Jump to content

piemadd

Member
  • Posts

    1,466
  • Joined

  • Last visited

Posts posted by piemadd

  1. It is possible to do, you'll need to get something like this for a few bucks:
    https://www.ebay.com/itm/143731103450. You can find one by searching "LGA 771 to 775 Adapter" on pretty much any online retailer. Additionally, you'll need to cut some pieces of plastic off of your cpu socket. Delidded.com has a nice guide on how to do everything, which can be viewed here: https://www.delidded.com/lga-771-to-775-adapter/

  2. If you want to see if this is a hardware issue or a software issue, I would try getting a sample blender scene (this is a nice and complex one from the blender website), make sure you have optix enabled in your blender properties (how to change it) and in your render properties make sure to use cycles and use gpu compute (how to). Then, change the preview mode to full render (how to) and duplicate items (left click to select and shift+d to duplicate) until you are satisfied with how much vram is being used.

    To truly answer your question, it is actually quite rare for video games to fill up their VRAM, and features such as the improved speed of GDDR6X and soon cpus having direct access to GPU memory really help optimize this.

  3. 1 hour ago, SPARTAN VI said:

    Personally would only use that short cable to setup Virtual Desktop Streaming instead. You'll only really need the cable for SideQuest to sideload Virtual Desktop. It's a bit involved, but if you've got a solid 5GHz WiFi setup, you might not ever need to use a physical cable for Oculus Link: https://uploadvr.com/how-to-play-pc-vr-oculus-quest-2/

    Yeah I was going to do that til I remembered the 5GHz wifi drops off in my room and 2.4 wouldn't cut it. I'm debating about just getting a 5GHz WAP for my room which could also help my other devices but since it will just be cheaper to get something from anker/aliexpress im probably going to go that route

  4. So I recently got a Quest 2 and it surprising arrived in just 5 days (was only on backorder for a day with newegg) and it arrived today, before I realized I never ordered a cable for the headset. I do plan on ordering one from anker to arrive in a few days, but want to at least get something temporary which will suit me well until it arrives. I started checking around the specs of the official quest link cable when i saw something which I just wanted to confirm. The official cable is USB 3.2 Gen 2 (5 Gbps) (see figure 1) but a shorter (and cheaper) cable from microcenter is only 3.1 gen 1, while still having 5Gbps of capacity. (see figure 2). 

    Would the cable from microcenter still work at the same speed as the quest link cable? Also do note this is a temporary solution so just a small bit of quality/frame loss is fine to me.

    Fig 1:

     image.png.5163f005ac47e24562e5513a5d839bbe.png

    Fig 2:

    image.png.3f5796ace424dd5394896a420201cb15.png

    image.png.c6ba5d5a619dd8b107d801545ed324ee.png

  5. Note: Yes I do know this topic is very grey area. I asked a moderator via pm if I should post this but as of now have not heard back. Chances are I have asked for this thread to be immediately locked upon posting since the fact that this topic is grey area brings the probability of it becoming off-topic in the comments below very likely. If you would like to read my original post about the parler archiving, it is linked below:

     

    I have also redacted some words at my own discretion to remove any mentions of politics in the first case.

     

    Summary

    A researcher/programmer is turning the data archived from parler into an interactive map of videos in {{ certain_city }} from {{ certain_date }} during the {{ certain_event }}. 

     

    Quotes

    Quote

    Article 1: A developer going by the name of Patr10tic has created an interactive map using GPS metadata and videos from the app Parler to offer a glimpse of what was happening inside and around the {{ certain_place }} during the {{ certain_event }} on {{ certain_date }}.

    The project is an expansion on similar maps like the one Gizmodo created to illustrate just how many Parler users were part of the mob that occupied the {{ certain_place }}, based off an archive created by a researcher known as @donk_enby. Patr10tic’s map (currently named “Y’all Qaeda”) uses the same archive and connects GPS coordinates with actual video, offering a map of red icons placed around the {{ certain_place }} for each associated video post.

    We're actually starting to see the data pulled from parler to donk_enby's archive. I expected like this to happen, but am kind of surprised it happened so quickly. No doubt we will see an expansion of the map as time goes on, but it is interesting to see all of the data collected compiled in a way which is easy to visualize. Kind of reminds me of Snapchat's snap map.

     

    Quote

    Article 1: Viewing the videos has an unsettling quality, not just because of the intersection of {{ political_topics }} on display, but also the perspective they offer on the {{ certain_place }} itself. Most people haven’t seen the chambers of {{ certain_place }} and the halls of the {{ certain_place }} outside of press photos or C-SPAN live feeds. Seeing protesters milling about in {{ certain_place }} feels strange, but it’s even stranger seeing them wander through the {{ certain_place }}, unbothered, with {{ higher_force }} looking on.

    I took a look at the videos myself, and it is fairly chilling seeing just the quality of the videos captured and uploaded. Most of the time when there is a major event with protestors/riotors, you'll get a few low quality videos reposted multiple times across snapchat, Instagram, Facebook, etc, but hundreds of full HD videos being released just like that (snap) just seems off and not common to me. 
     

