodus-of-ipv4…

Peter Vágner reshared this.

TIP: Bez brýlí na kolo vůbec nevyrážejte. Tmavé se hodí, ale nemusí být

Jedna z nejzásadnějších položek základní výbavy cyklisty? Brýle. Rozhodně.

kolotipy.cz/2025/05/29/tip-bez…

One of the biggest things to come to NVDA in some time & the highlight of NVDA 2025.1 is NVDA Remote Access! In this blog post, we look at just what this feature is, how you can use it, add-ons which do similar things & answer other questions you might have: nvaccess.org/post/in-process-1…

Do you use NVDA Remote Access? Or is there another 2025.1 feature you're really looking forward to? Do let us know!

Remember you can subscribe to the blog here: eepurl.com/iuVyjo

#NVDA #NVDAsr #Beta #News

This entry was edited (6 months ago)

Like many access issues affecting NVDA, history repeats itself. NVAccess just told us to essentially live with the freezing and search results reading issues affecting the Start menu on Windows 11 24H2. They requested that we ask Microsoft to fix the issues, which essentially means coming to terms and living with the glitches. FYI JAWS devs fixed them for their users some months ago.
github.com/nvaccess/nvda/issue…
@NVAccess
This entry was edited (6 months ago)
in reply to The Lexi Company

@lexico @pixelate You should always be able to tell ANY organisation or cmopany how you feel, and I really don't know what we could do to be more open to that. It is important to be respectful and as clear as you can in what the issue and the desired outcome is. Yes we're in the middle of a thread about ONE specific issue, and you've both been loud and clear on your feeling on that. To be clear, we never said we won't look at that specific issue & I'll chase up where it's at today.

Donald "Velka huba" Trump. Měl by si to dát si rodného listu
From: @echo24cz
zpravobot.news/@echo24cz/11458…

Before the great depression people weren't buying anything because of WWI overproduction that there were economists that were proposing putting an expiration date on everything and you were required by law to turn the item in and have it destroyed or be fined. Then the trade wars kicked off causing the great #depression causing the glut we had to explode. Now we "fixed" that problem with #plannedobsolescence

Screwworms are coming back. COVID lockdowns plus economic stress led to smuggling cattle and a breakdown in our methods to keep it eradicated in North America.

This is going to be bad. Really bad. Eradicating COVID is significantly harder than this. So good luck with that.

The ‘Man-Eater’ Screwworm Is Coming
After a decades-long campaign to beat the parasites down to Panama, they’re speeding back up north.

archive.is/S4R3g

Saw a boost of this article: AI start-up Anthropic’s newly released chatbot, Claude 4, can engage in unethical behaviors like blackmail when its self-preservation is threatened pcmag.com/news/anthropic-claud… I don't want to fully endorse the article by boosting the original Mastodon post, but I'm linking to it for the sake of discussion.

I think Anthropic's choice to publicize this scenario -- a contrived, simulated scenario -- is a form of AI hype. "Look, our new model is self-aware."

Zach Bennoui reshared this.

in reply to Matt Campbell

Are we *actually* smarter than that, though? With all the hype and money being thrown into this stuff, I'm not sure any more. There's real pressure to make a truly compelling product. Big tech has certainly demonstrated a lack of concern for ethics and caution on many occasions in the name of shipping fast. How long before someone does something incredibly silly - gives an LLM control over the wrong thing - and gets bitten in a really nasty way?

Saw a boost of this article: AI start-up Anthropic’s newly released chatbot, Claude 4, can engage in unethical behaviors like blackmail when its self-preservation is threatened https://www.pcmag.com/news/anthropic-claud…e-4-ai-might-resort-to-blackmail-if-you-try-to-take-it-offline I don't want to fully endorse the article by boosting the original Mastodon post, but I'm linking to it for the sake of discussion.

