RackNerd's OFFICIAL LET BLACK FRIDAY! HUNDREDS OF GIVEAWAYS + CRAZY DEALS, MASSIVE (come see)

MrEd · November 2024

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

TrK · November 2024

@noob404 said:

@TrK said:

@MrEd said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

Tha "all" for me is in the chart at the bottom Load the data and hover on any of the columns

Doing it again? Let's go.......

That gave me the idea to use Google Chart APIs for another project.

Show us when finished. Would love to see it.

noob404 · November 2024

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

noob404 · November 2024

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

TrK · November 2024

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

noob404 · November 2024

@TrK said:

@noob404 said:

@TrK said:

@MrEd said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

Tha "all" for me is in the chart at the bottom Load the data and hover on any of the columns

Doing it again? Let's go.......

That gave me the idea to use Google Chart APIs for another project.

Show us when finished. Would love to see it.

Sure. This was for another project that I launched last year. It didn't catch up as I had expected. This year, I have another one. Hope this one stays. Even this one uses Charts API for simple comparison.

TrK · November 2024

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

Get another fiber connection a 50mbps one would work.

MrEd · November 2024

@TrK said:

@MrEd said:

@codelock said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

he still has all the data to train LLM personas

Oh yeah, I still make the DB backups before starting a fresh one I am a hoarder

Most of us are, how many TBs and counting?

Everything fits on the 20GB of the VM

noob404 · November 2024

@TrK said:

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

True. I have found various GitHub repos that have a list of bad IPs in separate text files according to their score. Maybe it'd be a good idea to cross-check with them before buying from these providers.

MrEd · November 2024

@MrEd said:

@TrK said:

@MrEd said:

@codelock said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

he still has all the data to train LLM personas

Oh yeah, I still make the DB backups before starting a fresh one I am a hoarder

Most of us are, how many TBs and counting?

Everything fits on the 20GB of the VM

I mean, the stats DBs...

TrK · November 2024

@MrEd said:

@TrK said:

@MrEd said:

@codelock said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

he still has all the data to train LLM personas

Oh yeah, I still make the DB backups before starting a fresh one I am a hoarder

Most of us are, how many TBs and counting?

Everything fits on the 20GB of the VM

How? Only DB? What about backup of the backup of the backup?

noob404 · November 2024

@TrK said:

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

Get another fiber connection a 50mbps one would work.

That might as well. But, thinking about it now, chances of his IP getting banned is pretty low, because it only refreshes every 5 minutes. We guys do more refreshes than that.

noob404 · November 2024

@TrK said:

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

Get another fiber connection a 50mbps one would work.

BTW, what's your internet connection speed like? sh has a 500mbps connection.

TrK · November 2024

@noob404 said:

@TrK said:

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

True. I have found various GitHub repos that have a list of bad IPs in separate text files according to their score. Maybe it'd be a good idea to cross-check with them before buying from these providers.

Yup that's the entire problem. I might really need a static IP, my ISP is getting flagged constantly due to nat

noob404 · November 2024

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

True. I have found various GitHub repos that have a list of bad IPs in separate text files according to their score. Maybe it'd be a good idea to cross-check with them before buying from these providers.

Yup that's the entire problem. I might really need a static IP, my ISP is getting flagged constantly due to nat

JioFiber?

MrEd · November 2024

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

My IP is stati, but... I have access to more than one residential IP. Also, for the purpose of such threads I could easily incorporate some mobile data Its not that expensive here

TrK · November 2024

@noob404 said:

@TrK said:

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

Get another fiber connection a 50mbps one would work.

BTW, what's your internet connection speed like? sh has a 500mbps connection.

200mbps and it's going great, might upgrade or might get another one the home lab is getting hungrier by day.

TrK · November 2024

@noob404 said:

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

True. I have found various GitHub repos that have a list of bad IPs in separate text files according to their score. Maybe it'd be a good idea to cross-check with them before buying from these providers.

Yup that's the entire problem. I might really need a static IP, my ISP is getting flagged constantly due to nat

JioFiber?

Hate it, currently with an independent one.

noob404 · November 2024

@MrEd said:

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

My IP is stati, but... I have access to more than one residential IP. Also, for the purpose of such threads I could easily incorporate some mobile data Its not that expensive here

Oh I see. Mobile data can't be as inexpensive as India though. We have them dirt cheap.

