Keepassdx + Heliboard: both excellent apps in their own right that create a smoother experience of mundane phone use, but they also integrate rather well together and Heliboard will often pop Keepassdx in to its suggestion bar when you enter a log in page. It’s been really nice for me.
FUTO voice input: speech to text for those who don’t want to use Google speech services. Frankly, the FUTO app works better than googles app anyways, it always handles grammar correctly as long as you speak relatively clearly, and integrates with Heliboard nicely.
Tailscale: for those who need VPN access to their other devices.
Thunder : a Lemmy client with compatibility with Lemmy’s recent server side changes and also has a decent UI/UX
tasks.org: fantastic, customizable to do app with various syncing options.
Magic Earth: privacy respecting maps/directions for those who don’t want google maps. (NOTE: closed source. Here is the privacy policy, terms of use and description of their business model at the bottom of their FAQ)
I have been using FlorisBoard for a while now, and had no idea about Heliboard! I saw the rename but didn’t really pay any mind to what the app was. I liked Florisboard well enough but it was laggy and unresponsive with no predictions, so I made a ton of typos
That was exactly my experience with Florisboard too. It has tons of potential, but I couldn’t ever quite feel fluid with my typing on it. It was a relief to see Openboard get forked.
I’m not super sharp on the legalese of licensing so I may be missing something important, but the license for the voice input app seems to be pretty straightforward to me. It does retain more control over the source code than true open source licenses generally do, but that doesn’t bother me too much personally, provided the respect for privacy and user rights is there. I live more on the world of privacy policies and TOS since I don’t code anything beyond a few shell scripts on my desktop.
I should have made a note on my list that both FUTO voice and Magic Earth are not fully FLOSS, just privacy respecting. I’m not surprised there has been some drama though with it not being fully FLOSS. Along with that, one of the companies employees, Louis Rossman, is a YouTuber and has somewhat of a polarizing personality 😄.
I think overall I do generally get a good feeling from the company and the way this app has been functioning. The privacy policy is just about as clean as I think you could get it a speech to text app. Mostly though, it’s a lesser of two evils for me, as this is really the only option aside from Google speech to text that I’ve found. It fills a pretty niche spot on mobile devices right now.
Sorry I missed this earlier. And thanks for the update. I think we’re similar in our thought process. I appreciate your input. Also, Rossman is almost enough to make me walk the opposite direction. Almost. :)
That’s correct. I didn’t notice the c/opensource that this post came from, disregard that entry if you’re a %100 FLOSS user. I’m still happy to recommend it in general though. It’s privacy respecting as far as I am able to verify by their policies and TOS, and it has been a huge quality of life improvement for me on mobile. Hopefully organic maps or other mobile clients for OsmAnd continue to progress 🍻
I love tasks.org on Android but God I wish they’d add some NLP to parse out times. That’s the one thing I miss from other apps. IIRC The issue in GitHub basically says that there isn’t an easy library to do it with right now.
It’s so annoying to me that I’m considering trying to make a server side app to do this. Just pull the caldav and parse the titles and set the due date. Can’t be that hard right?
Haha, you’re over my head with all that. Is there some issues with setting time and date for reminders and due dates?
That is one of the things I like about Tasks.org (and really anything that interacts with CalDAV, messy though it can be), you could probably write some back end server element to interface with it and it would accept it just fine. It really gets me when I find a polished, well designed app… that will only work with the companies proprietary, hosted back end 🥲
Oh I gotcha, yeah that would be a really nice improvement. If you do end up writing something up, think about making it available in github and post about it here! I don’t know how much work that would take compared to a personal project though.
Interesting, I’ll have to check out the CPU usage on my server. I’ve been running it for months now and haven’t noticed anything, but it’s worth a second look.
I noticed because my server is a decently old Dell Optiplex I bought off Craigslist next to my TV and the fans went crazy every time usage spiked lol. There weren’t even any active connections yet.
Ah that makes sense, my server is a mini PC that has a quiet fan. I probably wouldn’t notice if it was popping on throughout the day.
I did a quick top check and it seems like the tailscaled service is pretty consistent at %0.7 CPU usage, and %0.5 memory usage. It seemed to stay consistent when I initiated a remote connection via ssh as well. I’m going to continue testing it though, I don’t like taxing my system. Ain’t got the money to replace it, so it’s gotta survive for as long as possible 😄
Interesting, it just started working automatically for me. I dont know exactly what combo of settings it might be, but try:
Go to Heliboard settings > Text Correction > suggestions. Then toggle on “show correction suggestions,” “personalized suggestion” and “next word suggestions.” Those are the ones I have on.
Go to Keepassdx setting > Form Filling. Make sure “set default autofill service” is toggled on. Then go in to “autofill settings” and toggle on “inline suggestions.” I think this is probably the issue.
Make sure Keepassdx is set as your default autofill service in android settings Settings > Passwords and accounts. You should see Keepassdx listed there as the form autofilling option.
+1 for Heliboard and KeepassDX. I didn’t know that feature existed, and that’s so much better than the pop-up I would normally get on Keepass2Android. Switched and happy
Try the package html2text. I haven’t used it before, but I think it might be what you’re looking for. Interestingly, it was originally developed by Aaron Schwartz, one of the co-founders of markdown.
Other than that I’ve seen people being pretty successful with a combination of wget and pandoc. I don’t know how well that would work in scripts though.
Wow, most of these were new to me. I’m testing them for a few minutes and so far they’re working great. I’ve been avoiding Google’s voice to text for a while, and hadn’t found a good maps app either.
