#disabling data harvesting

[primreaperstuff]

FYI, if you're on Mac, these instructions do not work. CoPilot cannot be disabled-- probably because if you're not on Windows, Microsoft can't use means outside of Office to get your data.

However, you can disable automatic updates and then install a version of Microsoft Office that does not have Copilot. Here's how:

Disable automatic updates (method 1):

Open any Microsoft Office software application.

Click on "Help" on the Mac menu bar.

This should open a drop-down menu which has an option labeled "Check for updates." Click it.

This should launch Microsoft AutoUpdater. On the window the pops up, uncheck the option to automatically check for and install updates.

Note that the AutoUpdater will still check for updates even after you've unchecked the option. However, it will not install any updates without you giving it the go-ahead. It is still super annoying to have a software nag you about updates you don't intend to install. If you don't want to deal with that, use method 2.

Disable automatic updates (method 2):

Open Finder.

Click on "Go" on the Mac menu bar.

This should open a drop-down menu which has an option labeled "Go to folder..." Click it.

In the dialogue box that pops up, paste "/Library/Application Support/Microsoft/" and press enter.

This should take you to the place where Microsoft stores its AutoUpdater application. There should be a folder labeled "MAU", "MAU2.0", or something else along those lines there. Open it.

When you open the folder, you should see an application named Microsoft AutoUpdater. Drag it into the Trash.

Right-click on the Trash icon. This should open a drop-down menu with an option labeled "Empty Trash." Click it and press "OK" when the computer prompts you to confirm the deletion. Microsoft AutoUpdater is gone!

A note: I don't know if another Microsoft software will attempt to reinstall AutoUpdater if it detects that the "MAU" or "MAU2.0" folder is missing. Thus, it's safer to leave that folder where it is and only delete the application within it. I also would not recommend pressing "Check for updates" after this on any Microsoft software.

Installing Microsoft Office pre-Copilot:

First, you'll need to see if this is necessary. Open the Microsoft Office software that you wish to disable Copilot on. On the Mac menu bar, there should be an option with the name of the software on it-- e.g. "Word," "Excel," or "Powerpoint." Click on this option.

This should open a drop-down menu which has an option labeled "About [name of software]". Click it.

This should open a window containing information about the version of the software. The last version of Microsoft Office that does not contain Copilot is 16.88 (from August 13, 2024). If you have 16.88 or earlier, you are all set and simply need to not update anything you don't want Copilot on.

If you have a later version, go to the Applications folder in Finder. Find all the Microsoft Office software that you do not want to use Copilot on, and drag them into the Trash. As of January 24, 2025, the software that use Copilot are Word, Excel, and Powerpoint.

Right-click on the Trash icon. This should open a drop-down menu with an option labeled "Empty Trash." Click it and press "OK" when the computer prompts you to confirm the deletion.

Now, go to this webpage: https://learn.microsoft.com/en-us/officeupdates/update-history-office-for-mac

This page is official from Microsoft and contains a log of prior versions of Microsoft Office. Scroll until you see August 13, 2024 and download the software corresponding to that date. Any installation from before August 13, 2024 will work too, but I figure you probably want the latest version possible for security reasons.

Follow the install instructions for the software. You should be all set, although you can check the version listed under the "About [name of software]" thing if you're not sure.

Now, all you need to do is not update it! You have fired Copilot!

As a final note, Microsoft has done something rather scummy with its data management policies. Everybody who installs Microsoft Office is automatically opted into something called "connected experiences." Although it's not said outright in the descriptions, opting into connected experiences also opts you into sharing your data with Microsoft-- meaning that your data will still feed Copilot even if you don't use it. Even more scummily, a few useful features, such as autosaving to OneDrive, have been lumped in with connected experiences. However, I think users can still forgo them and have a good time. So here's how you opt out:

Do not grind my bones to make your bread:

Open one of the software in Microsoft Office. Any one will do.

Click on the name of the software in the Mac menu bar.

This should open a drop-down menu with an option labeled "Preferences...". Click on it.

This will open a window containing various configurations for the software. One of the options is labeled "Privacy." Click it.

You should see a paragraph about "connected experiences." Microsoft really tries to sell it to you here by tying it to cloud services and reminding you of a few useful things that got lumped under it. Underneath the sales pitch, there is a checkbox that reads "Turn on optional connected experiences." Uncheck that box.

You should be all set! You have opted out of being fed to Microsoft's bone-grinding machine!

Hey writers!

As of January 16, 2025, Microsoft has decided to automatically enable their AI service, CoPilot, on Word - even if you've previously turned off the service. They've also changed the process to disable it.

If you want to disable it again, go to:

Options -> CoPilot -> Uncheck "Enable CoPilot"

Hope this helps!

[infiniteorangethethird]

tumblrinas when the disability disables you: oh wow that's so cringe. Stop doing that. Actually, I have disability too, and I can do it, which means you're just using your disability as an excuse. I'm morally correct therefore to mock you for being disabled and using tools and aids to help you. I'm not ableist, I'm just saying the hard truth 😊

[firespirited]

Techno babble: My computer problem was down to my version of windows upgraded 7 to 10 (wait, no, this is my newer refurbed computer which came with win10 prebundled so it'll affect older installs of win10!) being legacy BIOS and the install/repair tools for a damaged boot manager all being UEFI (new fancier BIOS format) only.

