FOSS, learning curve, and niche field? I got you. VLSI layout software or SDR (software defined radio) software.
VLSI layout is kind of software allows you to design electronics hardware at the silicone/transistor level. IOW you have to draw the layers of materials that form transistors (cross poly silicon and p/n-diffused silicon to form) and connect them to metal layers. Learn how NAND and NOR gates are actually formed. Make a 1 bit data storage (flip flop), half adder, full adder. Go as far as you want.
SDR. Less intense, but also niche and gives you lots of related things to learn/branch in to.
Can be used to receive and decode radio/satellite transmissions. Normally people would buy hardware and use software locally, but some people make their receivers accessible on the internet (websdr) which can give you a first intro to it. After you capture data you can run it through other software to decode (depending on what it is). For example, you can capture raw weather image data from satellites and decode it to get the images that the satellite is transmitting as it's capturing the data.
A list of some SDRs you can access over the internet (doesn't require hardware purchase to use): http://websdr.org/
It probably won't be immediately obvious how to use either SDR or VLSI software but you asked for niche and learning curve, so enjoy. ;)
A regular expression (a sort of language for pattern matching) that matches users. IOW there's patterns to filter users, not just communities and instances.
The video that I watched? It was from How to Drink, which iirc he said he decided to try and make his video after watching a video that talked about speeding up extracts via ultrasonic cleaner, but I don't recall what he said what was the video he watched.
A few months ago I saw a video of someone who used an ultrasonic cleaner to make extracts, so i tried it. I did 30 - 60 mins for each of orange peel, banana peel, pepper corn, anise, vanilla bean, cinnamon, and coffee. The coffee smelled absolutely disgusting, but this was like an hour rather than 3 mins. I had no idea how much time was appropriate. I also used vodka instead of water. Idk why I thought the coffee would work. The rest seem roughly what I'd expect. I've done traditional extracts previously. It makes sense to try water for coffee. I feel dumb for not trying that obvious combination.
That mindset is part of why I bought a Prusa originally, but felt betrayed by the MMU2S. That was an expensive and crap product that they never fixed. They basically stole from their customers with that one and I wish the community wouldn't let them get away with it without them making it right, which at this point would be difficult to do.
I don't know current best options, but last I was looking, the Voron design was where I was leaning. I'd suggest avoiding companies that violate open source licenses.
I used to be a Prusa evangelist, but I don't forgive them for the MMU2S. Huge amount of people never could get it to work. I'm curious as to the actual percentage working. I couldn't ever get it to work even with multiple complete rebuilds. That product should have been recalled or a free fix offered. Years later they released a new version that supposedly fixed the issues and you still had to pay to upgrade to it. Discounted if they had a support record for you. If not, full price.
From time to time I see a new Prusa printer and think "that looks good, I should buy it", and then remember the insane amount of time I wasted trying to get the MMU2S to work. Never buying Prusa again until they provide me the working MMU I paid for at no additional cost. As there's basically zero chance of that, my boycott continues and I suggest not supporting a company that fucked over their customers with an expensive poorly designed product that they then want them to pay to fix.
I think the numbers are worse. The 43 includes those with tumors that shrank or disappeared. In otherwords: 43 (shrank and disappeared) = 28 (shrank) + 15 (disappeared).
So what happened to the other 102 - 43 = 59? Still it's roughly 40% chance of improvement when other treatment methods have failed, which I think a lot of people would be willing to take. I'm often skeptical about early results though.
I have been trying to come up with a solution that retains anonymity while also preventing spam/scam/etc for a while now. I have a few ideas, but feel there's room for improvements.Im'm up for discussing related ideas.
Is it resolving to an IPv6 address? If so and you don't have fully working IPv6, try disabling IPv6 on Debian. I'm mentioning this as a possibility because I've personally had this issue with some new Debian installs a few times. The first few it took me more time than I'd like to admit to figure out the issue. If this is the case, then an IPv4 only mirror would avoid it, but you may have other random issues.
If it's not related to IPv6, then I'd just pick different mirrors until you find one that works. Are you able to access that fastly mirror from another device (via http)?
While the POC requires su, the underlying flaw potentially works on any setuid binary on systems with AF_ALG enabled (provided there isn't something else preventing it).
Sounds like the start of a project: change display control or automatically reset the existing time via a SBC that gets time via your computer or NTP directly.
Or worse idea -- point a webcam at the physical clock, OCR to get time and update your computer clock to the wrong time. ;)
Please actually compare the certificate when connecting to your server directly (bypassing Cloudflare) and connecting via Cloudflare. An easy way to do this is with openssl CLI:
Replace your-domain-here.org with your domain and your-ip-here with your actual server IP, but also do it with the Cloudflare IP.
