Login [Register]
Don't have an account? Register now to chat, post, use our tools, and much more.
Well as the topic implies I was wondering how feasible this would be. Well actually phero brought this up on IRC this morning and never posted a topic about it, so i figured I'd post it for him since i was interested in this as well. While the z80 is very slow, i think this would be more a a neat proof of concept than actually folding a lot of stuff at once. Tari was kind enough to say that normally folding requires a few MB for doing it. My idea was to split the folding data up either on a remote server and send it to the calculator, or, still have the main server split it up into little chunks and have it store it on the host computer and just have a program running on the the computer to send the data to the calculator to process. The main drawback(s) i see from this:
1) Batteries, might have to mod your calculator to run off ac
2) It would be more efficient to run it on the computer to fold, but it would be pretty cool to say you can fold from your 84+
Does anyone have any thoughts about this? How possible it would be, or how bad of an idea it was to post about this?
Honestly, I don't think it's a great idea. As my laptop takes about a day per protein while my PS3 does about five hours. I can't imagine what it'd be *on* a calculator; my Computer downloads the 5MB files, not sure which my PS3 downloads. It'd be neat if there was a application on the computer end that would intercept the progress (on that computer) and send that to gCn for "authorized" calculators to monitor.

Bonus Points for pausing, stopping and group features (changing groups, joining a group) as well as user statistics from your FAH page.
comicIDIOT wrote:
Honestly, I don't think it's a great idea. As my laptop takes about a day per protein while my PS3 does about five hours. I can't imagine what it'd be *on* a calculator; my Computer downloads the 5MB files, not sure which my PS3 downloads. It'd be neat if there was a application on the computer end that would intercept the progress (on that computer) and send that to gCn for "authorized" calculators to monitor.

Bonus Points for pausing, stopping and group features (changing groups, joining a group) as well as user statistics from your FAH page.

Well i with thinking having it split it up among multiple calculators. Cause having 1 calculator would take forever. I know even with a handful it wouldn't be super efficient, but i would be cool.
Forget Folding@Home -- use it to factor calculator public keys. That is the ultimate coolness -- using the calculators to factor their own keys.

Contrary to popular belief, there are still many keys to be factored -- in fact, as many as there are calculators. Each calculator has a unique "public key" -- TI must have a key vault somewhere of the private key component for every calculator they've ever sold, so that they can sign certificate updates with it. (That is how the calculator validates certificate updates (8XQ files).)

I agree it's not practical, but I'm all for this, and there actually is a concrete goal/end, unlike Folding@Home.

Thoughts?

EDIT: Also, building a circuit to run off AC is not hard. I was planning on doing it anyway.
BrandonW wrote:
Forget Folding@Home -- use it to factor calculator public keys. That is the ultimate coolness -- using the calculators to factor their own keys.

Contrary to popular belief, there are still many keys to be factored -- in fact, as many as there are calculators. Each calculator has a unique "public key" -- TI must have a key vault somewhere of the private key component for every calculator they've ever sold, so that they can sign certificate updates with it. (That is how the calculator validates certificate updates (8XQ files).)

I agree it's not practical, but I'm all for this, and there actually is a concrete goal/end, unlike Folding@Home.

Thoughts?

EDIT: Also, building a circuit to run off AC is not hard. I was planning on doing it anyway.
Wow, there are certificate updates? Wouldn't that be quite dangerous?
BrandonW wrote:
Forget Folding@Home -- use it to factor calculator public keys. That is the ultimate coolness -- using the calculators to factor their own keys.

Contrary to popular belief, there are still many keys to be factored -- in fact, as many as there are calculators. Each calculator has a unique "public key" -- TI must have a key vault somewhere of the private key component for every calculator they've ever sold, so that they can sign certificate updates with it. (That is how the calculator validates certificate updates (8XQ files).)

I agree it's not practical, but I'm all for this, and there actually is a concrete goal/end, unlike Folding@Home.

Thoughts?

EDIT: Also, building a circuit to run off AC is not hard. I was planning on doing it anyway.


Love it. That would be pretty cool to have it factor its own keys. I think that would be pretty cool to do. Tho, i still do like the folding at home, we could do both. But the keys have a way more coolness factor. The folding at home would be a way to get alot of attention to gcn because its very well know.
JosJuice wrote:
BrandonW wrote:
Forget Folding@Home -- use it to factor calculator public keys. That is the ultimate coolness -- using the calculators to factor their own keys.

Contrary to popular belief, there are still many keys to be factored -- in fact, as many as there are calculators. Each calculator has a unique "public key" -- TI must have a key vault somewhere of the private key component for every calculator they've ever sold, so that they can sign certificate updates with it. (That is how the calculator validates certificate updates (8XQ files).)

I agree it's not practical, but I'm all for this, and there actually is a concrete goal/end, unlike Folding@Home.

Thoughts?

EDIT: Also, building a circuit to run off AC is not hard. I was planning on doing it anyway.
Wow, there are certificate updates? Wouldn't that be quite dangerous?


It's perfectly safe -- the OS (and boot code) are very careful to receive and write such updates.