Seconding Newpipe, excellent app.
EDIT: added links to the terms, policies and FAQ of Magic Earth, as it is not open source.
I have been using FlorisBoard for a while now, and had no idea about Heliboard! I saw the rename but didn’t really pay any mind to what the app was. I liked Florisboard well enough but it was laggy and unresponsive with no predictions, so I made a ton of typos
Will definitely be giving Heliboard a run
That was exactly my experience with Florisboard too. It has tons of potential, but I couldn’t ever quite feel fluid with my typing on it. It was a relief to see Openboard get forked.
Re FUTO: how do you feel about their license? I’ve seen some drama around it. Doesn’t give me the warm fuzzies either. Nice list, btw.
Thanks! Glad to share.
I’m not super sharp on the legalese of licensing so I may be missing something important, but the license for the voice input app seems to be pretty straightforward to me. It does retain more control over the source code than true open source licenses generally do, but that doesn’t bother me too much personally, provided the respect for privacy and user rights is there. I live more on the world of privacy policies and TOS since I don’t code anything beyond a few shell scripts on my desktop.
I should have made a note on my list that both FUTO voice and Magic Earth are not fully FLOSS, just privacy respecting. I’m not surprised there has been some drama though with it not being fully FLOSS. Along with that, one of the companies employees, Louis Rossman, is a YouTuber and has somewhat of a polarizing personality 😄.
I think overall I do generally get a good feeling from the company and the way this app has been functioning. The privacy policy is just about as clean as I think you could get it a speech to text app. Mostly though, it’s a lesser of two evils for me, as this is really the only option aside from Google speech to text that I’ve found. It fills a pretty niche spot on mobile devices right now.
Edit: checked out the FUTO temporary license
Sorry I missed this earlier. And thanks for the update. I think we’re similar in our thought process. I appreciate your input. Also, Rossman is almost enough to make me walk the opposite direction. Almost. :)
Magic Earth is closed source
That’s correct. I didn’t notice the c/opensource that this post came from, disregard that entry if you’re a %100 FLOSS user. I’m still happy to recommend it in general though. It’s privacy respecting as far as I am able to verify by their policies and TOS, and it has been a huge quality of life improvement for me on mobile. Hopefully organic maps or other mobile clients for OsmAnd continue to progress 🍻
I do agree. I wish they were open source
I love tasks.org on Android but God I wish they’d add some NLP to parse out times. That’s the one thing I miss from other apps. IIRC The issue in GitHub basically says that there isn’t an easy library to do it with right now.
It’s so annoying to me that I’m considering trying to make a server side app to do this. Just pull the caldav and parse the titles and set the due date. Can’t be that hard right?
Haha, you’re over my head with all that. Is there some issues with setting time and date for reminders and due dates?
That is one of the things I like about Tasks.org (and really anything that interacts with CalDAV, messy though it can be), you could probably write some back end server element to interface with it and it would accept it just fine. It really gets me when I find a polished, well designed app… that will only work with the companies proprietary, hosted back end 🥲
On ticktick you can make a task like “cook dinner at 6pm” and it will auto set the reminder to be at 6pm.
I’d love that for tasks.org but like you said I can maybe implement it sever side :)
See: https://blog.ticktick.com/2020/09/03/ticktick-my-productivity-app/#Natural_language_input_recognition
Oh I gotcha, yeah that would be a really nice improvement. If you do end up writing something up, think about making it available in github and post about it here! I don’t know how much work that would take compared to a personal project though.
I tried Tailscale instead of Zerotier on my Ubuntu server but it was causing crazy CPU spikes
Interesting, I’ll have to check out the CPU usage on my server. I’ve been running it for months now and haven’t noticed anything, but it’s worth a second look.
I noticed because my server is a decently old Dell Optiplex I bought off Craigslist next to my TV and the fans went crazy every time usage spiked lol. There weren’t even any active connections yet.
Ah that makes sense, my server is a mini PC that has a quiet fan. I probably wouldn’t notice if it was popping on throughout the day.
I did a quick top check and it seems like the tailscaled service is pretty consistent at %0.7 CPU usage, and %0.5 memory usage. It seemed to stay consistent when I initiated a remote connection via ssh as well. I’m going to continue testing it though, I don’t like taxing my system. Ain’t got the money to replace it, so it’s gotta survive for as long as possible 😄
I just installed Heliboard but nothing seem to have changed? I still need to switch to Magikeyboard for login anyway. Am I doing something wrong?
Interesting, it just started working automatically for me. I dont know exactly what combo of settings it might be, but try:
Let me know how it goes.
Thank you so much! I did all the steps and it started working :D This is so much more convenient than before
Awesome!
+1 for Heliboard and KeepassDX. I didn’t know that feature existed, and that’s so much better than the pop-up I would normally get on Keepass2Android. Switched and happy
Is there something like markdownr for standard Linux, usable in sripts?
Try the package html2text. I haven’t used it before, but I think it might be what you’re looking for. Interestingly, it was originally developed by Aaron Schwartz, one of the co-founders of markdown.
Other than that I’ve seen people being pretty successful with a combination of wget and pandoc. I don’t know how well that would work in scripts though.
It feels wrong that I didn’t know about Myne
Myne is pretty awesome. I love project gutenburg, but their interface is hard to navigate through without a wrapper like Myne.
Wow, most of these were new to me. I’m testing them for a few minutes and so far they’re working great. I’ve been avoiding Google’s voice to text for a while, and hadn’t found a good maps app either.
Nice! I’m glad to be able to pass them along, I hope they serve you well.