glyphsearch
glyphsearch copied to clipboard
Adding more libraries
GlyphSearch was originally intended to support only the most popular libraries, which cover the majority of use.
But as we receive requests or contributions for new libraries (#30, #32, #33), we should discuss whether GlyphSearch should strive to be a comprehensive meta-resource or not. Sometimes less is more. Then again, sometimes more is more. :smiley:
If we decide to continue adding libraries, this requires a rethinking of the interface, especially with filtering libraries. Perhaps users can be selective about which libraries to include, and this setting can be saved across sessions.
The benefits of being inclusive are obvious. But are there any downsides or objections?
Go for quality works. So merge 33 if you like we have done good work and icons stand firmly.
I hope you will like ours. Thanks.
Props to @thomaspark for opening up this discussion here! :+1:
I definitely see the point that at some point too many libraries will simply clutter GlyphSearch, and make it too hard to easily use - and perhaps one of the great things about GlyphSearch is also the curation aspect. The fact that I can count on the libraries listed on GlyphSearch to be of a certain quality!
I don't know if it's possible to write down a set of requirements that a library will have to fulfil to be added to GlyphSearch, I think it's more of a subjective thing whether you(@thomaspark/community) find a library worth adding to the collection.
Perhaps users can be selective about which libraries to include, and this setting can be saved across sessions.
:+1:
@thomaspark did you try and ask Twitter what people thought of limiting the libraries?
Would be great to hear what the other contributors think of this: @redox + @jenil27
I'm a Glyphsearch user who uses the service a few times per week but I must admit that I mostly use it to search icons from a specific library, not cross-libraries (in my case: fontawesome 90% of the time). So the first thing I do before searching is selecting a library in the menu: it's still a way better than the official website of each of those libraries.
That being said, having subdomains to reach more easily what I'm looking for (fontawesome.glyphsearch.com for instance) + a kind of "Others" submenu listing all "not-so-popular" libraries is maybe worth it. So the only open-question is: do we need to search in all libraries when no-one is selected? Or maybe we just need to rank the results by library popularity? I mean 90% of users don't use Google's 2nd or 3rd page, but when you need it... Thank you Google, it's possible :)
So maybe we just want pagination here, to keep "popular" libraries on top. The "popularity" criterion can even be automatically fetched from Twitter http://urls.api.twitter.com/1/urls/count.json?url=PROVIDER_WEBSITE (http://urls.api.twitter.com/1/urls/count.json?url=http://fontawesome.io/ for instance) to feed Algolia's engine.
Hope that my 2cts helps :)
Hello guys,
Thanks @matiassingers for including me in this discussion & obviously @thomaspark for starting this one, which was indeed essential :+1:
The one thing I love about GlyphSearch is it's simplicity to use!
So like Thomas mentioned we would have to rethink the interaction of GlyphSearch.
There are a few thought running in my mind right now like:
- Our icons are currently stored in a JSON and I think we can take advantage of this fact, how about adding a functionality of offline storage using Application Cache? so that even if the libraries increase the application is running fast enough. [ref. devdocs.io]
- Also I like what @redox says, most of the users come with a library already set in their mind, so how about we store the users selections/options in the LocalStorage and serve them better? So this solves our do we need to search in all libraries when no-one is selected
- I think libraries should be made search engine crawl-able, we can probable use a
#!
in the URL and make is crawl-able. This way the popular libraries will automatically have a high ranking and users can discover icons even faster.
Hope this gives some insight :)
@redox that's a very good point. I hadn't really thought about it, but that's exactly how I primarily use GlyphSearch - searching for icons in the one or two icon packs I use in a lot of projects.
I don't really compare icons across libraries that often, maybe more of a need when starting out a project/choosing the library initially.
Also https://github.com/thomaspark/glyphsearch/issues/37
I think the way to go is to have the most popular libraries (Font Awesome, Glyphicons) as defaults, and allow users to filter other libraries.
Several changes are needed for this to happen, but in the meantime I'll keep track of requests here, instead of opening a new issue for each.
- Devicons (#32): http://vorillaz.github.io/devicons/#/main
- Icon Mart (#33): http://iconmart.in/
- Iconic: https://useiconic.com/open
- Map Icons (#37): http://map-icons.com/
- Material Design (#35): https://github.com/google/material-design-icons
- Semantic UI (#38): http://semantic-ui.com/elements/icon.html
Would you also consider including the Unicode Emojis? I think a few of them would intersect.
GitHub has a nice overview of the 15 most stared icon fonts: https://github.com/showcases/icon-fonts?s=stars
Just did a major redesign. Now GlyphSearch has a fixed sidebar, which gives us more flexibility in adding libraries.
I also added the Material Design Icons.
It makes sense to use a popularity criterion like GitHub stars to decide what other libraries to add. If we continue to add more, we may need to change the "All" filter to some sort of "Popular" filter that includes only a few of big dogs.
If we continue to add more, we may need to change the "All" filter to some sort of "Popular" filter that includes only a few of big dogs.
Or support selecting multiple libraries with checkboxes or pressing shift...
That would be a good solution as well.
A good list of Icons Fonts by the Sublime Text package: https://github.com/idleberg/Icon-Fonts-Sublime-Text#prefixes