It's how you pay for shareware applications -- the application is signed with a key unique to that application (not 0104), and then they send you a special 8XK with the application's data and the 8XQ certificate update embedded into it. (You can also have separate 8XK and 8XQ files.)

The certificate update contains the public key to match the signature of your shareware application; once installed, it is placed in your certificate. (This is exactly how Free83P works -- it places the 0005 key in your certificate (except it writes it itself, since we can't sign our own updates and have the OS validate and write it itself).)

The entire update itself is signed with the private key to match your calculator public key. The OS and boot code only write the update if it can validate it using your unique public key.

We don't even have ONE calculator ID public/private key pair -- even pulling off one RSA factor would be a significant achievement.

EDIT: Personally, I've wanted for some time to create a public service to do exactly this kind of thing -- anonymously submit your calculator public key, and we'll factor it for you (in time), so that you can sign your own certificate updates.

Certificate updates are cool for lots of different reasons, not the least of which is that you can embed your own custom "About" screen text (above TI's URL), such as your own name. (We can do this now, but not officially signing it, which is much safer to do -- the dangerous nature of it is the reason why such a tool doesn't already exist publicly.)

By the way, people have had 8XQs for years -- it's how developers initially wrote Flash applications using the SDK. TI would sign and send you a certificate update unique to your calculator with your own developer key (not 0104) -- that way you could write and sign applications and put them on your real calculator (but only your calculator).

EDIT2: In case you never noticed, this is the reason why TI asks you to register your calc ID on their web site (and always has) -- they need it in order to sign your shareware/paid application for you to download.

EDIT3: And, if you don't want to hook your calculator up to AC, you can always run emulators, which could be even better because you can speed them up to insane levels (like PindurTI).

We'd also probably want to encrypt blocks and/or keep part of the calculator side a secret to prevent tampering by people (injecting bad data, etc.).
BrandonW wrote:
JosJuice wrote:
BrandonW wrote:
Forget Folding@Home -- use it to factor calculator public keys. That is the ultimate coolness -- using the calculators to factor their own keys.

Contrary to popular belief, there are still many keys to be factored -- in fact, as many as there are calculators. Each calculator has a unique "public key" -- TI must have a key vault somewhere of the private key component for every calculator they've ever sold, so that they can sign certificate updates with it. (That is how the calculator validates certificate updates (8XQ files).)

I agree it's not practical, but I'm all for this, and there actually is a concrete goal/end, unlike Folding@Home.

Thoughts?

EDIT: Also, building a circuit to run off AC is not hard. I was planning on doing it anyway.
Wow, there are certificate updates? Wouldn't that be quite dangerous?


It's perfectly safe -- the OS (and boot code) are very careful to receive and write such updates.

It's how you pay for shareware applications -- the application is signed with a key unique to that application (not 0104), and then they send you a special 8XK with the application's data and the 8XQ certificate update embedded into it. (You can also have separate 8XK and 8XQ files.)

The certificate update contains the public key to match the signature of your shareware application; once installed, it is placed in your certificate. (This is exactly how Free83P works -- it places the 0005 key in your certificate (except it writes it itself, since we can't sign our own updates and have the OS validate and write it itself).)

The entire update itself is signed with the private key to match your calculator public key. The OS and boot code only write the update if it can validate it using your unique public key.

We don't even have ONE calculator ID public/private key pair -- even pulling off one RSA factor would be a significant achievement.

EDIT: Personally, I've wanted for some time to create a public service to do exactly this kind of thing -- anonymously submit your calculator public key, and we'll factor it for you (in time), so that you can sign your own certificate updates.

Certificate updates are cool for lots of different reasons, not the least of which is that you can embed your own custom "About" screen text (above TI's URL), such as your own name. (We can do this now, but not officially signing it, which is much safer to do -- the dangerous nature of it is the reason why such a tool doesn't already exist publicly.)

By the way, people have had 8XQs for years -- it's how developers initially wrote Flash applications using the SDK. TI would sign and send you a certificate update unique to your calculator with your own developer key (not 0104) -- that way you could write and sign applications and put them on your real calculator (but only your calculator).

EDIT2: In case you never noticed, this is the reason why TI asks you to register your calc ID on their web site (and always has) -- they need it in order to sign your shareware/paid application for you to download.

EDIT3: And, if you don't want to hook your calculator up to AC, you can always run emulators, which could be even better because you can speed them up to insane levels (like PindurTI).

We'd also probably want to encrypt blocks and/or keep part of the calculator side a secret to prevent tampering by people (injecting bad data, etc.).

I never knew all that BrandonW. Just wow, you've thought this out pretty good. I like this idea. Getting the keys by doing this from your own calc would be pretty awesome. I think we should go with this. Would be neat concept.
I definitely like the meta-ness and the coolness of using calculators to factor their own key, and I think CALCnet / globalCALCnet would be the simplest way to implement this. However, I'm concerned about the legality aspects. I know that TI never pursued the lawsuits against you and the others, but I still am paranoid about it. Thoughts?