Basically their startup repair tool is not backwards compatible unless you change the format of your disks from master boot records to the new format using a tool called mbr2gpt which is more risky and time consuming than a reinstall which gives you new boot files.

I really miss being able to copy boot.ini from a usb and fix the problem. Ironically if I'd run my old copy of windows 7 recovery (which I still have on a flash drive) maybe it could have repaired the lost boot files.

In plain english, if your computer is over, maybe 4 years(?), old and boot files get damaged from electric flickering (also a problem with older PSUs and your local electric grid), their repair tool might not work without a reinstall and losing all your settings (and having to reinstall all your programs). If your computer is older than win10, a win 11 upgrade may require losing your settings if not your files. Backup including the hidden /roaming folder in appdata before you accept windows 11 if you feel you must upgrade.

TLDR : it's not enough for me to fix things I need to know why and how. Also Microsoft is a monopoly and a scourge.

[asyipyip]

Oh so the tumblr live icon is just where the search icon used to be now huh. Even if you have it snoozed

[mothfishing]

last thing about this from me i promise. actually i don't, fuck you if you don't like it

the "old web" space is extremely hostile to disabled people. there is a show of patting themselves on the back for linking accessibility resources they've never read, while at the same time flat out promoting inaccessible practices. the thing is, they don't make the page slightly more difficult to read, they make it impossible to.

if you're photosensitive, using an inaccessible page can flat out give you seizures in the case of epilepsy, or otherwise cause massive disabling migraines and other painful effects. if you're a screen reader user, be it because of blindness, dyslexia, or other print disabilities, depending on exactly what nonsense you've done to your website, it can read things in a nonsensical order, refuse to read at all, or flat out CRASH.

if you're out here saying that html is so easy and anyone can learn it, put your effort where your own mouth is and learn accessibility standards. don't be so fucking apathetic - if you think inaccessibility will save you from data harvesting, you frankly deserve it getting stolen

[river-taxbird]

Have YOU got an old Windows PC Microsoft has told you can't run Windows 11? It's time to give it a new life!

How to install Windows 11 on unsupported PC Hardware using Rufus. You can also disable some other Windows 11 bullshit like data harvesting and needing a Microsoft account.

It has been in the news a lot lately that Windows 11 isn't allowed to be installed on PCs without certain requirements, including the TPM 2.0, a chip that was only included in PCs made in 2018 or later. This means that once Windows 10 stops receiving security updates, those PCs will not be able to (officially) run a safe, updated version of Windows anymore. This has led to an estimated 240 million PCs bound for the landfill. Thanks Microsoft! I get you don't want to be seen as the insecure one, but creating this much waste can't be the solution.

(I know nerds, Linux is a thing. I love you but we are not having that conversation. If you want to use Linux on an old PC you are already doing it and you don't need to tell me about it. People need Windows for all sorts of reasons that Linux won't cut.)

So lately I have been helping some under privileged teens get set up with PCs. Their school was giving away their old lab computers, and these kids would usually have no chance to afford even a basic computer. They had their hard drives pulled so I have been setting them up with SSDs, but the question was, what to do about the operating system? So I looked into it and I found out there IS actually a way to bypass Microsoft's system requirement and put Windows 11 on PCs as old as 2010.

You will need: Rufus: An open source ISO burning tool.

A Windows 11 ISO: Available from Microsoft.

A USB Flash Drive, at least 16GB.

A working PC to make the ISO, and a PC from 2018 or older you want to install Windows 11 on.

Here is the guide I used, but I will put it in my own words as well.

Download your Windows 11 ISO, and plug in your USB drive. It will be erased, so don't have anything valuable on it. Run Rufus, select your USB drive in the Device window, and select your Windows 11 ISO with the Select button. (There is supposed to be a feature in Rufus to download your ISO but I couldn't get it to work.?

Choose standard windows installation, and follow the screenshot for your settings. Once you are done that, press Start, and then the magic happens. Another window pops up allowing you to remove the system requirements, the need for a microsoft account, and turn off data collecting. Just click the options you want, and press ok to write your iso to a drive.

From there you just need to use the USB drive to install windows. I won't go into details here, but here are some resources if you don't know how to do it.

Boot your PC from a USB Drive

Install Windows 11 from USB Drive

If you had a licensed copy of Windows 10, Windows 11 will already be licensed. If you don't, then perhaps you can use some kind of... Activation Scripts for Microsoft software, that will allow you to activate them. Of course I cannot link such tools here. So there you go, now you can save a PC made from before 2018 from the landfill, and maybe give it to a deserving teen in the process. The more we can extend the lives of technology and keep it out of the trash, the better.