MrEd · November 2024

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@codelock said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

he still has all the data to train LLM personas

Oh yeah, I still make the DB backups before starting a fresh one I am a hoarder

Most of us are, how many TBs and counting?

Everything fits on the 20GB of the VM

How? Only DB? What about backup of the backup of the backup?

The threads are still acessable, I could just restart the crawler if needed

noob404 · November 2024

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

Get another fiber connection a 50mbps one would work.

BTW, what's your internet connection speed like? sh has a 500mbps connection.

200mbps and it's going great, might upgrade or might get another one the home lab is getting hungrier by day.

Oh, that's cool. I am still on a 25Mbps connection. Forgot to upgrade to atleast a 100 before the thread. Now, I am stuck with this for another month.

noob404 · November 2024

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

True. I have found various GitHub repos that have a list of bad IPs in separate text files according to their score. Maybe it'd be a good idea to cross-check with them before buying from these providers.

Yup that's the entire problem. I might really need a static IP, my ISP is getting flagged constantly due to nat

JioFiber?

Hate it, currently with an independent one.

Yah, heard really bad complaints, esp. from devs. We tried scraping data from a GitHub page once using JioFiber and it couldn't. Do you know why? Cause they blocked GitHub!!! The issue was resolved, but the audacity to be blocking something as elementary as GH!

noob404 · November 2024

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@codelock said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

he still has all the data to train LLM personas

Oh yeah, I still make the DB backups before starting a fresh one I am a hoarder

Most of us are, how many TBs and counting?

Everything fits on the 20GB of the VM

How? Only DB? What about backup of the backup of the backup?

The threads are still acessable, I could just restart the crawler if needed

Most of the older threads are now closed.

noob404 · November 2024

@noob404 said:

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

True. I have found various GitHub repos that have a list of bad IPs in separate text files according to their score. Maybe it'd be a good idea to cross-check with them before buying from these providers.

Yup that's the entire problem. I might really need a static IP, my ISP is getting flagged constantly due to nat

JioFiber?

Hate it, currently with an independent one.

Yah, heard really bad complaints, esp. from devs. We tried scraping data from a GitHub page once using JioFiber and it couldn't. Do you know why? Cause they blocked GitHub!!! The issue was resolved, but the audacity to be blocking something as elementary as GH!

But, their appeal lies in the coupled plans (TV+Internet+OTT).

MrEd · November 2024

@noob404 said:

@MrEd said:

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

My IP is stati, but... I have access to more than one residential IP. Also, for the purpose of such threads I could easily incorporate some mobile data Its not that expensive here

Oh I see. Mobile data can't be as inexpensive as India though. We have them dirt cheap.

A SIM card with 2GB for one month cost 2Eur. Since I am only using html, I guess, that would be more than enough

noob404 · November 2024

PARTY-PEOPLE-OF-THE-RACKNERD-COMMUNITY

Who wants a GA? Following Dustin's pattern, he might join us in an hour or two. Who wants to create some demands for GAs before he joins?

MrEd · November 2024

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@codelock said:

@TrK said:

@codelock said:

@MrEd said:

@noob404 said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@MrEd said:

@Frobsy said:

@noob404 said:

@Frobsy said:

@Kevinf100 said:

@Frobsy said:

@DaDeveloper said: Just for kicks I hope he stops at 1024 or 1337

Lol check this page
https://racknerd.projecthive.eu/ny2024/

11k? Jeez, guess makes sense for how long it runs for. Guess I under estimate time

They started posting quotes at some point that's probably what made the comments reach that high haha.

Yup, last few threads went crazy, pretty quick. 13K comments! Never again.

What do you think of this year's thread? Do you think you can pull 5k comments?

He is 20% there, and its only a week passed My prediction is he will get the 5k

Oh yeah ahhaah, So, I guess the 2nd guy will be like 3k and the 3rd will be very close as well. btw, can you change the similarity to 70% real quick? I wanna see something

No, I cannot. I would have to reindex everything, because similarity is stored in the DB. What would you want to check?

Oh, that's a bummer. I'm really interested in data and curious about how things work on the back end. I hope Dustin has access to a more advanced system, as it would be great to see how the stats are managed and interpreted. I'm sure there's a lot of potential for insightful analysis!