The section about the "Full (strict)" / "Full" is referring to how Cloudflare verifies the certificate (or not in the case of Flexible and off) between your origin server and Cloudflare -- this is not with respect to the client and Cloudflare. The Custom origin certificates are also with respect to Cloudflare and your server (has no impact on certificate used between the client and Cloudflare). Cloudflare still uses a separate certificate that they have issued to themselves and hold the private key to use for the client.
If you pay extra for their "Advanced Certificate Manager", this allows you to upload a custom certificate to be used between the client and Cloudflare, but you have to provide the private key to Cloudflare because they still terminate SSL/TLS at their servers. Even their "Total TLS" service (part of ACM and the word "Total" could be mistaken to be "total" as in from client all the way to your origin server) does not provide E2EE.
I may be unaware of a newer service offering, but the only way that I'm aware of to get true E2EE is on their Enterprise plan (Keyless TLS). I have a lot of experience with Cloudflare for both personal and Enterprise plan (I was the technical person in charge of the account and configuring and such). Granted, I've not been dealing with CF enterprise for a few years now and they may have a new service offering outside of enterprise that I'm not familiar with, but my quick look around still looks like everything aside from Keyless TLS requires either giving them the key (in the case of ACM custom certificates) or they use their own certificate for client <-> Cloudflare. When I did manage the enterprise plan, we actually didn't use Keyless TLS because we used features that required them to terminate TLS anyway, so I can't speak to the specifics of it.
I hope I'm wrong though. I'd love to have true E2EE while still getting the DDoS protection on my personal stuff.
You should check the certificate shown to clients when accessing your domain. I think you'll find that it is not the certificate that you created outside of Cloudflare. Cloudflare doesn't need your private key as they issue a certificate for your domain to themselves and use that for the connection with the client. The certificate you created is used between Cloudflare and your server. The only option I'm aware to route traffic through Cloudflare where they don't terminate SSL is an enterprise only feature.
I'm in very early stages of a similar project (platform fighter) and a similar issue (I can't do everything and was having trouble finding reliable people). I don't know if our project goals are similar enough to warrant working together, but I think it may be worth talking about a possible collaboration. Perhaps even just to make the game multi platform (I'm targeting a retro game console) given that we'll both need a lot of the same things even if the code itself has a lot of differences. Art, sound, music, story/text, but even things like defining character lists, abilities, and game balance related things is important and duplicative.
By very early stages I mean I don't have any game logic written yet. I'm targeting retro game console hardware and so far I've mostly been writing code (primarily C) to test my understanding of how the hardware functions/limitations (already found some bugs between emulator and real hardware that impacts some home brew games from other developers), and then writing functions that will become a game building library (I don't know that it's right to call it an "engine"). Granted, I'm making a lot of assumptions at this point about what I'll need in terms of features, but also in terms of how much system resources are safe to allocate to different pieces, so when I get things a little more understood and have some core library functions I'm happy with, I'll start writing game logic to see what more I need / what changes I need to make.
I've not worked on it for a few months as I've been busy with contract work that I was just informed this week is ending prematurely due to budgetary changes. As such, I expect to have time to pick it up again starting next week.
I think that depends what you consider a bug. I thinj Mario not being present makes sense as he's dead. The princess was saved even though Mario died, so I think it's reasonable to say the win condition is met. What I think is definitely a bug is that you can hear the death sound way late (a 2nd time) when you do this (jump and hit the axe at the same time bowser touches you).
Could you explain more? Is this just an experiment to see if you can line up and fuse 2 separately printed objects? Are the 2 parts different materials? I feel like I'm misunderstanding.
What I think you've done is print 1 object in TPU and then print a 2nd object, also in TPU, close enough to the 1st object such they fuse. Maybe your future plans would help me understand. I'm interested in learning about different techniques.
I had considered doing something like object fusing to create foldable objects, like print the first couple layers in TPU (for both objects as well as a connecting piece between them) and then print 2 separate objects on top of the TPU base -- think like a foldable phone case where rather than use a normal hinge, it would be an edge in TPU and the rest is PLA/PETG/whatever. Reason to do the whole base in TPU is that I thought just printing the part that connects the other 2 parts in TPU wouldn't fuse well enough and would separate with use. I've not actually done this.
FOSS, learning curve, and niche field? I got you. VLSI layout software or SDR (software defined radio) software.