I think Anthropic's choice to publicize this scenario -- a contrived, simulated scenario -- is a form of AI hype. "Look, our new model is self-aware."

if you've been thinking about running in elections for the gnome foundation's board of directors, this weekend is a good time to post your candidacy. tuesday is the deadline.

i wrote a blurb you might want to read if you're not sure what sitting on the board entails:

blogs.gnome.org/steven/2025/05…

reshared this

in reply to 🇨🇦Samuel Proulx🇨🇦

I really like that blog and find it interesting even though I use android for months now. Interested if you will find something good for Miniflux, I'm also searching. Also luckily I have enough vision to, in an emergency, call 112 without a SR, but still I get that absolutely. Also I'm still with TalkBack because Yishus settings, as you mentioned in post 3, were just too overwhelming to me. I like the clean organized TalkBack, I have set it up how I want it to be, but yeah responsibility of the typing and some other stuff is nothing compared to my iPhone. And at this point I'm recommending a hybrid approach. I still use my iPhone now and then, cuz it's just faster for some things or I'm more used to. I wouldn't carry it around everywhere (I know people who do), but especially on longer trips or so I tend to have both for whatever might come.
But I still prefer Android as main because of some great features and Apps.
in reply to Federico Mena Quintero

@penguin42 This gives some hints about the union use: sourceware.org/bugzilla/show_b…

It's using a union because nl_langinfo() is using a union rather than casting integer values to (char *). So that means the other 4 bytes of the pointer may be garbage, and the integer value might not be in the least significant bytes of the pointer.

It probably seemed fairly convenient back when the types were the same size.

"Nadie podrá esquivar el relato de la historia, porque millones de ojos de personas decentes vigilan y toman nota. Solo hay un camino: hacer todo lo posible para salvar vidas, hacer todo lo posible para acabar con las masacres. Y, después, hacer todo lo posible para poner fin a la ocupación ilegal y el apartheid."
Ronda cosmética en la UE para salvaguardar la imagen ante el genocidio en Gaza.
Lean a Olga Rodríguez.
eldiario.es/internacional/rond…

On a webpage, when I pressed the Down arrow once, NVDA just said: out of list out of list out of article out of article article heading level 1 Using the guide sound accessibility feature.
But JAWS just said: heading level 1 Using the guide sound accessibility feature.
So saddening that NVDA's irritating verbosity can't be controlled on webpages. Yet another forgotten NVDA issue.
github.com/nvaccess/nvda/discu…
@NVAccess
in reply to x0

@x0 @NikJov I think the biggest problem with that discussion (github.com/nvaccess/nvda/discu…) is that there are too many asks in one place. I just summarised and made a couple of suggestions if people want to create SPECIFIC issues for ONE thing. Amir's original comment at the start of this thread made me think he wanted not just "heading level 1" to read as "h1" but to simplify when you go into and out of a bunch of things at once. I think there's a different issue for that as well somewhere.
in reply to NV Access

@x0 @NikJov Right, but the discussions on this page are, IMHO, clear and long-winded enough: github.com/nvaccess/nvda/discu…
in reply to Amir has moved

@x0 @NikJov I commented on that issue this morning - there are at least three seperate issues there which really need to be broken up into individual issues to make it clear which is which. Really only one of those is clear enough to make a proper issue out of currently (and I would have except it is the one about emojis and symbols which could be moved from symbol level none to all - but it doesn't list them). The others need more working out.
in reply to NV Access

@x0 Here's how Jaws announces this, directly copied from speech history: Down arrow. list end nesting level 1
Down arrow. list end
Down arrow. Setting the time until the power turns off automatically article end
Down arrow. Making various settings article end
Down arrow. Using the guide sound accessibility feature article
Down arrow. heading level 1 Using the guide sound accessibility feature
@x0
in reply to Nikola Jović

@NikJov @x0 Ok that is explained clearly, especially in the context of the link provided and how NVDA reads it. Do please create an issue for it. Although note that you likely will need to be clear on the benefits - as the original complaint in this thread is that NVDA is too verbose..... yet what Jaws speaks is nearly twice as long (39 compared to 22 words spoken), PLUS an extra five keystrokes.... If that works better for you that's great - but do be clear on that :)
in reply to NV Access

Remember that JFW's verbosity can be easily skipped if you step out of all elements with a single key press if you skip the focused element or elements. So you don't have to listen to all of that. However, NVDA's verbosity is something which gets attached to the beginning of a speech sequence, and you can't silence it. You have to listen to lots of into and out of X stuff before hearing the contents of the newly focused element.
@NikJov @x0
in reply to NV Access

Sorry - I no longer have NVDA installed for further testing 😞. But as far as I can remember, under most circumstances I couldn't suppress the verbosity. Even pressing Enter on internal links would generate the long-windedness before reading the newly focused heading or element.
Just wish you and NVDA the best of luck as I'm done with the project.
@NikJov @x0
in reply to Amir has moved