    Quote

    Article 2: “About half of the metadata I have I got from the source before the leak occurred," Patr10tic told Motherboard. "I found a way to download just the first several kilobytes of each file, and use exiftool to parse the GPS coordinates. I got about half a million headers that way (from donks url list), but abandoned that effort after she just released the whole metadata collection herself. Exiftool is hella slow!”

    VICE actually got a chance to interview Patr10tic, and it is interesting how even if the leak didn't happen, there still would have been some sort of map of the events which unfolded on {{ certain_date }}. Additionally, this really enforces the idea that immediately after any data leak, the information leaked/archived is going to immediately be used, whether it is for legal or illegal purposes.

     

    Quote

    Article 2: "I didn't need terabytes and terabytes of videos, just the ones on the {{ certain_place }} grounds, so I sorted through the gps coordinates based on proximity, and obtained the video IDs for just those files," Patr10tic said. "Then I downloaded them from a mirror that someone else is hosting."

    Really interesting how transparent Patr10tic is on how they collected the data, just like donk_enby was. There is also quite clearly a community effort in this map, not just a single person, a repeat of how donk_enby's archive was created. 
     

    My thoughts

    It is definitely interesting to see the perspective of users of data from a leak/archive, as usually that sort of stuff happens in private message rooms on the dark web with illegally collected data. That is really all I can say about this matter without getting to a political stance. I hope you learned something from this writeup as I did, as it is not just about the Parler leak/archival, but more towards the incredibly small amount of time between such a leak/archival and it being used in a format which your everyday joe can look at and understand. 
     

    Sources

    Article 1: https://www.theverge.com/2021/1/14/22231749/parler-interactive-map-video-gps-capitol-attack
    Article 2: https://www.vice.com/en/article/pkdmm8/developer-makes-interactive-map-of-parler-videos-from-capitol-hill-riots
    Map: https://thepatr10t.github.io/yall-Qaeda/
    Github Repo: https://github.com/ThePatr10t/yall-Qaeda

  6.   

    33 minutes ago, Uttamattamakin said:

    The description of this Hack is completely in accurate.

    I updated that, Blade of Grass pointed it out. 

    33 minutes ago, Uttamattamakin said:

    The information was not stored in the clear.  They took our information put it into one archive file then ran a sha 256 hash.  That hash is saved.  

    Hashes are one-directional. You can't store data in a hash, its impossible. The only thing stored as a 256 bit hash would be your password and then also your ssl connection to their servers. 

    33 minutes ago, Uttamattamakin said:

    As the above reddit post shows NOTHING private was downloaded.  Think about it.  Do you know how rare multi terabyte hard disk are?  At 50-100 TB we'd have to be talking hard disk.  

    Considering the speed of even the best home internet it would take days and days to download all of the information.  Mathematically, and technologically what they are saying is impossible.  

    The data was split between many many computers (in all somewhere in the neighborhood of 50-100 people helped download data) which allowed them to download everything at the speed they did. It wasn't a single internet connection, but many. Hell, a single gigabit connection would be able to do the job in less than 5 days. I don't doubt multiple gigabit connections were doing the work.  

  7. 36 minutes ago, gabrielcarvfer said:

    A "researcher" stolen information, published it, claimed they failed to enforce their rules...

    Yes a researcher. Not stolen, scraped. Nothing illegal happened. They never claimed parler wasnt enforcing its own rules, only that it had become a platform harboring hate speech and violence.

    36 minutes ago, gabrielcarvfer said:

    What a coincidence this came out after every other social network, that also failed to do the same, started losing users to them...

    The amount of users lost was minimal. Also not a coincidence. This was started when parler was announced to be shut down in the next few days by apple, google, and amazon.

    36 minutes ago, gabrielcarvfer said:

    Because there's nothing wrong with Apple/Google prohibiting porn/hate speech on the stores but allowing Twitter and Facebook...

    They do their best to take it down. Can you find it if you look? Yes. will it be there 5, 10, maybe 15 minutes from now? Most of the time no. THats because they use automated systems to flag the content and take it down

  8. 12 minutes ago, Taf the Ghost said:

    Use data presented, yes. Use it in Court, 90% unlikely. It can be done, see the Dread Pirate Roberts case, but it takes some serious work on the Prosecutor's side to accomplish. They almost rarely won't do that amount of work. (In the Dread Pirate Roberts case, they had a FBI employee that scammed a huge supply of Bitcoins during the case, so he got a bunch of data illegally in the process.)

     

    As to the topic, somewhere around the 10 Gb range, this stops being "research" and became an attack surface discovery & exploitation. The only groups that have the type of use for that much data are Security Services or Intelligence Agencies. It wasn't called "Total Information Awareness" for nothing. There's still money to be made when you find a treasure trove.

    IMO i feel like the FBI is going to use the info from this leak to then know who they need to focus on. Then they might do something like use security camera footage in the [unnamed building] and also publicly available info from parlor (ie a selfie taken by a proposed defendant for [doing something in] the [unnamed building]) to gather enough evidence to get a case. Will the parlor leak be presented in court? probably not. will it be used in investigations? i dont doubt it 

  9. Disclaimer: Please do not talk about politics here! This post is meant to surround the data leak and the data leak only. It is again the rules of the forum to talk about politics!

