7 ms·

> it is overkill. > It definitely is. It is not, you pay $14 and you get basic analytics and this https://usefathom.com/news/anonymization https://usefathom.c

> it is overkill. > It definitely is. It is not, you pay $14 and you get basic analytics and this https://usefathom.com/news/anonymization https://usefathom.com/news/anonymization that's worth the money for me. > It's cheaper, faster (by a 3-4x margin), and simple. What about user privacy? are you anonymizing the users?

sdan 7y ago
As a student (and a developer) I see paying $14 a month a huge expense. To really promote the privacy and accessibility for all, I'm building this to make it free and will absorb the compute costs if I have to for the sake of value for customers. Finally, yes, I'm planning to utilize a variety of shuffling and hashing techniques to anonymize users, data, and all other aspects. At the moment, I'm focusing on getting under 90ms on average globally (which I have achieved, just not implemented into the master branch yet)... which is over 4.5x faster than Fathom's average of 420ms globally. More coming soon! Fill out if interested: https://forms.gle/BZA7thD878PbKnSs8 https://forms.gle/BZA7thD878PbKnSs8.
- JackWritesCode 7y ago
  Keep pushing man. We're getting between 300ms - 600ms response from your endpoint (from a single datacenter) but I wish you luck with hitting under 90ms average.
  - sdan 7y ago
    Actually addressed that on the website. Currently I'm getting an average of 1 second globally which is super slow. I'm going to migrate to the faster branch soon, which wouldn't be much of a problem, but am going to release a bunch of stuff alongside it.
- pauljarvis 7y ago
  Your stats are wholly incomplete compared to Fathom, lack any kind of GUI, graphs, or date-based filters. I think the comparison in speed is like comparing an apple to an elephant :)
  - sdan 7y ago
    Yeah I'm lacking a ton of stuff. No GUI, nothing in fact. But the tracker I've implemented (not into prod yet) does get under 90ms globally. Your service, which I've seen is serverless, is actually pretty fast compared to other services and for that I have to admire your work. However, as I've done some response time testing on Postman (https://twitter.com/notsuryad/status/1199201916850278400?s=20 https://twitter.com/notsuryad/status/1199201916850278400?s=2...) I've found that my service is getting right next to Google Analytics' response times. I'll release more details about how I've conducted this soon, but I hope the best for Fathom (I've used Fathom self hosted before but ran into some issues).

sdan 7y ago

As a student (and a developer) I see paying $14 a month a huge expense. To really promote the privacy and accessibility for all, I'm building this to make it free and will absorb the compute costs if I have to for the sake of value for customers. Finally, yes, I'm planning to utilize a variety of shuffling and hashing techniques to anonymize users, data, and all other aspects. At the moment, I'm focusing on getting under 90ms on average globally (which I have achieved, just not implemented into the master branch yet)... which is over 4.5x faster than Fathom's average of 420ms globally. More coming soon! Fill out if interested: https://forms.gle/BZA7thD878PbKnSs8 https://forms.gle/BZA7thD878PbKnSs8.

JackWritesCode 7y ago
Keep pushing man. We're getting between 300ms - 600ms response from your endpoint (from a single datacenter) but I wish you luck with hitting under 90ms average.
- sdan 7y ago
  Actually addressed that on the website. Currently I'm getting an average of 1 second globally which is super slow. I'm going to migrate to the faster branch soon, which wouldn't be much of a problem, but am going to release a bunch of stuff alongside it.
pauljarvis 7y ago
Your stats are wholly incomplete compared to Fathom, lack any kind of GUI, graphs, or date-based filters. I think the comparison in speed is like comparing an apple to an elephant :)
- sdan 7y ago
  Yeah I'm lacking a ton of stuff. No GUI, nothing in fact. But the tracker I've implemented (not into prod yet) does get under 90ms globally. Your service, which I've seen is serverless, is actually pretty fast compared to other services and for that I have to admire your work. However, as I've done some response time testing on Postman (https://twitter.com/notsuryad/status/1199201916850278400?s=20 https://twitter.com/notsuryad/status/1199201916850278400?s=2...) I've found that my service is getting right next to Google Analytics' response times. I'll release more details about how I've conducted this soon, but I hope the best for Fathom (I've used Fathom self hosted before but ran into some issues).