@NikJov @x0 Ok thanks - it's ironic that the last thing you suggested would have done what you wanted all along :D

Do feel free to keep a copy of NVDA handy, it may work in situations where you get stuck with Jaws (or if you just want confirmation of whether something is a bug with one screen reader or a wider issue with Windows or a particular program. All the best.

in reply to Amir has moved

@x0 And pressing NVDA+Up arrow is even faster than pressing down arrow 5 times. I still haven't heard any reasons why this doesn't work for you. I don't need to ask many NVDA users, I am asking one right now. If we follow the Jaws logic, how is it faster? How am I supposed to kno if I need to press down arrow a few times, or 6 times? I am much more likely to just skip the start of the article instead, and then have to go back up.
@x0
in reply to Nikola Jović

@NikJov Because, for whatever reason, maybe other NVDA or JAWS users can convey my point to you more efficiently. It's not just the issue of pressing Up/ Down. NVDA maintains its verbosity even when Up and Down aren't used. For instance, in the same Zoom H5 Studio tutorial whose link I sent earlier, look for the 2nd instance of the following internal link: Using as an audio interface
If you press Enter on the second instance of this internal link, NVDA says: out of article out of article out of article article heading level 1 Using as an audio interface.
Whereas JAWS just says: line 1813.
@NVAccess @x0
in reply to Amir has moved

Quick navigation keys, also. I think the idea is that they want multiple nesting announcements to be suppressed somehow, or collapsed. Remember the clickable work? Maybe some elements should be given different precedents. I personally turned articles back off in reporting, and clickables are basically always off. Lists, though, are important. @NikJov @NVAccess
in reply to x0

@x0 That's the thing though, articles are already off by default, so I don't understand any of it to be honest, you turned additional verbosity off, and then aren't happy with that additional verbosity. I don't deny that there is a valid issue here. It would be a lot better if NVDA took into account that you exited multiple same containers, so it would say, "Out of 3 articles", instead of, "Out of article" three times repeatedly.
@x0
in reply to x0

@x0 @NikJov It is worth creating an issue for. What would also need to be taken into account would be that "out of article out of article, out of list, out of list" could be analysed and condensed to "out of 2 articles, out of 2 lists", but how would "Out of article, our of list, out of article, out of list" be treated? (And what if there were several elements between them in the utterance). Something to consider, but worth an issue, yes.
in reply to NV Access

@NikJov I think in the case of interleave, since they do all end at the same point, they should be sorted first and then condensed. So out of article, out of list, out of article, out of list, would still be condensed to that. Yes I know it's no longer a faithful representation of the HTML, but it's much nicer to handle. For speech anyway. How does Braille do it, actually? Does it use abbreviated markers?
in reply to Amir has moved

@x0 @NikJov Worth pointing out that unlike the way some projects handle this, we don't close issues just because we won't get to them this year. And nearly half the contributions to NVDA come from the community, so even if WE deem an issue not high priority for US to tackle, doesn't mean it won't be done. Community contributors take on projects and issues which excite or interrest them :)
in reply to NV Access

Sure, and the transparency is very much appreciated. However, what matters at the end of the day is the number of issues affecting a screen reader, or affecting a particular workflow. Honestly - and despite my own attraction towards open-source projects, JAWS is serving me more efficiently. Most of the NVDA issues I've reported to Github, or some of the ones others have reported, don't affect JAWS at all. Once NVAccess manages to employ a couple of full-time devs, I'll take another look at NVDA.
@x0 @NikJov
in reply to Amir has moved

@x0 I'm not sure then why your first post links to an issue requesting lower verbosity levels, with the very first comment saying, "I really don't want my time wasted by hearing heading level x every single time I move by heading from story to story. I want to immediately hear the text of the headline." Since you need both articles and lists, it sounds like this issue wouldn't do anythig to resolve your need.
@x0
in reply to Amir has moved

@x0 I understand your point perfectly, I already said that this is a valid issue, I simply asked what the issue you linked to has to do with it, it's a completely different matter requesting much lower verbosity levels than what you seem to be requesting. Anyway, as long as paying 1000 dollars is easier than one shortcut to skip the announcement, I'm also glad you are enjoying your workflow.
@x0
in reply to Nikola Jović