     

    I have removed any mentions of political anything from my quotes, so again, I am doing my best to keep politics out of this. That is why some quotes seem to start and end in the middle of sentences.

     

    Summary

    I do good with bullet points so here goes:

    • Parler is built on WordPress
    • One of the WordPress plugins was vulnerable
    • A Security researcher was able to use this plugin to create admin accounts on the platform
    • These admin accounts were used to pull as much data from the platform as possible

     

    Quotes

    Quote

    Article 1: ...a lone researcher began an effort to catalogue the posts of social media users across Parler, a platform founded to provide [redacted] users a safe haven for uninhibited “free speech”...

    Security researcher did the work. Now lets take a look into what they uncovered and how.
     

    Quote

    Article 1: The researcher, who asked to be referred to by their Twitter handle, @donk_enby, began with the goal of archiving every post from January 6, the day of [absolutely nothing happening]; what she called a bevy of “very incriminating” evidence. According to the Atlantic Council’s Digital Forensic Research Lab, among other sources, Parler is one of a several apps used by the [redacted] to coordinate their [redacted]...

    So it seems like said researcher believes there is criminal evidence on this platform for an event which shall not be named.
     

    Quote

    Article 1: Operating on little sleep, @donk_enby began the work of archiving all of Parler’s posts, ultimately capturing around 99 percent of its content. In a tweet early Sunday, @donk_enby said she was crawling some 1.1 million Parler video URLs. “These are the original, unprocessed, raw files as uploaded to Parler with all associated metadata,” she said. Included in this data tranche, now more than 56 terabytes in size, @donk_enby confirmed that the raw video files include GPS metadata pointing to exact locations of where the videos were taken.

    This is where it gets crazy. If you missed it, Parler failed to removed identifying metadata from the videos hosted on their platform, giving gps location data of each of the users. But wait, it gets worse... 
     

    Quote

    Article 2: In their viral post, the Redditor asserted that one of Parler's hosting platforms, Twilio, accidentally exposed the app's security authentications via a press release. This in turn could have allowed any person to create a blank administrator account and access all of Parler's private content, which, besides message history and geo data, might have included users' driver's license photos, which were used to create a verified account.

    Oof is really all i can say here. All I can assume now is that there will be people who will match driver's licenses with geolocation data of images to forward any criminal activity onto the FBI and/or local law enforcement.

     

    UPDATE: As @Blade of Grasspointed out, the archive pulled its data from a poorly created api. You can read about it in the tweet below:
     

    Spoiler

     


     

    Quote

    Article 1: The privacy implications are obvious, but the copious data may also serve as a fertile hunting ground for law enforcement. Federal and local authorities have arrested dozens of suspects in recent days accused of taking part in [nothing]

    Beyond the privacy implications, we really don't know what will happen to those whose data is leaked. There is no doubt that a database of compiled data will be up for sale on the dark web within weeks, or even days for that matter.


     

    Quote

    Article 2: According to tech writer Matthew Sheffield, the breach was possible due to Parler's long-criticized lax security standards. Specifically, Sheffield blames the potential leak on the app "never actually deleting anything its users posted," while keeping the data accessible to administrator users.

    One final thing we can see here is that no data has ever been deleted from Parler, on top of their lax security. This is one of the reasons why the leakers could access so much information, even if it was deleted.
     

    My thoughts

    To put it simply, I am disappointed. I was not a user of Parler, but I do have to say if you market your platform towards a growing group of people not satisfied with their current social media options for one reason or another, you should do your best to make sure the security of your users is protected as much as possible. Who knows what will happen with the data, but we can for sure know that it is going to spread like wildfire.

     

    Sources

    Article 1: https://gizmodo.com/every-deleted-parler-post-many-with-users-location-dat-1846032466

    Article 2: https://www.rt.com/usa/512152-parler-hacker-data-leak/

  10. We've recently been redesigning the social view over here at Repl.it, and because of this, the current embeds just break Here is an example:
     


    Looks cluttered, and there also isn't a (very visible) run button. There is really a simple fix for this. If you look into the url of the iframe, you will notice the attribute "lite=true" has been added on. Now, lets see what happens when we change this to "embed=true". 

    image.png.b5e77ae2c023bc86bb7e2b1879769fee.png

     

    As you can tell, the experience looks so much better, easier to navigate, and there is also our signature green button. The documentation for this change was a part of repl.run being depreciated which can be found here in the official Repl.it docs: https://docs.repl.it/repls/repl-run 

  11. I feel like its crappy of Nvidia to do this, but they do have a point. 

    Hardware unboxed is basing their performance numbers pretty much only off of Rasterization. The thing is, while the improvements they have made there are kinda eh, we're going into a new era of everything is ray traced, down to linus' ball hairs the particles of sand in MC. Technologies such as DLSS and RT allow for us to start looking into what games in 2021 and 2022 will be using, most of which will probably be focused on ray tracing. This only makes sense, as better lighting = moar better. 

    This was 100% a sleep deprived Piero ramble but I hope you get my point.

×