VLSI layout is kind of software allows you to design electronics hardware at the silicone/transistor level. IOW you have to draw the layers of materials that form transistors (cross poly silicon and p/n-diffused silicon to form) and connect them to metal layers. Learn how NAND and NOR gates are actually formed. Make a 1 bit data storage (flip flop), half adder, full adder. Go as far as you want.
Example software:
SDR. Less intense, but also niche and gives you lots of related things to learn/branch in to.
Can be used to receive and decode radio/satellite transmissions. Normally people would buy hardware and use software locally, but some people make their receivers accessible on the internet (websdr) which can give you a first intro to it. After you capture data you can run it through other software to decode (depending on what it is). For example, you can capture raw weather image data from satellites and decode it to get the images that the satellite is transmitting as it's capturing the data.
A list of some SDRs you can access over the internet (doesn't require hardware purchase to use): http://websdr.org/
It probably won't be immediately obvious how to use either SDR or VLSI software but you asked for niche and learning curve, so enjoy. ;)
EDIT: Fixed some typos and grammar.
A regular expression (a sort of language for pattern matching) that matches users. IOW there's patterns to filter users, not just communities and instances.
The video that I watched? It was from How to Drink, which iirc he said he decided to try and make his video after watching a video that talked about speeding up extracts via ultrasonic cleaner, but I don't recall what he said what was the video he watched.
Can you be sure it wasn't the prior train arriving and leaving so late that it appeared to be leaving early?
A few months ago I saw a video of someone who used an ultrasonic cleaner to make extracts, so i tried it. I did 30 - 60 mins for each of orange peel, banana peel, pepper corn, anise, vanilla bean, cinnamon, and coffee. The coffee smelled absolutely disgusting, but this was like an hour rather than 3 mins. I had no idea how much time was appropriate. I also used vodka instead of water. Idk why I thought the coffee would work. The rest seem roughly what I'd expect. I've done traditional extracts previously. It makes sense to try water for coffee. I feel dumb for not trying that obvious combination.
That mindset is part of why I bought a Prusa originally, but felt betrayed by the MMU2S. That was an expensive and crap product that they never fixed. They basically stole from their customers with that one and I wish the community wouldn't let them get away with it without them making it right, which at this point would be difficult to do.
I don't know current best options, but last I was looking, the Voron design was where I was leaning. I'd suggest avoiding companies that violate open source licenses.
I used to be a Prusa evangelist, but I don't forgive them for the MMU2S. Huge amount of people never could get it to work. I'm curious as to the actual percentage working. I couldn't ever get it to work even with multiple complete rebuilds. That product should have been recalled or a free fix offered. Years later they released a new version that supposedly fixed the issues and you still had to pay to upgrade to it. Discounted if they had a support record for you. If not, full price.
From time to time I see a new Prusa printer and think "that looks good, I should buy it", and then remember the insane amount of time I wasted trying to get the MMU2S to work. Never buying Prusa again until they provide me the working MMU I paid for at no additional cost. As there's basically zero chance of that, my boycott continues and I suggest not supporting a company that fucked over their customers with an expensive poorly designed product that they then want them to pay to fix.
I think the numbers are worse. The 43 includes those with tumors that shrank or disappeared. In otherwords: 43 (shrank and disappeared) = 28 (shrank) + 15 (disappeared).
So what happened to the other 102 - 43 = 59? Still it's roughly 40% chance of improvement when other treatment methods have failed, which I think a lot of people would be willing to take. I'm often skeptical about early results though.
I have been trying to come up with a solution that retains anonymity while also preventing spam/scam/etc for a while now. I have a few ideas, but feel there's room for improvements.Im'm up for discussing related ideas.
Is it resolving to an IPv6 address? If so and you don't have fully working IPv6, try disabling IPv6 on Debian. I'm mentioning this as a possibility because I've personally had this issue with some new Debian installs a few times. The first few it took me more time than I'd like to admit to figure out the issue. If this is the case, then an IPv4 only mirror would avoid it, but you may have other random issues.
If it's not related to IPv6, then I'd just pick different mirrors until you find one that works. Are you able to access that fastly mirror from another device (via http)?
Doesn't look that far off from a lot of Unix socks posts with the color stripes between the user and multiple monitors.
While the POC requires
su, the underlying flaw potentially works on any setuid binary on systems withAF_ALGenabled (provided there isn't something else preventing it).Sounds like the start of a project: change display control or automatically reset the existing time via a SBC that gets time via your computer or NTP directly.
Or worse idea -- point a webcam at the physical clock, OCR to get time and update your computer clock to the wrong time. ;)
Your ummm... Your clocks are not synchronized.
