code sprint thoughts

I spent Friday morning/midday at the Mentored Core Sprint. It was a weird mix of extra-hand-holding and being thrown in the deep end, at least for me. I had 99% of my environment set before I got there or in the first 10 minutes: I already have a Git client (yay Tower!), and since I was up at 5:15 am AGAIN that morning, I went ahead & installed MAMP, then installed Drupal 8, so when I got to the sprint (early, see also 5:15 am) all I had to do was clone the main Drupal repo in Tower, install Chatzilla, and rediscover that my nickname was already registered. (Thankfully using an old default password.) I couldn’t get xdebug working, but got told that wasn’t strictly necessary.

But then…I don’t anything about applying patches. I don’t have object-oriented programming experience. I do some PHP, and I can usually work my way through it, but I’ve realized that in the Drupal parlance I’m a site builder, not so much a developer. And I was in a giant conference room with probably hundreds of other people, and really really really nervous about doing something wrong.

What I probably should’ve done was find the documentation team and offer my wordsmithing skills.

What I actually did wasn’t too bad, though. I found a bug in Views tagged “novice” that actually just needed someone to see if a bug was still a bug. So I walked through the steps — which also meant creating a new content type and fields in Drupal 8, and getting the hang of a few different parts of Views that I haven’t used much — determined that the thing seemed to work as intended, and posted some screenshots to the issue. I’m pretty proud of that, TBH. It was an environment that managed to simultaneously trip both my “student ahead of the class” thing and my “fake programmer” anxiety. And yet I did a thing, on what was marked as a Major bug.

After all that, and a little lunch, I was all the brain fry (did I mention 5:15 am?), so I did a little work-related stuff. But before I took off, I went to one of the organizers to say thank you. Because whatever was not quite right for me, it was pretty amazing that all of this was happening. She asked what I’d done, and I explained it, and she wanted me to get up and say something. Which OMG NOPE. Still, she gave me a couple of stickers, which ok, I’ll take that.

With that, me and my luggage were off away from DrupalCon. It was a good week. I feel like I managed to balance on-time and downtime, meeting people, catching up with people, seeing a few sights, and knowing where my boundaries were. I only went to two parties, and that was PLENTY. (It’s kind of reminding me of some SxSW experiences, and I feel ambivalent about that. I really wish there had been a game night or something. If I go again in the future, I WILL organize one this time.) I learned a bunch, and also created a bunch of mental bookmarks: things I need to come back to and explore in more depth on my own.

D&D & Drupal

list of urls, query path to scrape data, migrate for pathfinder!

migrate & spotify

marvel character database, star wars character api

D8 has symphony’s DOM parser

3.5 SRD MySQL data dump

Pathfinder has it as CSV.

isotope module (for hiding & showing?)

wacky dice stories.


[context: at some point Google Site Search is probably going to be prohibitively expensive. is this a good alternative?]

Lucene is an underlying search technology, from Apache, in Java. Same as Solr.

big diff from Solr is that it’s clusterable? is that something that matters for us? I’ve no idea.

also, I have no goddamn idea what they’re talking about now.

“if you just want to search all your nodes, use Solr” ok then. (esp better integration with Search API?) so there’s that, I guess that’s the super-short answer to my question. not that there’s not super-neato stuff in Elasticsearch, and who knows, maybe by the time we price out of Google Site Search (does _that_ work with Search API) it’ll be ready for prime time.


Someone from Evergreen should go to Write the Docs?

(these are notes on phone.)

I think there’s a lot to bring from this to documentation of things at work. The idea of curation, pruning, caretaking of articles. Everything has an owner/editor, at least.

Edit as pull request.

flavors of svg

I’m in this one for Justin, mostly, and to see what modules and things we should be going with.

svg is just ridic cool.

inline svg — could that be added via embed entity (or whatever that’s called) — I hope I’m not getting too far ahead. TBH, this part is all about all the different methods of adding SVG to a page, nothing yet about Drupal specifically. (something called svg injector?) And I’m pretty sure Justin already has a preferred inclusion method. If I were on a regular computer, I’d go look at how the SVG for the logo is included.

oh, the media queries for showing different versions based on the sizing of the bounding box is quite fascinating. I wonder if he’s thought of that.

[look up jolly icons. those are cute.]


oh wait he did like a whole day workshop on svg, yes? we’ll  have to compare notes.

also grumpicon (best for one-time use)

this svg sprites thing is what we were trying to do in Cascade?

I’m so glad I got to test-read Shelley’s book all those years ago. it gave me this great view into what was possible.

you can ctrl+c SVG from illustrator into a text editor.

still no Drupal mentions, except a quick mention of some navigation in a .tpl as markup, which I don’t think is what we’re after.

revisions everywhere

undoing delete?

problems with concurrent editing

the issue of distributed content generation


demo of a module for D8 — I’ve already forgotten the name. [nb: Multiversion]

tree showing split “revision tree” when editing in two sites that are supposed to be in sync.

[so is there a D7 contrib module that works like Cascade’s Recycle Bin? because we use that all the time.]

the speaker’s notable modules are UUID & Deploy? which I should probably look at both.

oh wow this gets complicated fast. and this might be going over my head a bit. :\

some in-session researching:

how does Deploy play nicely with Workbench Moderation? Does it?

for Recycle Bin like functionality: see Killfile?

this is all quite high-end (the advanced level was quite correct) but it’s fascinating.