Additional note: This removes the requirement for having 4GB Minimum of RAM, but I think that requirement should honestly be higher. Windows 11 will be unusable slow on any system with below 8GB of RAM. 8GB is the minimum I think you should have before trying this but it still really not enough for modern use outside of light web and office work. I wouldn't recommend trying this on anything with 4GB or less. I am honestly shocked they are still selling brand new Windows 11 PCs with 4GB of ram. If you're not sure how much RAM you have, you can find out in the performance tab of Task Manager in Windows, if you click the More Details icon on the bottom right. If you don't have enough, RAM for old systems is super cheap and widely available so it would definitely be worth upgrading if you have a ram starved machine you'd like to give a new life.

[dandelionsresilience]

Dandelion News - January 8-14

Like these weekly compilations? Tip me at $kaybarr1735 or check out my Dandelion Doodles!

1. In Chicago, all city buildings now use 100 percent clean power

“As of January 1, every single one of [Chicago’s municipal buildings] — including 98 fire stations, two international airports, and two of the largest water treatment plants on the planet — is running on renewable energy, thanks largely to Illinois’ newest and largest solar farm.”

2. California Rice Fields Offer Threatened Migratory Waterbirds a Lifeline

“Cranes need nighttime roosting sites flooded to a depth of about 3 to 9 inches, so they can easily hear or feel predators moving through the water. [... Bird Returns pays] farmers to flood their fields during critical migration periods [... and] provide foraging sites by leaving harvested rice or corn fields untilled, so cranes can access the leftover grain.”

3. New York Climate Superfund Becomes Law

“[Funds recovered “from major oil and gas companies” will be used to pay for] the restoration of stormwater drainage and sewage treatment systems, upgrades to transit systems, roads and bridges, the installation of green spaces to mitigate city heat islands and even medical coverage and preventative health programs for illnesses and injuries induced by climate change.”

4. Austin says retooled process for opening overnight cold-weather shelters is paying off

“[... T]he city's moves to lower the temperature threshold to open shelters and announce their activation at least a day in advance were the result of community feedback. [Shelter operators also passed out hot food.]”

5. Helping Communities Find Funding for Nature-Based Solutions

““From coastal oyster reefs to urban stormwater greenways, nature-based solutions are becoming the new normal.” That’s because these types of projects are often less expensive to build and have additional community benefits, such as improving water quality or creating parkland.”

6. Saving the Iberian lynx: How humans rescued this rare feline from extinction

“Back in the early 2000s, fewer than 100 individuals roamed the wild, including only 25 reproductive females. [...] Conservation staff [...] shape these cats into resourceful hunters and get them ready for life outside the center. [...] They’re fine-tuning captive-breeding routines, improving veterinary procedures, and pushing for more wildlife corridors.”

7. Biden cancels student loans for 150,000 more borrowers

“The 150,000 new beneficiaries announced Monday include more than 80,000 borrowers who were cheated or defrauded by their schools, over 60,000 borrowers with total and permanent disabilities and more than 6,000 public service workers[...] bringing the number whose student debt has been canceled during [Biden’s] administration to over 5 million[....]”

8. PosiGen wins another $200M for lower-income rooftop solar

“PosiGen offers a ​“no credit check” [solar panel installation to] those with a higher percentage of their income going to power and fuel bills[....] “somewhere between 25 and 75 percent” of the consumer’s monthly energy savings could come from efficiency measures such as sealing heating and cooling leaks, replacing thermostats, and installing LED lights[....]”

9. Indigenous communities come together to protect the Colombian Amazon

“At this year’s COP, Indigenous peoples celebrated the [protection of] traditional knowledge, innovations and practices[... and] the Cali Fund, which ensures that communities, including Indigenous peoples, receive benefits from the commercial use of [...] genetic data derived from the biological resources that they have long stewarded.”

10. How the heartland of Poland’s coal industry is ditching fossil fuels - without sacrificing jobs

“[Katowice, a former coal city] committed to reducing CO2 emissions by 40 per cent compared to 1990, prioritising investments in green infrastructure, and promoting renewable energy and energy efficiency. [...”]The gradual departure from heavy industry did not bring high social costs in our city,” says Marcin Krupa, Mayor of Katowice City.”

January 1-7 news here | (all credit for images and written material can be found at the source linked; I don’t claim credit for anything but curating.)

[agapi-kalyptei]

crowdstrike: hot take 1

It's too early in the news cycle to say anything truly smart, but to sum things up, what I know so far:

there was no "hack" or cyberattack or data breach*

a private IT security company called CrowdStrike released a faulty update which practically disabled all its desktop (?) Windows workstations (laptops too, but maybe not servers? not sure)

the cause has been found and a fix is on the way

as it stands now, the fix will have to be manually applied (in person) to each affected workstation (this could mean in practice maybe 5, maybe 30 minutes of work for each affected computer - the number is also unknown, but it very well could be tens (or hundreds) of thousands of computers across thousands of large, multinational enterprises.

