Getting Things Started
Table of Contents
The Preamble
Hello and welcome to the first edition of this newsletter! Your invaluable support helps me offset development costs, so I can take care of menial tasks like “producing the documentation needed to ensure this project can succeed”.
Going quick, cause I have to get back to f i n a l l y
coding Boba.
Keeping up with Ms. Boba
As you very well know, we’re recently announced Realms, also upping our support channels to 3 (newsletter, store and RobinBoob). This effort took more than a month.
I can’t wait to code again.
Before that, however, I had a couple things to tie up:
The Newsletter Survey
We have a “topic of the month” survey to figure out what my target audience (you) wants. Of course, I’m not going to do this manually. Instead, I set up a whole fancy pipeline to automatically take care of the logistics.
One step further: I documented it all on my own technical blog, dug up from its grave for the occasion. May this be the first of many documentations to come!
Privacy Concerns
As someone warned me (thank you!), the blogging software I’m using is taking member names from Stripe despite me having disabled them completely in the signup. I’ve opened a support ticket to get it resolved. In the meantime, as you can read there, I did some code magic and hacked away the behavior.
If this is a concern for you, you can remove your name from the account settings page. Sorry about the issue! I do have a rant about how needlessly hard it is to not have other people’s names leaked to you when running an online business, which might find its way here at some point.
What’s next for Boba
Now that I can go back to what people are supposedly paying me for, I need to resist the urge to just code code code. No one more than me wants to put their head down and whip up Realms in a month, but that will just make it harder and harder for others to help me out in the future. If we go too quickly now, we’ll pay the price later.
So here’s a rundown of the focus areas I will need to untangle and fit into a ~coherent plan~:
-
Performance: we definitely aren’t where I want us to be, especially in longer threads. While some of it will be solved on its own once React Concurrent Mode launches, there’s some work that needs to be done regardless (e.g. lazy-loading of embeds, serving smaller images, swapping gifs for videos*).
-
Quality of Life Improvements: somewhat related to performance, but not only. For example, there’s definitely something going on with random 500 errors. Of course, I can look at the logs and solve these (and I will), but I really need better instrumentation so I don’t have to wait for people to complain before I act. This is likely to be one of the first things I’ll focus on in the coming week.
-
Design Work: “design twice/code once” sort of efforts. For example, there’s a lot we need to fit in those thread/board sidebars, and I’d rather have a path charted before I start incrementally coding improvements. How do we collaboratively work on the final result? How do I empower people to do this work without me? That’s part of the challenge that needs solving.
-
Documentation: stabilizing Boba for Realms is a perfect occasion to write documentation and help volunteers do meaningful, impactful work autonomously. And documentation is also the perfect chance to standardize conventions and kick down some technical debt. For example, I’m looking into adding OpenAPI-based specs to Boba’s backend endpoints.
-
Current Hotspots: working on more Realms doesn’t mean the single ones don’t also need some love: from better support for deep threads, to search, to crossposting, to accessibility, to… a lot of small things, “the work must go on”.
-
Realms, of course: there’s a couple different ways I’m planning to approach this. In the short term:
- Admin Tools: Realm admins will need to make new boards, new roles, new
identities… all things I’ve been doing directly through the DB. No more:
it’s time to get us some nice UI.
[BobaLore: the 4/20 “decoration settings” page was prioritized knowing we’d need the work for Realms]. - API work: All our server requests have, until now, not carried any information about the Realm they originated from (understandably, since there was only one). I’ll need to go through each of them and add this information, which is also a great chance for the documentation work highlighted above.
- Backend Modularization: does this user have permission to make this post? Which set of identities should we select from when assigning a new one? These checks, which are already happening, will be based on more and more complex logic as time goes on. Some good architecture design done now, will save pain later.
These aside, the design work mentioned above will also feature some realms-based effort, like the “eagle eye view”/“welcome page”/“rules explanation” flow & screens.
- Admin Tools: Realm admins will need to make new boards, new roles, new
identities… all things I’ve been doing directly through the DB. No more:
it’s time to get us some nice UI.
-
Code Candy: those things that make my code worth coding even though they might not be the “highest priority”. Right now, “emojis”, “user badges”, “realm user pages”.
…and that’s… a lot. Please do let me know if you have thoughts, either with the usual anonymous form, or by emailing me, or on Discord, or in any way you can think of. If you’re interested in more (or less) details, you can let me know through the survey. wink wink
From Ms. Boba’s Calendar
I’m going to have a bunch of meetings with prospect volunteers, stakeholders, and random people in the following weeks, but here’s the coolest one: I’ve been introduced to Zulip’s CEO, and will have the chance to ask him about how to foster a good open source collaboration environment (especially in these early stages). From their main page:
Zulip has a significantly larger and more active development community than other modern open source group chat solutions like Mattermost, Rocket.Chat, and matrix.org.
What’s the secret? I don’t know. I’ll report back with thoughts.
Final Notes
…and with this, we’re done for the first super quick edition! If you subscribed to this newsletter, you should have received the survey by now. If you haven’t, let me know!
See you soon!
* Re: gifs vs videos
I expect people to have a knee-jerk “no!! I hate videos in place of gifs!” reaction. I’m right there with you. But there’s 2 main reasons this is needed:
- Gifs are really much, much slower to load than videos. Any web performance guide hammers this point down.
- Gifs can’t be paused, and I need to be able to pause gifs for accessibility reasons.
Personally, I find not being able to save videos as gifs frustrating, so that’s one thing I’m going to make sure can still happen. If you have other particular frustrations, do voice them and I’ll look into it. This change isn’t going to happen overnight anyway, cause it’s a non-trivial addition to the Boba backend architecture.
Note: Who has access to this?
[UPDATE]: This is now publicly available!
For these first editions, I’m currently surveying and giving immediate content access to both paying and free members (but not the public, which will likely get it ~3-4 weeks after release). Paid members will still get “stronger dibs” on the topics. I know who you are.
In the future, this will likely change. I’m also looking at other offerings that might entice people to give some extra support. All this aside, the most important offering is “Boba keeps getting worked on and Ms. Boba cam afford to hire help so she doesn’t crash and burn”. If you can, consider becoming a paid member. If you can’t, consider letting me know what’s stopping you.