Oh, and powering calculators from AC: an easier option is from nice regulated USB power. Cf. DShiznit's topic, for example.
http://www.cemetech.net/forum/viewtopic.php?t=5638
KermMartian wrote:
I definitely like the meta-ness and the coolness of using calculators to factor their own key, and I think CALCnet / globalCALCnet would be the simplest way to implement this. However, I'm concerned about the legality aspects. I know that TI never pursued the lawsuits against you and the others, but I still am paranoid about it. Thoughts?

Oh, and powering calculators from AC: an easier option is from nice regulated USB power. Cf. DShiznit's topic, for example.
http://www.cemetech.net/forum/viewtopic.php?t=5638


TI will not pursue this. There's nothing to pursue. We are simply factoring numbers for our own amusement.

The only argument is why we are factoring the number. TI abandoned the shareware application concept long ago as soon as they released the 0104 private key to create our own freeware applications. As soon as they did that, they made it very easy for anyone to get their hands on a shareware application (from anyone at all who has it), re-sign it using the 0104 key, and then send it to anyone. You'd be hard-pressed to prove how TI is being damaged by factoring a number that is unique to only ONE calculator and cannot be used to obtain new applications or OSes.

EDIT: Whether you agree with the factor-on-a-calculator component of this or not, it's a worthwhile thing to do. If you show to the world that even 6MHz calculators are factoring RSA public keys, it will improve the public's perception of it. Remember, we're guilty ourselves of thinking it wasn't practical -- and then we showed the world that 512-bit RSA CAN be broken and that it's practical to do so.

We could even have a PC program that this talks to, and both PC and calculator can contribute to the factoring.

Not that curing cancer and all isn't a worthwhile goal, but this is something you can contribute to knowing you're going to be around to see the end of it.
This is why BrandonW has a medal next to his name =)
elfprince13 wrote:
This is why BrandonW has a medal next to his name =)

Exactly. Also if TI isnt going to purse anything i think we should find a way to go along with it. It would be really nice to factor all those keys.
BrandonW wrote:
We could even have a PC program that this talks to, and both PC and calculator can contribute to the factoring.
I agree this would be awesome, this is what i was thinking with the folding at home client, but more or less having the computer program just split up the data for the calc and store it till its ready to send it. I think we should do both, but brandons idea first as its a great idea.
I agree it would be cool and definitely something worthwhile if TI wouldn't get angry about it. I guess the question is where would we go from here? We'd need a server to connect to the metahub and coordinate workload distribution, and of course the calculator-side client.
That sounds about right to me =)
KermMartian wrote:
I agree it would be cool and definitely something worthwhile if TI wouldn't get angry about it. I guess the question is where would we go from here? We'd need a server to connect to the metahub and coordinate workload distribution, and of course the calculator-side client.


Research into GGNFS and how in the world to write a calculator-side client. It's all moot if we can't figure that part out.
BrandonW wrote:
Research into GGNFS and how in the world to write a calculator-side client. It's all moot if we can't figure that part out.

Cabamap seems like a good place to start.
KermMartian wrote:
I agree it would be cool and definitely something worthwhile if TI wouldn't get angry about it. I guess the question is where would we go from here? We'd need a server to connect to the metahub and coordinate workload distribution, and of course the calculator-side client.


Well If you need a server i can set-up whatever you need me to run on mine. Ill gladly offer some server space for this and any other gcn hubs.
Thanks rcfreak0! BrandonW, I'll have to read more about GGNFS (here's the one page intro for the rest of you). I'm a little vague on what the factoring process actually entails.
The original United TI thread about key factoring (if it still exists) is also a good reference, as it has the original discussion about applying GGNFS and msieve in a distributed format.

There are several steps, the first major one being polynomial selection, which is done by one person. Then the actual sieving can be done by all of us until enough relations are obtained, and then one of us takes that information (many MB or GB pushed to a central location) and performs the final step on it to produce the factors.

That's majorly oversimplified, but the thread (and the page you linked to) goes into more detail.
Well for reference, here are the 3 threads i could find on UTI about factoring keys. I think these where the ones you meant, BrandonW, otherwise its possible i missed another one.
I do believe this was the original one here.
The second one here.
The third one here.
The third one is more nspire related, but it still has factoring in it.
BrandonW wrote:
There are several steps, the first major one being polynomial selection, which is done by one person. Then the actual sieving can be done by all of us until enough relations are obtained, and then one of us takes that information (many MB or GB pushed to a central location) and performs the final step on it to produce the factors.
Space wise i can host a lot of data for this if i have too, thankfully shouldn't be a problem at all.
  
Register to Join the Conversation
Have your own thoughts to add to this or any other topic? Want to ask a question, offer a suggestion, share your own programs and projects, upload a file to the file archives, get help with calculator and computer programming, or simply chat with like-minded coders and tech and calculator enthusiasts via the site-wide AJAX SAX widget? Registration for a free Cemetech account only takes a minute.

» Go to Registration page
Page 1 of 2
» All times are GMT - 5 Hours
 
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

 

Advertisement