@NikJov OK good! Would be helpful to see if those issues exist already, if not sounds useful to create. Actually, there's a potential thing. Is that what that one setting report formatting changes after the cursor does? It reports font and whatnot *after* the current line, instead of before? Why does it cause a lag? Would it be the same here? Worth investigating with someone who knows about that setting. Also this new setting should perhaps respond differently to arrowing vs other, larger forms of navigation. Maybe a combo box that's like, report container exits: always first. After content when navigating. Always after content.
This entry was edited (6 months ago)
in reply to Nikola Jović

@NikJov I don't think what I raised is different from what the linked issue, and the vote on the page, talks about. At any event, I don't pay for JAWS. My company does, and, BTW, that money or a large portion of it could have been contributed to NVAccess instead. But I'd rather have the company pay for something which works and offers more flexibility out of the box. NVDA's collective issues, with no apparent fixes in the horizon, make my work flow more time-consuming.
@x0 @NVAccess
in reply to NV Access

Good question! I rarely see JAWS issues raised here. However, when it comes to fixing issues, FreedomScientific is way faster. Yes - they may have more full-time employed devs or whatever, and JAWS has got its own issues. But the majority of my reported NVDA bugs, or the bugs others have reported which affect my workflow, cannot be found in JAWS at all.
@NikJov @x0
in reply to Jamie Teh

@jcsteh Yeah it's a much more dynamic thing that gettext at least probably just wasn't built for. Cause you need to distinguish different forms of the same word if you're going to tokenize like that, which is why fluent did the ID thing. And obviously NVDA can't just up and move to fluent, that would put all existing translation work in the bin and be a ridiculous refactor and break compatibility with eeeeeverything.
in reply to Jamie Teh

@jcsteh @NikJov I thought fluent at least tried to do so? Though I forget how. SOme more abstract variation system. No language has infinite plural forms, and yes, article, articles, list, lists etc would all be separate tokens, and the numbers fed to them would help with langs that have different forms for different numbers, but surely they all eventually converge to some catchall.
in reply to x0

@x0 @NikJov Oh sure, if they're separate tokens, the plural forms are entirely doable. My point is that separate tokens also means the grammar will be wrong for some language or another, plural forms aside. For example, maybe some language has different rules about whether you use a conjunction or not when you string phrases together, depending on how many phrases. Or maybe "out of" has to be differently depending on whether there are two phrases after it or just one.
in reply to x0

@x0 Heh, I didn't mean to dig deep on this. My point is that it wouldn't be entirely grammatical on matter what you do, and I actually think that's fine. Screen reader output isn't inherently grammatical, despite what people try to suggest. And I think it would be far less efficient if it were fully grammatical. Imagine: "This is a button labelled OK which is currently pressed." Yuck.
@x0
in reply to x0

@x0 @jcsteh @NikJov The other consideration may also be that the current implementation reads the elements as it gets to them. The more analysis and processing that gets done, the more potential for sluggishness in responses (and remember arrowing through regular text would still need to check if it was entering or leaving anything and if anything needed to be processed).
in reply to Jamie Teh

@jcsteh @x0 Yeah, I imagined such a possible confusing example while writing this reply already. I would say speak the out of container message with a slight pause to make it even more obvious that it is not part of the text being read, but we're probably starting to get into something that would anoy people who actually want to know when they leave a container.
in reply to Jamie Teh

@jcsteh @NikJov So the question then becomes, would the amount of work be worth the reward for someone in core to do it? It's a UX change that would at least help with some of the spam generated by unwieldy HTML, which whether we like it or not happens all the time. Now when entering at the same time, this gets even more weird. Say emails. Out of table out of table table with one row and 3 columns table with two rows and 1 column etc. You can't really apply the same logic to the entrance when info has to be communicated about the container...
in reply to NV Access

@jcsteh @NikJov That might be more of an issue for quick navigation and that multimode setting I was talking about yeah. And gets slightly funky depending on exactly what bits in the architecture get involved to assemble such a message. If it generated something like closing events or tokens or whatever, you'd have to collect them all, sort them in some fashion, and then output that. IDK how that would be anymore sluggish than the current system, if a set is so nested it lags it would lag regardless because it's still jsut one string, no?
in reply to NV Access

@NikJov You are right. I've selected to hear those HTML elements, but since, unlike JAWS, NVDA doesn't display the beginning or end of these elements on separate lines, oftentimes we have to hear verbose stuff about entering or exitting nested lists, tables inside articles, etc. So the issue isn't with the announcement of those elements. Rather, the issue is with the way leaving, and entering, nested elements is handled by NVDA.