Status update, June 2018

A lot happened this month. I’m usually used to focus on a single project for extended amounts of time, but this month was different. While continuing to frequently review and send patches to my “big projects” (such as wlroots), I also started a bunch of little side projects. I’m going to try to sum up all of this in this post, I feel like stepping back could help driving me forward next month.

First, what’s up with wlroots? I’ve been working on a new type, wlr_buffer. It offloads some complexity off of wlr_surface, fixes some bugs that caused glitches and enables new use-cases. I won’t be able to go into details, because buffer and surface management on Wayland is complicated: you have different buffer types (shared memory in RAM vs. zero-copy buffers in the GPU) which have different semantics (some need to be uploaded to the GPU, with damage tracking support), and on top of that you also need to handle synchronization (which also depends on the buffer type) and cleanup (when the buffer is removed). Anyway, wlr_buffer allows for atomic layout and animations when closing windows (used by Wayfire). I’m still continuing this work and improving the wlr_surface API.

Another Big Thing is the export-dmabuf protocol. We’ve been working with atomnuker, who has a ffmpeg background, to design an efficient protocol to capture your screen. It’s using DMA-BUFs under the hood, which means that clients can directly read the screen contents without needing to copy anything. This will be useful for screen recording and screencasting, we already have a working — even if somewhat arcane — screen recorder (which is able to use hardware-accelerated video encoding!). I’m still working on another protocol to cover the cases where export-dmabuf either doesn’t work or is overkill.

I’ve also sent another iteration of the xdg-decoration protocol for inclusion in wayland-protocols. This protocol allows the compositor and clients to negociate server-side decorations for better system integration. Getting a protocol right takes an insane amount of time, even for a very simple one like this one!

Outside of the sway world, stuff happened too. vil has been working on mako, my Wayland notification daemon, to add criteria support. This means you can now style differently notifications depending on the app they come from, their urgency, or even their category.

A new contributor, freemountain, has started to work on my Twitter → Mastodon bridge project, emuarius. The goal of this project is to be able to follow Twitter accounts from Mastodon, and to be able to interact with them seamlessly. The “follow Twitter users from Mastodon” part isn’t too far from working, but still has bugs. It’s also using the legacy OStatus standard, we’ll eventually need to migrate to ActivityPub.

Now to the new projects. I’ve started mrsh, a new shell. I want to build a simple and minimal library for POSIX shells. Existing projects, even dash, aren’t cut for this job. My workflow so far has been to get a subset of each feature working, to prove that the approach is viable, and organically grow the feature set when the full chain is operational. So far I’ve got a subset of the parser/AST and a subset of the command handling working. My next focus will be job control and variable expansion. I’m not yet happy with the parser, I’ll try to make it nicer (the POSIX standard doesn’t help with that).

Another project I’ve been working on is emailthreads. My friend sircmpwn is working on sr.ht, an open-source collection of services to manage software projects. We’ve been discussing about the best way to submit patches and review them. We’ve been wondering if it’d be possible to design an easy-to-use interface for git-send-email. emailthreads’ job is to parse replies to emailed patches into a tree of comments to allow them to be displayed in a web interface. To my own disappointment, it has been working pretty well so far on real-world test cases. I still think it won’t ever be perfect because there are too many edge cases, but I’m looking forward to testing it on more discussions!

Last, I’ve bootstrapped wleird, a collection of weird Wayland clients that use obscure Wayland features in counter-intuitive or uncommon ways. Test clients are lacking when implementing the compositor side of Wayland protocols. There are only two test clients so far, one using subsurfaces and the other buffer positions, but they both don’t work (yet) on wlroots. We also have ideas for more weird clients!

What a busy month! And I’ve not even mentionned the few patches I’ve pushed or reviewed for matcha, slurp, grim and go-imap. Anyway, I’m looking forward to even more awesome stuff next month. If you’re interested in any of these projects, make sure to drop me an email or ping me on IRC!

Questions, comments? Please use my public inbox by sending a plain-text email to ~emersion/public-inbox@lists.sr.ht.

Articles from blogs I follow

Status update, May 2023

Unfortunately I had to take a little leave for personal reasons, so this update will be rather small. My main adventures last month were in Alpine packaging. My motivation was to set up a package repo for Anemos. Alpine's apk is pretty great, but I found …

via blogfehler!

Friday Updates

Team Updates A number of members of my team at Valve don’t blog, and I like to bring visibility to their work when I can. Here’s a quick roundup of what some of them have been doing, in no particular order.

via Mike Blumenkrantz


It kind of crept up on me. One day, sitting at my workstation, I stopped typing, stared blankly at the screen for a few seconds, and a switch flipped in my head. On the night of New Year’s Eve, my backpack was stolen from me on the train from Berlin to Amste…

via Drew DeVault's blog

Generated by openring