wefindx / metadrive Goto Github PK

....

ok

Right now, the data is returned unnormalized, and without the polycontext metasymbol (MFT-1) * attribute by default. However, developing the data-browser (viewer) then requires to adapt the ReactJS widgets to each situation.

Since the metaform package already exists, and provides required features, open source it, and make it a dependency. Let it the API default to normalized data fields and values.

Note: this requires fixing the latencies (pre-caching) of schemas for metaform. It is very slow now, because it does not cache schemas from wikis.

Possible solutions:

1. Make it auto-install the new driver versions if present by default.
2. Make it ask to upgrade.
3. Make not upgrade.

So, it is best to provide a setting in the ~/.metadrive/config about the drivers update policy, say, with a default not equal to 2. (which hangs API)

This also affects the use of yolk3k==0.9, which has hard-coded XML_RPC_SERVER = 'https://pypi.python.org/pypi' in .env/lib/python3.7/site-packages/yolk/pypi.py, which prevents use of the package in some regions. The 'https://pypi.org/pypi'` is generally more available.

You need metadrive, because you want to design one interface to all world's services, so that you don't have to manually interact with them. In that sense, any person, who has to work with various systems, needs a metadrive to avoid having to learn to work with each of them. Also, every intelligence agency needs a metadrive, to be able to automate data collection and actions.

As a person, using metadrive, you will be able to be a "universal peer" that freely interacts with all protocols. In daily life, e.g., write your updates in any editor you like in one place, and have it published in any places (any social media or hardware groups) you like, and have all interactive comments from all of those places without having to manually interact with those interfaces.

As an organization, using metadrive, you will be able to take multiple applications, and automatically collect and merge data about all of them, to create AI insights automatically, without any manual human integration work, and also, execute actions on those systems as a result of AI decisions.

Metadrive is the syntax sugar for merging world into one hypergraph, where each driver provides hyperedges.

Creation of config asks for github username. Solution could be use .gitconfig's username, or leave as None. if None, the code should throw proper errors and notifications, if the feature requires editing config to enter GH username.

Currently, only subtools like _selenium, _requests utilize the proxies defined in ~/.metadrive/config, however, when bootstraping (e.g., looking up for drives in PyPI, etc.), the direct connection is used. Reuse the proxies defined in the config for everything.