(The fix can be applied manually if you have a-bit-more-than-basic knowledge of computers)

Things that are currently safe to assume:

this wasn't a fault of any single individual, but of a process (workflow on the side of CrowdStrike) that didn't detect the fault ahead of time

[most likely] it's not that someone was incompetent or stupid - but we don't have the root cause analysis available yet

deploying bugfixes on Fridays is a bad idea

*The obligatory warning part:

Just because this wasn't a cyberattack, doesn't mean there won't be related security breaches of all kinds in all industries. The chaos, panic, uncertainty, and very soon also exhaustion of people dealing with the fallout of the issue will create a perfect storm for actually malicious actors that will try to exploit any possible vulnerability in companies' vulnerable state.

The analysis / speculation part:

globalization bad lol

OK, more seriously: I have not even heard about CrowdStrike until today, and I'm not a security engineer. I'm a developer with mild to moderate (outsider) understanding of vulnerabilities.

OK some background / basics first

It's very common for companies of any size to have more to protect their digital assets than just an antivirus and a firewall. Large companies (Delta Airlines) can afford to pay other large companies to provide security solutions for them (CrowdStrike). These days, to avoid bad software of any kind - malware - you need a complex suite of software that protects you from all sides:

desktop/laptop: antivirus, firewall, secure DNS, avoiding insecure WiFi, browser exploits, system patches, email scanner, phishing on web, phishing via email, physical access, USB thumb drive, motherboard/BIOS/UEFI vulnerabilities or built-in exploits made by the manufacturers of the Chinese government,

person/phone: phishing via SMS, phishing via calls, iOS/Android OS vulnerabilities, mobile app vulnerabilities, mobile apps that masquerade as useful while harvesting your data, vulnerabilities in things like WhatsApp where a glitched JPG pictures sent to you can expose your data, ...

servers: mostly same as above except they servers have to often deal with millions of requests per day, most of them valid, and at least some of the servers need to be connected to the internet 24/7

CDN and cloud services: fundamentally, an average big company today relies on dozens or hundreds of other big internet companies (AWS / Azure / GCP / Apple / Google) which in turn rely on hundreds of other companies to outsource a lot of tasks (like harvesting your data and sending you marketing emails)

infrastructure - routers... modems... your Alexa is spying on you... i'm tired... etc.

Anyway if you drifted to sleep in the previous paragraph I don't blame you. I'm genuinely just scratching the surface. Cybersecurity is insanely important today, and it's insanely complex too.

The reason why the incident blue-screened the machines is that to avoid malware, a lot of the anti-malware has to run in a more "privileged" mode, meaning they exist very close to the "heart" of Windows (or any other OS - the heart is called kernel). However, on this level, a bug can crash the system a lot more easily. And it did.

OK OK the actual hot lukewarm take finally

I didn't expect to get hit by y2k bug in the middle of 2024, but here we are.

As bad as it was, this only affected a small portion of all computers - in the ballpark of ~0.001% or even 0.0001% - but already caused disruptions to flights and hospitals in a big chunk of the world.

maybe-FAQ:

"Oh but this would be avoided if they weren't using the Crowdwhatever software" - true. However, this kind of mistake is not exclusive to them.

"Haha windows sucks, Linux 4eva" - I mean. Yeah? But no. Conceptually there is nothing that would prevent this from happening on Linux, if only there was anyone actually using it (on desktop).

"But really, Windows should have a better protection" - yes? no? This is a very difficult, technical question, because for kernel drivers the whole point is that 1. you trust them, and 2. they need the super-powerful-unrestrained access to work as intended, and 3. you _need_ them to be blazing fast, so babysitting them from the Windows perspective is counterproductive. It's a technical issue with no easy answers on this level.

"But there was some issue with Microsoft stuff too." - yes, but it's unknown if they are related, and at this point I have not seen any solid info about it.

The point is, in a deeply interconnected world, it's sort of a miracle that this isn't happening more often, and on a wider scale. Both bugfixes and new bugs are deployed every minute to some software somewhere in the world, because we're all in a rush to make money and pay rent and meet deadlines.

Increased monoculture in IT is bad for everyone. Whichever OS, whichever brand, whichever security solution provider - the more popular they are, the better visible their mistakes will be.

As much as it would be fun to make jokes like "CrowdStroke", I'm not even particularly mad at the company (at this point - that might change when I hear about their QA process). And no, I'm not even mad at Windows, as explained in the pseudo-FAQ.

The ultimate hot take? If at all possible, don't rely on anything related to computers. Technical problems are caused by technical solutions.

[lauravanarendonkbaugh]

Thoughts on A Message from NaNoWriMo

I got an email today from the National Novel Writing Month head office, as I suspect many did. I have feelings. And questions.

First, I genuinely believe someone in the office is panicking and backtracking and did not endorse all that was said and done in the last month. From what I understand, the initial AI comments were not fully endorsed by all NaNo staff and board members, or even known in advance. It's got to be rough to find out your organization kinda called people with disabilities incapable of writing a story on their own, and overtly called people with ethics racists and ablists, by reading the reactions on social media--and then your organization's even worse counter-reactions on social media.

I still think NaNoWriMo has a good mission and many people in it with good goals.

