breck7 / ohayo Goto Github PK

There's a cool new python word cloud package. Would be neat to implement this an ohayo tile(s) to replace current wordcloud.

https://github.com/minimaxir/stylecloud

Would it be useful to make tile(s) for this:
https://cloud.google.com/bigquery/public-data/

Should have a show static tile where you can just plugin a hard coded number.

If you just want to drop a couple of columns, and keep most of them, would be nice to have the inverse of columns.keepy

https://github.com/treenotation/ohayo/blob/1574a7f7efeccec98c9a5eb1ab9a6b48e10c3078/flow/packages/columns/columns.gram#L77

https://data.commerce.gov/

Let's add a spotlight tree component that shows available commands and has autocomplete...Like sublime text...Especially as number of commands grows this would be helpful...

Could be a generic reusable tree component in jtree

Trying to add the Gutenberg tile (#24), I realize for most interesting public datasets even if you just keep the minimum in CSV it's still going to be a huge download which means huge parse time as well (at least 3MB uncompressed min for this particular set).

As the saying goes "easier to move the compute to the data than the other way around".

I wonder if there's a simple way to turn any CSV into a web service? Perhaps using BurntSushi's awesome xsv library or some equivalent in Go.

https://github.com/woop/awesome-quantified-self

• Plus sign in the bottom right?
• Click to add?
• toggle full screen
• toggle gutter keeps space there
• can’t right click
• can’t double click to add
• need better suggestions

Would be great to have access to all project gutenberg books right in ohayo

https://www.gutenberg.org/

Would love tiles that provide suggestions (potentially can even choose best autocorrect method and implement that) for correcting missing data.

https://cran.r-project.org/web/views/MissingData.html

Perhaps the flow might be something like:

web.get somefilewithmissingdata.csv
 table.basic
 missing.fix
  table.basic

Would be nice to be able to compile OhayoLang programs to:

• PNG
• Animated Gif
• PDF
• HTML

https://roughjs.com/

[] when running tests should we check to see if fabServer is started? perhaps start it for having the web.get test pass?
[] have tests for server methods for desktop version. perhaps create a temp project directory?

Would be cool to have a tile(s) for this:

https://github.com/awesomedata/awesome-public-datasets

Should it be called Flow? Forest?

perhaps some tiles from this project:

https://www.gdeltproject.org/

in particular the 3.5m book dataset (with 1m full text) on google big query.

Would be nice to remove as many dependencies as possible from the core.

These are good for starters:
breck7/jtree#99
breck7/jtree#100

We should also be able to remove a number of others.

Dependencies in package folders are fine and not something to worry about.

The link https://github.com/breck7/ohayo/blob/master/2017-06-21-show-hn-programming-is-now-two-dimensional.html points to a 404.

Lots of wikipedia tiles would be great

Would be neat to have some tiles for this:

https://aws.amazon.com/blogs/aws/aws-data-exchange-find-subscribe-to-and-use-data-products/

A sample program might look something like this:

html.h1 Amazon Data Exchange Providers
amazon.dataExchange.providers
 table.list

Would be cool to have a tile for this:
https://github.com/awesomedata/awesome-public-datasets

Perhaps the best way to deal with breaking changes to Maia is:

• going to ohayo.computer just redirects to latest stable version
• deep links thus will always work since those versions will always be up
• we could have a migrate.ohayo.computer or something, basically branches dedicated to migrating versions if someone wants to auto do it

Idea: scan the program. If you encounter a "hidden", then default program tile is visible. If you encounter a "visible", then the default is hidden.

The autocomplete is pretty dumb right now. Let's move some stuff from the prototypes over to make it smarter.

Add an easy command to export/import all data from localStorage to faciliate easy switching between servers and backups

experiment with sibling data tile flow

instead of:

samples.portal
 vega.bar

should we allow:

samples.portal
vega.bar

Could potentially use newlines as sort of a "clear".

Of what the rows are.

The vega tiles are obviously critical. Would be good to get more help with them.

For meta studies, it seems like it would be cool to import multiple maia programs. Perhaps it's best to do this as a build script, and create a simple meta language for it?

So, out of curiosity, how is this different from things like TreeSheets, or "sweet syntax" or various lisps that use deeper tree notations (think SRFI-101-based lisps) or things like Lush or Klone? YAML and Rebol seem to be similar predecessors…

I'm just curious if I missed something obvious, but the papers didn't demonstrate anything that made these wholly unique compared to other previous attempts.

Thanks!

https://www.w3.org/TR/?tag=data

Add a better auto height layout that just uses browser if possible for computing height.

Datasette has some great datasettes. https://github.com/simonw/datasette

Also might make more sense to use that for this: http://datasets.ohayo.computer/

Or just improve experience/design so we need less scrolling?

This is a neat project and there are perhaps templates + plugins to share:

https://github.com/huginn/huginn

Would be cool to have a tile(s) for this:

https://www.tensorflow.org/datasets/catalog/overview

repro steps:

1. go to ohayo.computer
2. create reddit template
3. in source editor, change line 9 scatter xColumn

expected:
correct column names suggested in autocomplete

actual:

https://usafacts.org/

• let's rigorously explore and define the "." qualifier and how it will work going forward, and how contributions will work

• experiment with a "doc.use samples vega" syntax for shortening tile names?

• experiment with short dictation dialect

repro:

1. go to ohayo.computer
2. shift+m

expected:
i should see all files in folder

actual:

sample program:

https://ohayo.computer/?filename=myprograms.maia&nodeBreakSymbol=%7E&edgeSymbol=_&data=editor.files%7E_hidden%7E_rows.sortBy_link%7E__hidden%7E__editor.gallery

At the moment FlowTiles is the main focus, but could lead to code improvements and technical clarity I think if we had some basic editing scenarios for tiles files.

Imagine having every academic paper ever, at your fingers in Ohayo.

http://ma-graph.org/

Might be fun to have wayback machine tiles and do things like:

wayback.get http://nytimes.com 1999 2019
  wordCloud

https://github.com/timqian/chart.xkcd