cbenz

Tap-to-dictate voice input for Android — on-device or cloud transcription, AI-powered cleanup styles, works in any app. No backend, your keys, your data.

This software allows to use a local or remote (cloud) service for audio transcription and text cleanup.

I've been using it for 2 weeks, it's working great and development is active.

Voice dictation changed my life, but I didn't want to pay for a proprietary app or service. This app is GPL-v3 licensed.

I found the floating button approach better than the keyboard approach : I can keep my favorite keyboard while having a small floating button that is only shown when focusing a text input field.

Another useful feature (not yet mentioned in the README): you can select text in an input field, select "share" then "ramblr" and apply one of the cleanup prompts to transform the selected text (these same prompts that are used for cleanup after audio transcription).

Keyboard driven, and a lot more!

I love the following patterns for GUIs:

- command-driven (e.g. command palette of VSCode) - keyboard shortcuts call commands, are customisable - scriptable: ability to call those commands from the outside, controlling the running app, for example with CLI calling D-Bus (or socket), and not necessarily having macros from inside the GUI - reorganisable: the different GUI elements like panels, tabs, should be moved by the user and remembered (ideally having layout presets) to let the user adapt the GUI to his/her needs.

I'm thinking of VSCode, Blender

I'm not especially aware about common toolkits like Qt or GTK, whether they provide tools for that.

In web front-end dev, I'm using Svelte, and there are awesome UI libraries like shadcn-svelte, however I didn't find anything rather standard implementing these architecture patterns.

Same here. The only problem is that I "only" have 24Gb of RAM. I wish I could upgrade but it's a hard limit. And keyboard quality seems to have been degrading over the years since 2020. Is this new model good in terms of keyboard?

Yes via the "--system" option of the different commands.

Or via the global "python-preference" option set to "only-system".

Cf https://docs.astral.sh/uv/concepts/python-versions/#adjustin... and https://docs.astral.sh/uv/reference/settings/#python-prefere...

Absolutely. It changed my Python developer life \o/

I would tend to rely on a single tool, do you see advantages to use both?

Do you happen to have discrepancies between both? (e.g. an error raised by one and not the other)

ruff, then uv, then rye All of them changed my pythonist everyday life, one after another.

Great thanks to all the team.

I had to forbid Happy Tree Friends to my son after he tried... no, just joking

Indeed I can't see the point of shipping the "AI first" feature as a whole editor. It seems to me that it should be a feature of any editor, or else I'd like to be convinced by arguments less marketing oriented than those I see on the home page.

Sorry, too late, I'm french.

I tried and just typing letters is slow, they appear ~1 second after typing. Deleting text is slow also. I'm under Linux+Chrome.

This argument sounds to me like security through obscurity [1], the same way some people incriminate emergency evacuation maps of buildings when there is a shooting because they helped attackers find their way inside...

Worldwide shared knowledge versus protection against potential attacks.

> It happens more often than people think

This argument seems to be a strawman, what is the real probability of a wood thief to commit their business, and in particular with the help of that map?

[1] https://en.wikipedia.org/wiki/Security_through_obscurity

That's exactly the reason why we created https://db.nomics.world

Dat is more about distribution (décentralized, distributed, P2P) but it's not possible to make queries.

Great news, we're using it at [1] and like it very much. Congrats to Tailwind team.

[1] https://db.nomics.world/

Using the Svelte JS library (https://svelte.technology) you can have both: server-side rendered components [1] and compile to web components (custom elements) [2]

Another advantage is that the components compile to vanilla JavaScript, so we don't rely on a runtime library to run the application.

[1] https://svelte.technology/guide#server-side-rendering

[2] https://svelte.technology/guide#custom-elements

Some links:

git-ssb (Git over Secure-Scuttlebutt):

- https://git.scuttlebot.io/%25n92DiQh7ietE%2BR%2BX%2FI403LQoy...

- https://github.com/noffle/git-ssb-intro

Secure Scuttlebutt (https://www.scuttlebutt.nz/)

I prefer Scuttlebutt because of the P2P architecture which allows offline-first and no-backend. With Mastodon I still have to have an internet connection, and create an account on a server.

Users manipulate applications named Patchwork or Patchbay, for instance. Scuttlebutt is the protocol.

Or Secure-Scuttlebutt, which vehiculates the concept of identity as a first citizen.

https://scuttlebutt.nz

Here, the social aspects could be useful, like following my friends mind maps etc.

Is the source code available? For the whole product, or else for the table (or grid) component?

As an Elm developer I would be interested to dive into your implementation and perhaps contribute.

Despite the "damn" and "hell" words you use, let me develop my opinion.

It is not the user rights which allow them to do whatever they want. I totally agree that one must not violate the terms expressed by the author.

I don't like to talk about "my" creation or "yours" or somebody's. Because every creator is inspired from previous ones. Someone obviously works for a certain amount of time, but the concept of paternity is strange given all the inspiration. When I write free software, I always consider it would not have been possible to do it without libraries and other inspiration. So it's normal to give back the product of my time to the world, which gave me the base to do it.

I create things not to do business, but to make the world match my ideas and hope people will follow and improve. And I would really be pleased if somebody has a complaint or criticism, and takes the time to "fix" it concretely by coding it, not just expressing it.

In arts (music, painting, etc) this is natural, as in science and research. About computer science, how people can improve then if it's not free software? That would be a "use it or drop it" situation, not adapted for collective improvement.

So using free software licenses, as a creator, makes the world better. The contrary doesn't make the world worse, but freezes it until free software appears. That's waste of time in a sense!

That's why I don't recommend people to use proprietary software. Because it maintains the world in a situation which is not ideal, globally.

I understand people want to make money for a living. But there are economic models quite OK with free software. We have to find a way, I mean collectively, to finance free software. That would be an improvement for society,better than defending proprietary software.

Finally pardon my English, I'm not native.

Freedom of users is more important than freedom of vendor. That's because users are the society, the citizens, everyone in fact.

So defending the freedom of the creator to sharing software under her own terms is way less important than defending the freedom of users.

Frontend: definitively Elm.

Backend: Node for popularity, Elixir or Haskell to try something new, or Python Flask which I already know.

I use http://workflowy.com/

In short it's a tree with potentially infinite depth, search, tags (#work) and contacts (@mummy), and the ability to change the root of the tree.

It's also possible to share nodes in read or read/write with other users. It's exportable to Markdown and other formats.

Sadly Collate notes nor Workflowy are free software (as in libre), and I cannot install them on my server :-(

There is no "starter kit" to bootstrap a new country, but the project architecture is designed to clearly separate the core engine from the modeled country.

As this question has occured many times, we're certainly going to work on this topic and add a chapter to the documentation "Howto bootstrap a new country".

For now starting a new country consists of creating a new repository like openfisca-<country_name>, implement what we call entities (ways to group people together like households, families, etc.) and implement the formulas corresponding to the legislation taxes and/or benefits you want to compute.

This involves a lot of reverse-engineering work! If someone is interested, please contact us by opening a thread on the forum (https://forum.openfisca.fr/) or a GitHub issue (https://github.com/openfisca/openfisca-core/issues)

Hi,

One of the developers here :)

One of the original authors has both nationalities French and Tunisian and he started to implement Tunisia, but it's a "work in progress", without UI yet. Source code: https://github.com/openfisca/openfisca-tunisia

The team is happy to welcome newcomers for new countries.