Please actually compare the certificate when connecting to your server directly (bypassing Cloudflare) and connecting via Cloudflare. An easy way to do this is with openssl CLI:
Replace
your-domain-here.orgwith your domain andyour-ip-herewith your actual server IP, but also do it with the Cloudflare IP.The section about the "Full (strict)" / "Full" is referring to how Cloudflare verifies the certificate (or not in the case of Flexible and off) between your origin server and Cloudflare -- this is not with respect to the client and Cloudflare. The Custom origin certificates are also with respect to Cloudflare and your server (has no impact on certificate used between the client and Cloudflare). Cloudflare still uses a separate certificate that they have issued to themselves and hold the private key to use for the client.
If you pay extra for their "Advanced Certificate Manager", this allows you to upload a custom certificate to be used between the client and Cloudflare, but you have to provide the private key to Cloudflare because they still terminate SSL/TLS at their servers. Even their "Total TLS" service (part of ACM and the word "Total" could be mistaken to be "total" as in from client all the way to your origin server) does not provide E2EE.
I may be unaware of a newer service offering, but the only way that I'm aware of to get true E2EE is on their Enterprise plan (Keyless TLS). I have a lot of experience with Cloudflare for both personal and Enterprise plan (I was the technical person in charge of the account and configuring and such). Granted, I've not been dealing with CF enterprise for a few years now and they may have a new service offering outside of enterprise that I'm not familiar with, but my quick look around still looks like everything aside from Keyless TLS requires either giving them the key (in the case of ACM custom certificates) or they use their own certificate for client <-> Cloudflare. When I did manage the enterprise plan, we actually didn't use Keyless TLS because we used features that required them to terminate TLS anyway, so I can't speak to the specifics of it.
I hope I'm wrong though. I'd love to have true E2EE while still getting the DDoS protection on my personal stuff.
You should check the certificate shown to clients when accessing your domain. I think you'll find that it is not the certificate that you created outside of Cloudflare. Cloudflare doesn't need your private key as they issue a certificate for your domain to themselves and use that for the connection with the client. The certificate you created is used between Cloudflare and your server. The only option I'm aware to route traffic through Cloudflare where they don't terminate SSL is an enterprise only feature.
I'm in very early stages of a similar project (platform fighter) and a similar issue (I can't do everything and was having trouble finding reliable people). I don't know if our project goals are similar enough to warrant working together, but I think it may be worth talking about a possible collaboration. Perhaps even just to make the game multi platform (I'm targeting a retro game console) given that we'll both need a lot of the same things even if the code itself has a lot of differences. Art, sound, music, story/text, but even things like defining character lists, abilities, and game balance related things is important and duplicative.
By very early stages I mean I don't have any game logic written yet. I'm targeting retro game console hardware and so far I've mostly been writing code (primarily C) to test my understanding of how the hardware functions/limitations (already found some bugs between emulator and real hardware that impacts some home brew games from other developers), and then writing functions that will become a game building library (I don't know that it's right to call it an "engine"). Granted, I'm making a lot of assumptions at this point about what I'll need in terms of features, but also in terms of how much system resources are safe to allocate to different pieces, so when I get things a little more understood and have some core library functions I'm happy with, I'll start writing game logic to see what more I need / what changes I need to make.
I've not worked on it for a few months as I've been busy with contract work that I was just informed this week is ending prematurely due to budgetary changes. As such, I expect to have time to pick it up again starting next week.
I think that depends what you consider a bug. I thinj Mario not being present makes sense as he's dead. The princess was saved even though Mario died, so I think it's reasonable to say the win condition is met. What I think is definitely a bug is that you can hear the death sound way late (a 2nd time) when you do this (jump and hit the axe at the same time bowser touches you).
Could you explain more? Is this just an experiment to see if you can line up and fuse 2 separately printed objects? Are the 2 parts different materials? I feel like I'm misunderstanding.
What I think you've done is print 1 object in TPU and then print a 2nd object, also in TPU, close enough to the 1st object such they fuse. Maybe your future plans would help me understand. I'm interested in learning about different techniques.
I had considered doing something like object fusing to create foldable objects, like print the first couple layers in TPU (for both objects as well as a connecting piece between them) and then print 2 separate objects on top of the TPU base -- think like a foldable phone case where rather than use a normal hinge, it would be an edge in TPU and the rest is PLA/PETG/whatever. Reason to do the whole base in TPU is that I thought just printing the part that connects the other 2 parts in TPU wouldn't fuse well enough and would separate with use. I've not actually done this.