So I ported AwesomeTiles and TopHat Gnome extensions to Cinnamon Applets using Claude Fable and suprisingly they are working flawlessly

[x9sim9]

So moving from Gnome to Cinnamon I had to say goodbye to 2 of the my favourite Gnome Extensions or did I?

So out of curiosity I asked Claude Fable to convert them and was really surprised with just how good a job it did.

Both are working flawlessly for me in Cinnamon

https://github.com/x9sim9/cinnamon-tophat

https://github.com/x9sim9/cinnamon-awesome-tiles

I don't know enough about the original extensions to become a maintainer but would someone be interested in adding these into the cinnamon spices?

TopHat - Screenshots in Cinnamon

Awesome Tilest - Screenshots in Cinnamon

[Fantu]

I suggest you do a thorough search before creating something that already exists. It might not be useful. If something similar already exists but can be improved, it would be better to try to improve that.
Search shouldn't be done exclusively with LLMs, because it can easily become unreliable if the sites it needs to query aren't responding, if it relies on memory that's too old, etc. (half a year, a year, or even more).
Regarding this case, with applets, it's a good idea to look in the applet repository (some may be also externally).
If you intend to have the LLMs do something serious, you should download everything you need locally and have it search there for precise information. I've seen this to be the main difference. Having it search everything in its memory (of what it's been trained on), and online, you waste more time, you use many more tokens, you make more requests to sites that are often overloaded due to the LLMs (I think the Cinnamon applet server, for example, was also having problems recently), and you get worse results.
In the case of creating or modifying an applet, the essential things would be have local repositories of applet, Cinnamon and CJS, and perhaps others as well. This way you give the llm agent the ability to have correct and up-to-date information on how to create the applet or how to contribute to an existing one, also getting the exact functions it needs to use (which is why cinnamon, cjs and maybe other things will also be needed)
Then, if you really want to port something that doesn't exist, you could also use the Gnome applet repository (of what you want port). In your case, you've easily obtained something that works because you started from that (and Cinnamon was forked from Gnome), but providing the exact information about Cinnamon actually needs is certainly better.
Furthermore, another important thing: LLMs don't have intelligence, and that's up to the user. They're designed to always provide results, even if they're poor or incorrect, so it's up to the user to have or acquire at least some knowledge about what they want to do in order to be able to guide and correct the LLMs in the work they're asked to do.
I hope I've given you some useful information.
Regarding this specific case, please search the existing applets carefully to make sure there isn't something already identical or similar to what you need. It might not be necessary to make additional applets. If you want to improve existing applets, try to only propose useful and well-made things and avoid wasting reviewers' time. Unfortunately, with LLMs, it's easy to run the risk of many inexperienced people filling up submissions with poor or useless content, wasting reviewers'/maintainers' time, and thus damaging the projects.

[x9sim9]

Thanks for the detailed response, at this point I am not really looking to do anything serious these are just proof of concepts at best.

I have over the past 3 years tried pretty much every applet none of the existing applets work as well as TopHat does for resource monitoring, and all the window managers are drag and drop not snapping by keyboard commands like Awesome Tiles does. It terms of applets like gTile the gnome version has way more features that the cinnamon one does.

So this was just an experiment to see what I could achieve with 30 mins of my time.

I am a programmer by trade and have a lot of experience with the headaches that AI driven code can cause, and considering how much I was able to get done in just 30 minutes I suspect there will be a flood of AI generated applets and extensions flooding cinnamon over the next few years so its right to be cautious.

That being said I have a use case that maintainers simply don't have the resources to do... porting gnome extensions to cinnamon.

My hope is that someone will see the value in these ports and become a maintainer of them so we can all enjoy them in cinnamon, I have even contacted the original authors to see if they have any interest.

[Fantu]

In fact, similar applets already exist, like gtile that you mentioned, and for resource monitoring there are already multiple ones. Instead of continuing to add more similar ones, it is better to first try to improve the existing ones to have a few but better quality ones rather than many similar ones that are of poorer quality and in some cases not well maintained.
and if you make new ones or want to contribute it's better to do it well rather than something quick and raw and then immediately propose it for inclusion.
otherwise discussions like these end up being classified as "AI slop" and are simply ignored, and any issues and/or PRs in the relevant repositories, if not good, risk wasting the reviewers' time (the small Mint team is already overloaded) and/or are ignored.

[x9sim9]

In an ideal world you are definitely right less apps and higher quality is the way to go but if that was feasible then you would expect gTile to have the same feature set between Gnome and Cinnamon which it doesn't the gnome version has way more features.

I would argue that the main issue we struggle with is time and that many of the applets have been abandoned or at the very least have not had new features added because the maintainers simply don't have the time.

I could of spent several weeks of time adding these features painstakingly into the existing applets but I just don't have that time available and by the looks of it neither do the maintainers.

So between a choice of not having the features in the cinnamon ecosystem at all or having them in but with some duplication surely the 2nd option is better?

In terms of features TopHat is by far the best resource monitor extension in the gnome ecosystem and it has way more features than the current applets in cinnamon so surely this is a win for the ecosystem allowing users to have access to it?

These extensions were not vibe coded, they were built by maintainers for the gnome ecosystem, all I've done is ask AI to make them compatible with Gnome.

[Fantu]

Indeed, we keep adding similar ones, and we end up with many of them of inferior quality and poorly maintained.

I recently saw an example while searching for a project for a self-hosted Rustdesk server with some essential features for free. They've made a lot of them; I found almost a dozen (but there could be more), all built with AI, almost all by a single developer. NONE of them are good enough to be chosen for production use on hundreds of devices, requiring little time, low risk of bugs, and likely to survive and maintain long-term compatibility.
The most used one is managed by a person with no experience in programming or management and is therefore a continuous alpha/beta, many users report bugs and then through AI it fixes many of them but with the fact that it continues to add things at the same time it never has a stable enough version. And this is just one example.

Regarding some cases of things I need but the existing projects aren't enough.
For example, a NextCloud app I need at work (extract) was abandoned. I looked for an alternative and found one, but it didn't have what I needed.
Many people in the same situation would make another with LLM or at most fork it separately. However, I thought it best to see if the other (files-archive) was still supported. Since it was, I started contributing to it so that we have a better one, with an existing user base and more possibilities in the medium/long term.

An example of an important project improved by joining forces with a fork that had been differencied a lot and was working on itself is memtest86+, instead of 2 projects with a single developer now they are a single better project with 2 expert developers and a third who has joined who helps.