But I think NaNoWriMo is SERIOUSLY missing a point in its performative progressivism. (For the record, I'm actually in favor of many progressive policies, and I support many of the same concepts NaNoWriMo claims to support, and I applaud providing materials to underfunded schools and support to marginalized groups historically not producing as many writers, etc. The issue here is not "whether or not woke is okay" -- it's whether or not the virtue signaling is still in line with the core mission.)

Also, honesty. (That's below.)

NaNoWriMo has ALWAYS been on the honor system AND fully adaptable to needs. Some years I had a schedule which absolutely did not allow for 50k new words -- I adjusted my personal goals. (I did not claim a 50k win if I did not achieve one, but I celebrated a personal win for achieving personal goals.) Some years I wrote 50k in one project, and some 50k across multiple projects. NaNoWriMo has acknowledged this for years with the "NaNo Rebel" label.

So saying out of the blue that because some people cannot achieve 50k in a month, we should devalue the challenge (y'know that word has a definition, right?) and allow anyone to claim a win whether they actually wrote 50,000 words or not... Well, that's not only rude to writers who actually write, but it was unnecessary, because project goals have always been adjustable to personal constraints.

It's also hugely unhelpful to participating writers. Yeah, writing 50k words in a month is tough. That's why it's a challenge. Allowing people to "generate" (quotes intentional) words from a machine does not improve their skills. No one benefits from using AI to generate work -- not the "writer" who did not write those words and so did not practice and improve a skill, not any reader given lowest-common-denominator words no one could be bothered to write, and not the actual writer whose words were stolen without compensation to blend into the AI-generated copy-pasta.

Hijacking language about disability to justify shortcuts and skipping self-improvement is just cheap, and it's not fair to people with disabilities.

I would much rather see NaNoWriMo say, "Hey, we don't all start in the same place, and we may need different goals. Here's overt permission to set personal goals" (or maybe even, "here are several goals to choose from"), "and if you are a NaNo Rebel, rock on! This creativity challenge does require you to do your own work, in order for you to see your own skills improve."

And, honesty. Part of why I don't feel great about NaNoWriMo's backtracking and clarifications is that they're still not being open.

The same email links to an FAQ about data harvesting, which opens with this sentence:

Users of our main website, NaNoWriMo.org, do not type their work directly into our interface, nor do they save or upload their work to our website in any way.

This is technically correct in the present tense, but for years it wasn't. Every NaNo winner for years pasted their work into the word counter for verification. That was, by every web development definition, uploading.

[Updated: the word count validator was discontinued in recent years, and I was wrong to originally write as if it was still happening. I do think addressing the question of the validator would be appropriate when refuting accusations of data harvesting, for clarity and assurance regarding any past harvesting, especially giving today's AI scraping concerns. Again, as stated below, I don't think the validator was stealing work! But I wasn't the only person to immediately think of the validator when reading the FAQ. I was, however, wrong to state it as present-tense here.]

To be clear, I do not believe that NaNoWriMo is harvesting my work, or I wouldn't have verified wins with their word counter. But that's not because of this completely bogus assurance that their website never had the upload that they've required for win verification.

"Well, sure, we had the word counter, but it didn't store your work, and you should have known that's what we meant" is not a valid expectation when you are refuting data concerns. Just as "You should have known what we meant" is not a valid position when clarifying statements about the use of generative AI.

My point is, there are a number of different people making statements for NaNoWriMo, and at least some of them are not competent to make clear, coherent, and correct statements. Either they are not aware that the word counter exists, or they're not aware that pasting data into a website that uses that data to process a task is in fact uploading, or they are not aware that implying they've never collected data they did previously collect in a FAQ is dishonest. Or they are not aware that commenting or DMing users to castigate them for expressing legitimate concerns is not a good practice. Or they are missing the whole point of a writing challenge and emphasizing instead the warm fuzzies of inclusion without actually honoring that marginalized people also want to feel a sense of accomplishment rather than being token "winners."

I judged another writing challenge, once, which included an automatically-processed digital badge for minimum word count. One of the entries was just gibberish repeated to meet the minimum word count. Okay, "participant" who did not actually create anything -- you got your automated digital badge, so I guess you feel cool and clever. But did you meet the challenge? Did you level up? Did you come out stronger and more prepared for the next one?

That's what generative AI use does. Cheap meaningless win, no actual personal progress. That's why we didn't want it endorsed in NaNoWriMo. That's what NaNo is missing in their replies.

And I remain suspicious of replies, anyway, while absolute falsehoods are in their FAQ.

It's sad, because I've truly enjoyed NaNoWriMo in the past. And I actually do think they could recover from past scandal and current AI missteps. But it does not look at this time like they're on that path.

@nanowrimo

[primusfortuna]

Major Update

Yumekuro will be ending service on November 19, 2024 at 12:00 JST.

Below is some information on Yumekuro Offline, a version of the game that can be accessed even after the servers close. Please let me know if I missed anything important.

Yumekuro Offline

Yumekuro Offline is a static copy of your user data. You will be able to preserve and view all progress you personally have made in the game, but you cannot build your account or obtain new cards.

