tolc VS Reduct Storage

I love CLI11! It has been pretty easy all the way through. Used to use lyra before but since I wanted to have subgroups in Tolc I had to switch. Great job on CLI11 if the author is in the chat :)

I'm working on a project called Tolc that is generating bindings from C++ to other languages. When creating bindings to a class MyClass, it needs to be defined before any code using that class (for example a function that returns an instance of it). Therefore I needed to know in which order to define things. Honestly I just had so much fun solving this problem (using some C++20 and features) that I wrote a post so sum it all up:

Hi everyone!

Some friends and I were unhappy with how much work it was to use C++ from other languages. We were working in the optimizations industry and often had to prototype with python. Eventually, we created a tool to make it easier for C++ to talk to python (by generating pybind11). Later, javascript via WebAssembly was added as well. It's now at a point where it's very easy to just create a C++ library and use it from any of those languages without change. We're planning on slowly adding more languages as needed/requested. The next on the list are Swift and Kotlin.

It does not require any change to your existing public interface, but simply reads it and creates the bindings off of that. It should also work on Linux (Debian), MacOS, and Windows (Visual Studio). Here's a small demo if you'd like to test:

https://github.com/Tolc-Software/tolc-demo

And here are the repositories with the source code:

https://github.com/Tolc-Software/tolc - The executable

https://github.com/Tolc-Software/frontend.py - The python bindings generator

https://github.com/Tolc-Software/frontend.wasm - The WebAssembly bindings generator

https://github.com/Tolc-Software/Parser - The C++ parser

It is dual licensed with AGPL and, if someone wants, a commercial license as well.

Would be cool if someone finds it useful!

https://github.com/Tolc-Software/tolc - The executable and CMake wrappers

I’m just gonna drop this here: https://github.com/Tolc-Software/tolc

When it comes to computer vision, data storage is a critical component. You need to be able to store images for model training, as well as the results of the processing for model validation. There are a few ways to go about this, each with its own advantages and disadvantages. In this post, we’ll take a look at three different ways to store data in computer vision applications: a file system, an S3-like object storage and Reduct Storage. We’ll also discuss some of the pros and cons of each option.

If you've been using the Reduct Storage client-side SDKs for Python, JavaScript or C++ to interact with your storage instances, then you'll want to upgrade to the latest version. This week, all of our SDKs were updated to support Reduct Storage v1.1.0 with Token API. With this new release, we've made some important enhancements and bug fixes that will make your development experience smoother and more productive. So make sure you upgrade today!

Reduct Storage v1.1.0 comes with a new Token API to create tokens with different permissions. This allows you to give different users or applications access to specific data or storage functions. For example, you could create a token that gives full access to all data, or one that only allows writes to a certain bucket. The Token API is simple to use and gives you a lot of flexibility in how you use Reduct Storage.

A bucket is a top-level data container of Reduct Storage. You can consider it as a folder of a file system. If you delete it, you delete all files inside the bucket. Also, it has common settings for all entries, like the quota or maximal block size.

The client has ben updated for compatibility with v1.0.0 of the storage engine's API, which is described here.

As you can see, Reduct Storage is always faster for write operations (16 times for 10 Mb blobs!!!) and a bit slower for reading when we have many small objects. You may notice that the speed decreasing for both databases when we reduce the size of chunks. This can be explained with HTTP overhead because we spend a dedicated HTTP request for each write or read operation.

You also can download binaries and run them:

Hey everyone, I just released the Reduct Storage SDK for C++ v1.0.0. Now it supports Reduct Storage v1.0.0.

We released the next major version of Reduct Storage and broke compatibility with the current client SDKs 😄. No worries, we'll fix everything soon.