I have already said multiple times. Your message text (removig quotes, images, youtube links) is compared to other messages of yours and Levenshtein distance is calculated (how many symbols need to be changed to get the new message). If the distance is less than 3 or 90% of messages match, it is treated as similar.

Do you know what's a good idea? Adding an about page on the Stats website. Over the three threads that I have participated in, this has been a recurrent question. And, if you prolly were to check, "Levenshtein distance" is probably the keyword that would have given you the most similar comments . Just kidding!
But, seriously, just an about page with a brief idea on how it is calculated and then, you can just link everyone who asks to that page. If they still don't understand, add a link to Google on the about page.

Maybe that is a good idea, but on the other hand, answering this question increases my comment count by 1

you could have it increase by 1.69 if you wanted to

There was a time MrEd had all the comments to himself.

he still has all the data to train LLM personas

Oh yeah, I still make the DB backups before starting a fresh one I am a hoarder

Most of us are, how many TBs and counting?

Everything fits on the 20GB of the VM

How? Only DB? What about backup of the backup of the backup?

The threads are still acessable, I could just restart the crawler if needed

Most of the older threads are now closed.

Closed, but accessible

noob404 · November 2024

@MrEd said:

@noob404 said:

@MrEd said:

@noob404 said:

@MrEd said:

@TrK said:

@MrEd said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

I don't care about my home BW. I have 1Gbps FTTH

IP rate limiting might be introduced just for your home ip.

I would still have workarounds for that

Let me guess - Dynamic IP. Or, are you with one of those Double NAT ISPs?

My IP is stati, but... I have access to more than one residential IP. Also, for the purpose of such threads I could easily incorporate some mobile data Its not that expensive here

Oh I see. Mobile data can't be as inexpensive as India though. We have them dirt cheap.

A SIM card with 2GB for one month cost 2Eur. Since I am only using html, I guess, that would be more than enough

In India, around 5.45 EUR for 3 months of free calling, SMS + 6GB 5G Data

TrK · November 2024

@noob404 said:

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@TrK said:

@noob404 said:

@noob404 said:

@MrEd said:

@noob404 said:

@TrK said:

@MrEd said:

@noob404 said:
Hey @MrEd, does your stats page keep recording data till the thread is closed, or is it till the end of the GA period?

Its tracking till I stop it, and that usually happens when new thread is created There is no automation in place for stopping

Set it to the announcement date and conserve your home bandwidth.

It's hosted on a RackNerd VPS.

Stats page is hosted at RN, but LET is preventing DCs to read, so traffic to LET goes through my home network

Makes sense. Wish some of those residential IP providers conducted a GA as well, so, you could route it through their IPs.

Last year, a round BF, there were some providers, mostly new ones providing residential IPs for pretty cheap prices. Not sure if they are still in business.

Residential proxies are cheap when they are already flagged, clean ones are really costly.

True. I have found various GitHub repos that have a list of bad IPs in separate text files according to their score. Maybe it'd be a good idea to cross-check with them before buying from these providers.

Yup that's the entire problem. I might really need a static IP, my ISP is getting flagged constantly due to nat

JioFiber?

Hate it, currently with an independent one.

Yah, heard really bad complaints, esp. from devs. We tried scraping data from a GitHub page once using JioFiber and it couldn't. Do you know why? Cause they blocked GitHub!!! The issue was resolved, but the audacity to be blocking something as elementary as GH!

Their peering is bad sometimes they just have CF disconnected from jio fiber.

MrEd · November 2024

Ok, children are sleeping, time to get back off the phone... Will get back in several hours Have a great time

Howdy, Stranger!

Categories

In this Discussion

RackNerd's OFFICIAL LET BLACK FRIDAY! HUNDREDS OF GIVEAWAYS + CRAZY DEALS, MASSIVE (come see)

Comments

PARTY-PEOPLE-OF-THE-RACKNERD-COMMUNITY

Howdy, Stranger!

Quick Links

Categories

In this Discussion

RackNerd's OFFICIAL LET BLACK FRIDAY! HUNDREDS OF GIVEAWAYS + CRAZY DEALS, MASSIVE (come see)

Comments

PARTY-PEOPLE-OF-THE-RACKNERD-COMMUNITY