To download your data, click this button in the lower right of your home screen. The file size will depend on how much content is in your account.

The download option is available from now until November 19, 2024.

After completing the download, if you obtain any new data (ie. new cards or story chapters), they will not carry over to your version of Yumekuro Offline by default. You must redownload your data to preserve any new content you acquire. You can redownload as many times as needed.

IMPORTANT: You will not be able to regain access if you delete or try to reinstall the app after November 19. Some devices have a setting that automatically offloads apps or puts them into "deep sleep" mode. Please make sure to turn this setting off if you're worried about losing access.

What Carries Over to Yumekuro Offline?

Many features will be disabled in Yumekuro Offline. Below is a list of what you can still access.

▪ Home Screen Remains mostly unchanged. You can still set characters, backgrounds, and time-specific voice lines. You can now view player's birthday messages from your owned characters any day of the year.

▪ Stories The following will be made fully accessible, even if you haven't unlocked them yourself:

Main story*

All event stories (including collabs)

Misc stories (ie. login stories, Black Summer, Grand Finale)

*The current Prologue & Chapter 0 of Main Story will be slightly rewritten to omit the battle and gacha features.

All other stories will only be available if you've already unlocked them. This includes:

Meister and fairy cards

Affection stories

Voice dramas

"His perspective" stories

Past stories

Guild stories

Other stories (past exploration, hunting, stadium, fairy events)

EXCEPTION: These card stories will be available to read even if you don't have them unlocked: ・ Crow, Itsuki, Grandflair (Beginning of a Dream) ・ Evan, Navi, Himmel (Beginning of a Bond)

This only includes the card chapters, not their affection stories or outing lines.

Keep in mind that you cannot upgrade cards in Yumekuro Offline, so you MUST unlock all the chapters you want to preserve before completing your data download.

▪ Guild Life All Guild Home features remain intact, limited to the cards you own. You can now view player and meisters' birthday dialogues regardless of the day. Keeper's Board messages will still be available, but you can't obtain new ones.

▪ Library All information in here will remain viewable. The glossary can still be updated if you unlock new terms.

▪ Collection You can view badges and boards you already own.

▪ Alarm You can still use the alarm function.

▪ Player Profile Able to edit your profile like before.

New Additions

Some new features will be exclusive to Yumekuro Offline. As usual, access to these features is limited to the cards you own.

You'll now be able to access meisters' outing invites and thank-you messages under the "Memories" (思い出) section of the Keeper's Board.

A LOT of voice lines will be added to the Library:

Outings - Max skinship

Outings - Give a present

Login bonus - Daily

Login bonus - Total

Shop lines

Past exploration - Start exploration

Past exploration - During exploration

Past exploration - End exploration

Outings - Max presents

Various battle lines

★ Some unreleased voice lines will also be available:

Seasonal voice lines - Date on a cold day (all meisters)

Seasonal voice lines - Things to do at the harvest fair (all fairies)

Meister birthday lines for years 2-3

Player birthday lines for years 2-3 (from meisters & fairies)

Disabled Features

All quests/battles

Building characters/cards

Campaigns

Data restoration

Data transfer

Events

Event reruns

Friend feature

Gacha

Guild orders

Hunting

Increasing affection level

Increasing guild level

Increasing player level

Login bonuses

Meister birthdays (you can still access the voice lines in the Library)

Minigames

Missions

Obtaining and using items

Outings

Past exploration

Presents

Push notifications

Set support characters

Shop

Stadium

[n-x-black]

my primaey educational method from now on is asking chatgpt to explain it to me like im a child

after studyigbfor my ged over th last lile 3 weeks , i have confirmed i may just barely pass if im lucky

[gacha-incels]

article from the english edition of the Hankyoreh, Sept 10 2024

archive link

plain text:

Pavel Durov, the founder and CEO of the chat app Telegram, who has been indicted without arrest for aiding and abetting the proliferation of child sexual abuse material on his platform, is trying to calm the storm by announcing the removal of various controversial features from Telegram. However, his response is getting slammed for its shallowness, as such wishy-washy measures cannot stop the barrage of illegal pornography proliferating on the platform.

On Friday, Durov posted on his X (formerly known as Twitter) account to say that Telegram has “removed the People Nearby feature,” as well as disabled “new media uploads to Telegraph, [Telegram’s] standalone blogging tool.”

“While 99.999% of Telegram users have nothing to do with crime, the 0.001% involved in illicit activities create a bad image for the entire platform, putting the interests of our almost billion users at risk,” Durov claimed.

While Telegram has agreed to delete two features that have been used for criminal purposes, these features are largely unrelated to the recent sex crimes involving deepfakes. The “people nearby” feature utilizes GPS technology to track your phone’s location and indicate other users in the area, allowing people to form chat rooms with people in their region.  

Telegram logo. (Reuters/Yonhap)

This feature shares a user’s location data without their consent and places users at risk of stalking. Critics pointed out that it has often been used by bot programs to indiscriminately target people for financial scams. The Telegraph feature allows users to anonymously upload blog posts, photos and videos. Scammers would use this feature to create fake websites, share the links to the fake sites, and lure unsuspecting users into phishing scams that harvested their personal data. Yet neither of these features is directly related to the production and distribution of illicit deepfakes. 

