Tiddlywiki5: Plugin Finding -- can they broadcast their existence?

[TiddlyTweeter]

I find the finding of plugins that have essential utility very laborious.

Is there not a way that when its created a plugin can BROADCAST its existence to a central repository? Are we so distributed we can't find a decent centralised solution?

My focus here is on getting TW used more. Much of its utility is revealed most via plugins. Currently finding them is a huge mess. Its consequently v. hard, particularly for beginners, to find functional solutions that plugins deliver.

This is damaging TW's uptake imo.

[sukima]

There was talk at about moving plugin's into NPM modules. By doing so TW plugins can easily be discovered through NPM.

[TiddlyTweeter]

Ciao @sukima would that mean you'd need be on node?

A recent effort by David Gifford to list TW resources, mainly plugins ... https://dynalist.io/d/zUP-nIWu2FFoXH-oM7L7d9DM ... I think illustrates the issue that is escaping solution. Its been started before on Reddit by Riz ... https://www.reddit.com/r/TiddlyWiki5/wiki/pluginsandresources. And Erwan's (good but very underused) aggregator ... http://erwanm.github.io/tw-community-search/#CommunityAuthors:CommunityAuthors underlines it.

Problem is we have zillions of good plugins (and likely some bad ones) all in a lost forest. Many developed outside any shared framework. Simply as TW. No shared code base. Just individuals having great ideas.

I'm not god. And I'm not a programmer either. I'm not sure what a decent solution could look like.

But I AM clear that the way that major additions to TW by independent coders currently works often makes it hard for their work to find its lasting place.

Overall, I think its damaging TW in two ways:

• Beginners find it very difficult to find solutions because there is very poor indexing of what is;

• Its (in a way) disrespectful to the plugin makers--they left to the entirely arbitrary Darwinism of the Google Group whether 1 or 50 read them on the day of launch.

I'm sure we can do better. But I can't see quite how without some method to automatically aggregate?

[sukima]

would that mean you'd need be on node?

Not necessarily. My comment was in reference to establishing a package idiom all plugins follow. NPM offers this out of the box. There for is you run TW in node you gain the advantage of npm install --save my-awesome-plugin but if you use just the HTML version then you simply head over to npmjs.com and search for some established keyword like tw-plugin.

Ultimately we need a package manager. whether that is home grown or a different tool. My personal opinion is to use the resources already available and let NPM do the package management.

[danielo515]

I was the one that proposed to use NPM for plugin distribution, and I still think it is a great idea.
If you are using the node version, or compiling a TW from node, or you build automatically any TW on Travis (like my tutorial/template of automatic TW on github pages) you just have all this for free.
For those using the browser we can build a small server-less architecture that serves the plugin the TW plugin library asks for. Many server-less providers offer reasonable free tiers, so this can come also for free.

[TiddlyTweeter]

If a developer who was simply using their browser to develop plugins could also get it known this way, without too much effort, it could work. One thing clear to me is that the current way we publish add-ons is a mess and its counter-productive. It should be advancing TW & it isn't.

[danielo515]

There are some devs. building plugins on the browser , Ok, they are crazy enough. I don't see how making this available first to node ones can hurt in any way