Not enough documentation to understand Tejú Jaguá.

[trcrsired]

I still don't fully understand how the algorithm works. Can you join my Discord to explain to me or the Jk-jeon? Jk-jeon (the author of Dragon Box) is my friend and on my Discord.

If this could work out, I may change my floating point implementation of fast_io from Dragonbox to this one.

https://discord.gg/b968Rt7GnF

[cassioneri]

@trcrsired Thank you very much for your interest in Tejú Jaguá.

I appreciate that, at the moment, it's not easy to understand the algorithm due to the lack of documentation. The talks that I gave at C++ conferences provide an overall idea of what goes on behind the scenes but are not enough to have a full understanding of Tejú Jaguá.

As the project's README.md states, this is a working in progress and in the next few months I'll be working on providing more detailed documentation. Specifically, I'm going to write an academic paper with a more detailed description of the algorithm alongside its proof of correctness. Unfortunately, this will take a lot of time since my work on this project is entirely carried out in my spare time which is quite scarce.

Like you, others have expressed interest in understanding and implementing Tejú Jaguá in different libraries. Unfortunately, again because my scarce time, I can't explain the details on a 1-2-1 basis. Please be patient and come back in the future when, hopefully, the documentation will have answered all your questions.

[trcrsired]

@trcrsired Thank you very much for your interest in Tejú Jaguá.

I appreciate that, at the moment, it's not easy to understand the algorithm due to the lack of documentation. The talks that I gave at C++ conferences provide an overall idea of what goes on behind the scenes but are not enough to have a full understanding of Tejú Jaguá.

As the project's README.md states, this is a working in progress and in the next few months I'll be working on providing more detailed documentation. Specifically, I'm going to write an academic paper with a more detailed description of the algorithm alongside its proof of correctness. Unfortunately, this will take a lot of time since my work on this project is entirely carried out in my spare time which is quite scarce.

Like you, others have expressed interest in understanding and implementing Tejú Jaguá in different libraries. Unfortunately, again because my scarce time, I can't explain the details on a 1-2-1 basis. Please be patient and come back in the future when, hopefully, the documentation will have answered all your questions.

You can come to my discord and explain that to jk jeon. He is my friend and may help you.

[trcrsired]

Also, I want to know whether there are better algorithms for parsing strings to float.

[cassioneri]

As I said, unfortunately, I don't have time to go on a 1-2-1 basis to explain Tejú Jaguá. I'm really sorry about that. :cry:

Regarding converting string to float, I'd suggest taking a look at https://github.com/fastfloat/fast_float by Daniel Lemire. I doubt there's anything better than that.

[trcrsired]

ty ty