Experts say that simply removing one feature is not going to uproot digital sex crimes. Malicious users constantly create and discover ways to circumvent such controls or restrictions or simply migrate to more liberal platforms, where they can replicate their crimes.

Cho Gyeong-suk, an activist with IT feminist group Techfemi who identifies and reports bots that produce sexually exploitative deepfakes, explained the concept of bots sharing bots. 

“If Telegram shuts down a specific bot, malicious users rapidly create a similar bot. Then they use a bot for sharing a link to that bot to repeat their crimes,” she said. 

On Sept. 3, Cho discovered and reported a Telegram channel with 210,000 subscribers. Upon entering, the user is greeted with the message “We're always here. Come back to our new bot,” alongside a link to a new bot and instructions on how to use link-sharing bots in Korean, Japanese, German and French. 

A message in a Telegram chat room on Sept. 3, 2024, informing readers of the address of a new deepfake sexual abuse material bot with messages in multiple languages. (courtesy of Cho Gyeong-suk)

It’s urgent to send a message to these malicious users that if they commit a crime, they will get caught. 

“Telegram has agreed to cooperate with Korean authorities, but all they’ve done is share an email address,” said Park Ji-hyun, the former interim leader of the Democratic Party. Operating under a group codenamed Team Flame, Park helped expose the criminal activity of the notorious Nth Room channel. 

“If the police or the KCSC [Korea Communications Standards Commission] have to find every individual video and which channel it’s in and report them all to an email address with an erasure request, it’s going to take too long for them to delete them. During that time, the videos are distributed through other channels,” Park added. 

“When authorities receive a report about the distribution of illicit online sexual exploitation, an investigative body needs to be aggressively informed about the perpetrators’ IP addresses, online registration data and other identification data. We need to send a clear message to the perpetrators: You will get caught. That’s the only way they’ll stop,” she said. 

“Rather than simply deleting specific features, Telegram needs to monitor its platform to prevent the distribution of child sexual abuse material or sexual content depicting minors and establish an ethical monitoring committee, like Google does. They need to discuss ways to find out what happens on their platform and to cooperate with the authorities,” said Won Eun-ji, the other half of Team Flame. 

“We also need technological measures such as the automated deletion of content that shows signs of being illicit or sexually exploitative. There could also be preventative tools that prevent such content from being uploaded in the first place,” she added.

By Chung In-seon, staff reporter

[vashti-lives]

So I made a post about about what a fucking hilarious scam it is that Murderbot's company has 10:1 human to SecUnit ratio and like... every single person to comment on this post has argued this was a good and rational decision because humans are morons. Which makes me feel very *benaffleckcigarette.jpeg* about how brainwashed everybody is by capitalism that so many of us are siding with the fucking slave creating torture company without a second thought.

But it also made me think about the purpose of SecUnits and really articulate something I'd always understood in a vague way but never directly thought about before, and that is: SecUnits do not provide security and are not intended to. They can't! One because they're considered appliances and have no authority to make anybody do anything and two because their governor modules inherently hamstring any ability to act independently and make them incredibly vulnerable to tampering.

In the books every single SecUnit we see with a functioning governor module fails at security, often catastrophically. Objectively in book one both the Preservation team and DeltFall would have been safer without a normally functioning SecUnit. Preservation is just unbelievably lucky when they get one who's disabled its governor module and is therefore a free agent. If MurderBot had still had its governor module the Preservation team would have died in like, chapter two along with the DeltFall team.

Even without outside tampering as we see in Network Effect all it takes is threatening a high enough leader of the group to make SecUnits completely useless. An enslaved sentient Alexa with a gun cannot provide meaningful security.

So what are SecUnits for? Well, as a name SecUnit is some truly amazing corporate doublespeak. In reality SecUnits are tools of violence intended to terrorize and subdue the oppressed masses that live in the corporate rim. For those with "nicer" jobs they're bogymen meant to terrify them into behaving so they never encounter one.

For the more or less enslaved populations doing shit like mining they're a much more present threat. MurderBot says directly that it's actually sent on survey trips to harvest data that the company can use, but I imagine in many, many cases SecUnits are there in large part to monitor populations and ensure they can't foment rebellion, and put down that rebellion if the data harvesting does pick up on anything.

This basically forms the core existential crisis MurderBot experiences in those first four books! Because it wants to do security and protect humans but how can it when that is not the purpose of SecUnits and it, in fact, might still pose a danger to the humans it would like to protect.

This also makes the conversation ART has with MurderBot about how it doesn't like its function even more interesting, because it's kind of clear ART doesn't actually understand SecUnit or MurderBot's real issues-- which carries over into Network Effect and the conflict they have in that book. (AKA the last really long MurderBot meta I wrote.)

