Who can share a video demo of Glide OS in action?
I’m keen to see what’s new!
I’d like to get a preview too. I thought they we’re inviting the whole waitlist today?
I built this with GlideOs from scratch. I did have a website to give it a starting point and understand what I wanted vs typing it all out. There were somethings where I had to ask it several times before I got the desired results, which burned unnecessary credits. Also I found some easy things I wish I could do myself but had to ask that also burned through credits. All in all, it cost me about $50 in credits and a few hours to create. My original glide app similar to this took me a lot more time and i personally think this one works better.
Hi Joe, thanks for sharing. Nice app… I see this find-a-recipe topic is becoming quite a craze. The successful ones seem to start with a meal budget and include a lot of images. Have you thought about that? How do you feel about the design? With respect, I believe a Vibe coded version using ChatGPT would also look beautiful whereas the Glide App looks more functional (apologies for being direct).
I’m interested to learn 2 things:
- That app could be done quite easily in Glide Classic although admittedly it would take a few hours. A tricky feature would be to generate spontaneous image options for dishes. Something LLM’s can do easily but it’s difficult to integrate into classic Glide without a lot of external wiring. Can OS do this on the fly?
- I would like to see a demo of someone using OS to develop an OS Glide App… a classroom demo so to speak.
Lastly, this does not seem great value… 50$ for a basic app.
Hey Simon, thanks for the comments. A few things I’ll add related to Glide OS and not the app itself.
- it cost about $50 in credits to build, but I have no idea how much it will cost to use. It could just cost $25 a month on the basic plan (which allows you to have other projects or apps as well.) you could also spend $1000 to build and only pay $25-50 a month to run, no idea.
- I think I saw in a separate post where David said the credits are charged at cost. Whether building this app with a different program would have used a lower number of credits to get the same result, I am not sure of.
- For people who arent strong in databases, this app required zero knowledge of all the column types such as ITE, Relations, Roll ups etc. it handled everything for you. Could be a major plus for some users and not so much for others.
- Design is much easier and requires zero CSS knowledge. For a simple example, I wanted a button to be sticky at the bottom of the screen. This was able to do it in 1 minute vs trying to learn CSS and apply it to the right class.
- My biggest pet peeve is burning credits doing something I could do myself and burning credits when the result isnt what I wanted and it keeps having to redo it. What might have cost 2 credits to do right the first time ultimately cost 10 credits. This isnt reiterating a component, but just not doing it right.
Regarding the app idea, it was designed go be simple and requires less thinking to combat dinner decision fatigue. Choose your mood, foods you dont want (easier to eliminate than say yes to) and it pops out a simple dinner recommendation. You get 3 choices. It then generates the full recipe where you can print/pdf it. Side note, the pdf setup on GlideOS just required a prompt like “create a button that generates a PDF of the recipe” and bam it was done. No PDF integration and setting up templates etc. that was impressive.
Hi Joe,
Lots of great points. Thank you. I particularly like the PDF description. That sounds like a game changer.
I guess I’m used to vibe coding away from Glide and using the custom AI component inside Glide. Both experiences tell me it’s easy to burn through credits trying to get what you need. That’s fine if you are working on a company budget, but when it’s your own cash this is going to make for nervous programming.
Another thing you could do design wise is take a S/S of your current Glide OS build (first pic), feed it into a different design AI app and ask to improve the design. Once you have that settled, upload the s/s to Glide OS and make it use that design (see second pic). Took about 15 minutes to go from 1st to 2nd. With original Glide, this redesign would have taken me days.
For the resign, I used Claude Design with free credits. So having GlideOS resdesign the app prob cost like $1-2 in credits. Well worth it imo.
Love this. So are you saying we could regenerate key aspects of an existing app by feeding screen shots and the results are likely to at least as good if not better AND you have the benefit of being able to further develop new screens and features using OS?
Promises a lot. Seems daunting but impressive if possible. Were you able to confirm that tricky steps were regenerated perfectly - like on-boarding screens that trigger workflows that add rows to helper tables that are then used to manage row owner rights?
Still doing testing but I did ask it to apply row owners so users can only see their data and allow multiple users to use the app at once, which it created unique session IDs. No need for helper tables anymore.