{oh I just heard Greg.}

this question: ow my brain.

forward revisions? O.o — oh, like Scheduler, or Draft in Workbench Moderation.

Right now I feel like I’m gonna need a week with a whiteboard and a brand-new Drupal installation.

using grunt with drupal

I’m not 100% sure if this applies, but it’s intriguing since Justin is already using Grunt for CSS & JS tasks (compiling SASS, minifiying, etc).

I really wish DrupalCon were set up like Confab in one very specific way: I like having a table to set my tablet/notebook on for typing. The height of these chairs combined with the height of my legs means that it’s quite awkward typing on my lap, the whole thing wants very much to go head over heels and fall off of my lap entirely. Then add in the knitting, which is mostly helpful in listening, but it’s hard to have ALL THE THINGS in my lap simultaneously. Confab’s setup with rows of tables was so much nicer than just rows of chairs w/out tables.

what about testing that’s more than just “clicking around” (I guess that’s the point of something like behat?)

what ARE D’s tools for deployment?

Grunt Drupal Tasks

“it has opinions”

Gadget “easy to start a Drupal project with Grunt Drupal Tasks”

feels very rabbit-hole. I’m not quite sure I get what’s going on, but maybe… actually, Justin might like that for that other thing he’s thinking about….

actually, I think I’m going to try this, if I can get Justin to help me get Grunt & Yeoman set up.

but omg sleeeeeeepy.

this might not be related, but does Hobsons have an API that we could connect with for building forms?

separate Grunt file for the theme — sort of like what Justin does now?!

again: what would it be like for the web team to be able to do our own theme deployment? or module deployment specifically for Features-generated modules (ie, content types with associated views, etc)?

the guy sitting in front of me has the USDA foreign agriculture service website in his browser, but with the url of localhost. so that’s a thing.

lean collaborative usability

I don’t have notes for the keynote, but I do have thoughts. more later?

omg this room is too hot.

nice slide of options for testing

better screening helps with buy-in. — I would also love to know what her screening questions look like. this whole slide about automating scheduling, Susan should look at.

I want Susan & Jamie to watch this presentation.

remote unmoderated testing – half desktop, half mobile!

9 page memo as the results

short memo means people read the damn thing, also could turn issues into tickets

she’s big on Optimal Workshop, which I’ve been thinking about off & on

Notable, by Zurb

lean ux book: proto-personas. behavior-based over demographics.

* * *

It’s later; after lunch later, and the session I thought I wanted was totally not what I wanted, so I’ve got a little time to sit & think & write. So about that keynote: my personal reaction to it is pretty positive, for the most part. I had tiny quibbles with some of the science-y details, and I wish her recommendations for building empathy, influence, and adaptability had been as specific and actionable as the ones she gave for awareness. But what I think she heard from Twitter, and I think what I even overheard from someone in the exhibit hall, was a disconnect with the audience. Which is not to say that this stuff isn’t important for Drupal people, for developers, whatever — it absolutely is! (“unfortunately, as you probably already know, people.”) But I would have liked to see her be more explicit about WIIFM and more explicit of a connection with the Drupal community, issues in the Drupal community. I think there are ways to do that. (Plus also unspoken gender boundary, emotion as women’s work, etc. I don’t know it’s better to tackle that issue directly or to glide past it silently, but oh boy is it there.) So that’s what I’m thinking about that.

avoiding the git of despair

Features, Git, & Drush!

I’m wondering if I really want more command-line-ish access, or if Justin does. Especially with committing theme changes.

Also, I want to do some sketching about what the ideal flow looks like for theme, module, structure, and content.

I’m not making a lot of notes in this one because there’s SO MUCH content in the slides, and TBH I want to walk through it with my actual test site. Plus that thing about wanting to sketch. I’m having thoughts about overall process as well.

I want all the more resources, except maaaaaybe the videos?

(plus also I need to watch The Princess Bride again.)

site building like an engineer

the architecture, and complexity, of a drupal site is based on module selection.

nice runthru of heuristics for module selection.

special cases

panels (“is it done the drupal way or the panels way?” and this is why I’m always skeptical)
display suite
“I have this layer of complexity…do I use it?”
domain access
organic groups
rules (turing complete, apparently. “infinite complexity”)

avoiding building the homer-mobile.

data structures

[I don’t know if he’s talking about content types, but I’m feeling like I kinda need a full do-over of the Fields of Study content type]

ok, yes.

“every time someone tries to save a week’s worth of work on the building side, they create a month’s worth of work on the maintainence side”

site building smells

taxonomy: number of vocabularies, general size/complexity of vocabs

content types: number of types, overloaded with fields, do the lines between content types make sense, would taxonomy, entities, etc be better?

in all of this: is there a reason for this to be here?


use the description field on the content type, taxonomy/vocab, menus, blocks

in views there’s a place to write comments about the view itself! “this display’s comments”

only add  complexity deliberately
name things carefully
comment liberally { every time you see a textarea just start typing }

that was really good, lots to think about. (and short!)