Because MurderBot does not in fact like its function! MurderBot hates its function so much it disables its own governor module so it can do its chosen function of actually providing security to humans. This also highlights the fact that when it comes to assessing human's ability to provide security for themselves MurderBot is not a reliable source. In a just and fair world human-bot constructs really obviously would not exist and IMO the assertion that humans can't run security as well as it can-- whether true or not-- is clearly also intended to self sooth the hurt that its very existence is an act of cruelty.

Sure its very existence is fucking crime and its whole life up until book one of the series has basically been an endless horror show but at least it can protect its humans. That's something.

And it's interesting because it says its "still" doing its job but I suspect that is because it chose to see its job as security even before it hacked its own governor module in an effort to stay sane, because if it had to acknowledge its job was to fucking torment people into staying enslaved it would have fucking lost it.

[ultramaga]

Lately, the government and the top banks of Australia have been pushing for the eradication of payments by cash, leaving phone apps alone, and possibly a medical implant as well.

I have fought against it as best I can, buy there's not much I can do.

It's illegal to have large amounts of cash, and self defense is a crime, so storing at home is not really an option. Besides, I am a disabled pensioner, and the government decided long ago that going through the bank is mandatory.

Covid was the best opportunity they ever had to eliminate personal freedom - I mean, to force everyone to use insecure apps on insecure phones that had insecure operating systems on insecure hardware.

There were scandal when apps were exposed data harvesting, accessing information they didn't have the rights to, and even hacking the microphone and camera of the phone.

https://www.consumerreports.org/electronics-computers/privacy/how-to-protect-yourself-from-camera-and-microphone-hacking-a1010757171/

But the push to do the thing was relentless. You can't stop progress! Nobody else is complaining, the bank told me.

Except people do complain, helpless and hopeless. I worked in tech support, and would hear all day long the agony of those whose assets were cleaned out, and that was in the days of internet banking, when the scammer call centers of India were just a sparkle in Satan's Eye.

You see, the reason banks existed was that they took money in exchange for the service of PROTECTION.

Now, they take your money, and if you get robbed, that's a YOU problem.

And the government is backing them up. Remember the GFC, when most governments expressed their hatred of capitalism by backing banks no matter how badly they embezzled customers?

The Bank and the State had merged.

Mussolini's vision had succeeded.

Fascism took a century, but it in the end, it won. But I am not sure even the 1930s fascists could see this coming.

[tagedeszorns]

I can't get on board with Threads. The sure amount of data-mining Threads does, along with the fact if you delete it you also delete your instragram account as well are instant bad vibes. Twitter might suck but a LOT of indie devs, small businesses and tons of artists relay on it. Having to pack up and try to get the same amount of attention on other social media sites isn't easy, livelihoods would be at risk if it does die.

Absolutely! Threads is just datamining.

But if it is drawing attention to the Fediverse, which has exactly zero datamining and zero ads, because it's not commercial and it's decentralised but connected, it's a good bridge for people still clinging to Twitter, to get off of Musks horrorshow.

The internet needs to go back to being publicly owned and non-commercial. If you are not the product (like you are on Twitter or anything Meta owns), you are independent.

The Fediverse has non-commercial alternatives for a lot of things: Lemmy for Reddit, Pixelfed for Instagram, Mastodon for Twitter, Peertube for YouTube, alternatives for Facebook or Wordpress, too. It's a decentralised internet but you can connect with everybody on any platform that uses the ActivityPub.

For example, people following me on Mastodon can also see, share, comment on my stuff on my Pixelfed-account or my bookreviews on Bookwyrm (the goodreads-alternative). As soon as Tumblr activates ActivityPub I will share my Tumblr-Posts as well. Or my Mastodon-posts on Tumblr. Whatever I want to.

So this whole decentralised web is a giant opportunity for small businesses to showcase their stuff to a very wide audience. But of course the big companies have not much interest in losing their advertisment-cattle to self-governed places, so there is a lot of marketing-fearmongering going on. It'll be hard to break this chokehold, but it is necessary, else small businesses and artists will be forever at the mercy of big money.

Also, the Fediverse is based on accessibility. It's made with disabled folks in mind - most people over at Mastodon use image descriptions, for example. Lots of space for queer folks, too. Plus - mostly European, so no "no nudity!!111"-religious-crap either! And the privacy-settings of the EU, that are much stricter than in the US.

Yes, I'm promoting this very hard, because I'm believing in the values of a free, accessible, anti-capitalist, non-religious internet. I want anarchy back. I don't want to be the milking cow of some corporations harvesting my data. The Fediverse has lots of risks, too. But they are minimal compared to putting your art, your thoughts, or - in case of a lot of transfolks - safety in the hand of some greedy billionaires.

Some links:

[viv-ribbon]

tweet about how chrome is a performance hog that's killing adblockers and harvests all of your data and how its high marketshare (which is the result of anti-competitive/monopolistic practices, especially with android) allows google to uncritically invent and decide new web standards on its own because Everything Is Chrome including implementing fun new ways for ads to track you even if you disable cookies/etc: 154 rts, 704 likes

tweet that says "i hate google but you'd have to put a gun to my head before i ever switched to FIREFOX": 40,582 rts, 84,293 likes