Geektoolkit
patreon
Mid March Dev Update!
🕑 Added 2023-03-16 07:47:39 +0000 UTCQuick Mid march update:
The team is heads down debugging issues with the March Beta, taking feedback, and working through things. The video hang issue we feel is resolved, and I want to thank those that contributed data which helped us narrow this down! This includes:
Andrew Crider, Daren Westman, Norbert Weber and Matt Landon, and Jan Borghoff
The team was able to correlate the info provided to eliminate many of the possabilities that could've been the issue getting us to a resolution much quicker. The issue turned out to be something that was introduced back in a December change that was exacerbated by January changes. We believe we have this one figured out. Your contributions directly helped us and the project, and we appreciate you! If I forgot anyone apologies, we handle a ton of data between messages, posts, patreon discord and sometimes even email.
The overlays were broken...text overlays were unable to be updated for fonts/colors and such. I have that bug fixed which was our second blocker on the March release.
The last issues we are aware of are related to networking. The reality is that while each of us knows a bit about networking, none of us are linux networking experts. Taking out a networking stack that was written by another developer over years and trying to swap in our own we knew would be risky, but we also knew that it had to happen. It was the number one reported support issue we had, and using third party code limited what we were able to do. We feel strongly the new networking stack is the right way to go, and that the experience is the most flexible possible, and we're close to having the last of the bugs figured out. We're stopping all development on anything further till we get this locked down...it's a necessary discipline that the team needs to get things stable and out.
That being said, once it's out I'm going to take a week to document. I've been promising documentation for a while for things like the plugins, and haven't taken the time because we're always chasing 'one more feature'. It's time to slow down, get that done, then look at the bugs that are reported, the feature requests, and the roadmap and get things into a bit more order.
To be honest, this is very new to myself and the team. We are working hard to learn from our mistakes. I get excited and talked about a few features that were half done with the intent of finishing them, and got ahead of myself and where the project was. It is SO easy to do with a project like this which is not my day job, while having just moved houses, and while also having a daughter who I'm raising. That means that I need to balance my ambition with reality, and that is..well..something I'm learning. I can only promise that I'm working to learn, to be better, to improve, and figure out ways to continue to make this the best digital art frame ever.
I keep saying there's three of us on the team...but really there's over 200 of us..because the feedback that you provide, the issues, the successes, the ideas, sharing your builds, helping each other out in the discord...it all helps the community out. So thank you for being such a great community. When as a dev I have a rough day or am frustrated about a bug, it's comforting to know I've got a community that loves the project and is backing us up. So thankyou!
Back to testing and debugging (and learning!). Next update should be a beta with DNS issues fixed up so we can go back to getting this release out!
Joe, and on behalf of Bdlou and QWK.
Comments
Joe Farro
Do you mean like automating pulling images to the frame? If there’s a downloader or you can write one to pull the images, then you can download them to the image share and that would let you show them on the frame. Midjourney is rough because they have no private apis yet. They’re working on them but the only access I’m aware of is to the gallery via a web scraper. For stable diffusion you can copy over images to the network share on the frame. That’s how I get my SD images over today
Vectorr66
Hey team, It has been a while since I have setup a frame. I want to make one and mount it on my wall to keep running. I see you always hard at work but haven't kept up too much. I am curious since I personally have been using Stable diffusion on my local system what kind of repositories can we plug into? I mean I would love to be able to pick certain genres or something and have them scroll. Is that possible? I mean as far as possibly viewing mid-journey feeds or what not?
