The road to the next major release (2.2)

Linc

A few updates since we recently synced up on our expectations for our next open source release internally.

1. PHP: The minimum PHP version is moving from 5.2 to 5.4. A version checker needs to be added or refined.
2. Addons: We are definitely including Advanced Editor. We are definitely not releasing Reactions yet. There's too much work left there.
3. Features: Our main feature hurdle is getting our emoji parsing squared away. We need to review it so we don't get incorrect/unfinished implementations in the wild.
4. Bugs: My biggest bug concern is our Markdown support. It's kinda gross right now. Related: Spoilers support.
5. Versioning: We're still discussing our version number strategy. Suffice to say: the skip-a-number strategy isn't set in stone yet.
6. Workflow: I'd like to see outstanding pull requests winnowed significantly before the release cycle starts.

I am very tired of backporting fixes to a version that is 1200 commits behind (hello, conflicts and double chores), so I'm looking forward to getting this party started soon. If there are things you want to see resolved before the next release, start filing issues & pull requests.

Anonymoose

@Linc said:
1. Features: Our main feature hurdle is getting our emoji parsing squared away. We need to review it so we don't get incorrect/unfinished implementations in the wild.

Emoji in the database will require mySQL 5.5.3 or higher that supports and uses the utf8mb4 charset. See: http://vanillaforums.org/discussion/21667/solution-to-support-additional-unicode-characters

Linc

@Anonymoose said:
Emoji in the database will require mySQL 5.5.3 or higher that supports and uses the utf8mb4 charset.

"Emoji", as in our take on emoticons & a system for using your own sets, not the official Unicode standard natively in the db.

hgtonight

This is more of a support request, but tightening the versioning between core plugins and the releases would be awesome.

I think that is more a request for the Community repo.

Bleistivt

@Linc said:
1. Features: Our main feature hurdle is getting our emoji parsing squared away. We need to review it so we don't get incorrect/unfinished implementations in the wild.

I really like the system currently in master where you can just use Emoji::setFromManifest() to deploy Emoji sets via plugins.

R_J

@Linc said:

>

4. Bugs: My biggest bug concern is our Markdown support. It's kinda gross right now. Related: Spoilers support.

It's really no fun, but the library you are using exists in a newer version and if you look at that list: https://packagist.org/search/?q=markdown&search_query[query]=markdown&page=1 (michelf/php-markdown) it seems to be very popular. It also seems to be quite fast: http://kzykhys.com/blog/benchmarked-markdown-parsers-written-in-php/ Though parsedown (which has alreadybeen suggested by Bleistivt) is more than 3 times faster in that comparison.

Isn't it possible to simply update the used library or has it changed so much that it would require further customizing?

Anonymoose

No moose emoji in the unicode pipeline. 😈

vrijvlinder

nope ☠ , but it there was it would be filtered. Here at least

http://getemoji.com if you add the emoji js to your forum it can work. But here they are filtering as it may seem , but no butterfly or moose or gecko

Anonymoose

Any sort of image replacement method is hacky and Chrome 🔧 was the last browser to move forward ⏩ with native support for emoji. Every modern browser and os today has the fonts and capability to display emoji. 🙈🙉🙊 But not every database can store them in native form. 👹

hgtonight

Damn kids and their emoji.

Back in my day you had to turn your head sideways to understand pictures!

vrijvlinder

@hgtonight said:
Damn kids and their emoji.

Back in my day you had to turn your head sideways to understand pictures!

back in my day, polaroid was high tech....

vrijvlinder

@Anonymoose said:
Any sort of image replacement method is hacky and Chrome 🔧 was the last browser to move forward ⏩ with native support for emoji. Every modern browser and os today has the fonts and capability to display emoji. 🙈🙉🙊 But not every database can store them in native form. 👹

I know , it is incomprehensible .... 👹

Anonymoose

@vrijvlinder said:
back in my day, polaroid

Please post some.

vrijvlinder

nope

JasonBarnabe

@Anonymoose said:
Any sort of image replacement method is hacky and Chrome 🔧 was the last browser to move forward ⏩ with native support for emoji. Every modern browser and os today has the fonts and capability to display emoji. 🙈🙉🙊 But not every database can store them in native form. 👹

For what it's worth, I run my Vanilla install with a utf8mb4 DB. Works fine, except for https://github.com/vanilla/vanilla/issues/2082

JasonBarnabe

@R_J said:
It's really no fun, but the library you are using exists in a newer version and if you look at that list: https://packagist.org/search/?q=markdown&search_query[query]=markdown&page=1 (michelf/php-markdown) it seems to be very popular. It also seems to be quite fast: http://kzykhys.com/blog/benchmarked-markdown-parsers-written-in-php/ Though parsedown (which has alreadybeen suggested by Bleistivt) is more than 3 times faster in that comparison.

>

Isn't it possible to simply update the used library or has it changed so much that it would require further customizing?

The library has already been upgraded on master. I've been using these changes and haven't had any problems. I believe @Linc is referring to the Vanilla-specific Markdown customizations, like spoilers and stuff that lets you write haikus.

JasonBarnabe

Workflow: I'd like to see outstanding pull requests winnowed significantly before the release cycle starts.

Yes please! I have to backport my pending PRs on each new release. It would be great to not to have to for 2.next, at least initially.

Bleistivt

Some of my most-wanted (to be fixed) issues for the next release would be

https://github.com/vanilla/vanilla/pull/1942
https://github.com/vanilla/vanilla/issues/1410

and what do you think about refactoring Emotify into an emoji set?

Linc

The Livequery deprecation is already merged.

what do you think about refactoring Emotify into an emoji set?

I like it a lot. Might wanna wait til we finalize the support tho.

Linc

I reorganized the GitHub milestones today to signify that I believe 2.1 is winding down and we should be targeting master with most non-critical fixes in anticipation of the next version. I am attempting to get Advanced Editor moved to the primary repo as soon as possible.

Linc

Soliciting feedback on altering the plugins distributed in the next version:

https://github.com/vanilla/vanilla/issues/2420