wakabayashi

Hi everyone, hello @datakick, @Smile and @Acer My premium membership just expired, which felt like the right moment to pause and reflect on the current state of the project. I’ve been here since day one and truly appreciate everything achieved over the years. thirty bees has been a solid foundation for a long time, but I am seriously concerned about its future. Official commits on GitHub have become rare; while community PRs are still coming in, they often seem to go unnoticed. There is a lack of transparent communication regarding the roadmap. This leads me to a point where I have to ask: Is it still worth building on the thirty bees core, or is the project effectively dead as an open-source endeavor? Technical Hurdles and Workarounds In my daily work, legacy issues in the core are slowing me down significantly. Address handling is cluttered, and features like multishipping (used by maybe 5% of merchants) make the code unnecessarily complex and bug-prone. A prime example is the "splitting order" issue that hits me every few months—a bug known in the PrestaShop community for 15 years. To keep the system extendable, I’ve developed a "best practice" over the last few months using classes like OrderDetailExtension or ProductExtension that share the ID key to manage new columns. It works, but it’s a lot of overhead that only makes sense if maintaining backward compatibility with the core actually provides long-term value. If the project is stagnating, it would be more efficient for me to drop compatibility and modify core files directly. Waiting for Features Over a year ago, I had an intensive talk with Petr about a credit system for customers. Since I need exactly that, I waited—but I’m still standing here without a solution. In recent weeks, I haven't been able to reach Petr at all. While there was a recent sign of life on GitHub, it’s not enough for professional planning. I would love to see a Version 2.0 that modernizes the system radically: A rigorous code rewrite (even if it breaks old modules). Support only for currently supported PHP versions and updates for components like Smarty. A Backoffice designed around merchant needs, not just a collection of controllers. Clean Code as AI Foundation: clean, unambiguous codebase is essential today. If the core is logically structured, any AI can easily generate high-quality modules. If the base is "spaghetti," the AI will only produce more spaghetti code. Conclusion Is this vision of a Version 2.0 shared by the team, and is it something being actively worked towards? If not, that is perfectly fine. But then I have reached the point where I will likely move in this direction alone and radically decouple my own codebase from the core. Best Regards Emanuel

Bro you are not a normal thirtybees user, you are a module seller 🤷‍♂️

I don't know the regulation and I am not affected by it. But in general I agree with @Yabber how it should be handled. A clean/standardized process, that is triggred by a button. Afterwards I would check if the order involves multiple products. If yes, I would ask the customer, if he wants to cancel the whole order or only some products. A contact form might work as the oldfashioned solution, but imo it has only drawbacks (for merchant and customer). I would only disagree with @Yabber, that cancellation and refund are totally different things. From a coding perspective, they are pretty strongly related. In both processes you have questions like: Needs the current_state (order) be changed? Is there a money refund involved? Is the whole order involved or only some products? In both cases restock might be needed. So in short I would say, cancellation is a simple case of returns.

Sure, there are some risks. But buying random modules on prestashop store is actually risky as well. Or downloading code from github. Isn't it even more risky? In my case it would just be a waste of not using AI. I have become a decent coder recent years, but some AI solution wear cleaner than I would have written it 😊

I never worked with claude yet. It seems to be the best model. But right now I can use Codex 5.3 with no limits in phpstorm. I only pay the 20$ plan. Codex 5.3 is very strong as well. Maybe there are "political" reason to leave OpenAI but pricing and quality aren't an issue for me right now 😅

Its completly sick, what you already achieved without writting code your own 😅

@vincentdenkspel amazing! 😏 Don't forget: be careful. I would always recommend to work with testshops at least. Also try to learn some basics (in case you don't know anything right now). But in general this so cool. Keep going! @30knees This was my first guess too. But we might be wrong. It could actually be, that thirty bees is a winner of this. We have an open architecture here. It's accessible for all AI tools. And the system itself was built for extensions/moduls.

It's a very interesting discussion here. I can understand both positions. It's really a chicken-egg game. But imo there is a huge game changer: AI. It has become way more simple and fast to write code. I am also not aware of the plans/roadmap of TB. But with the new AI tools, it's even possible for no coders to start modifying some stuff. Ofc it's always better, if you have some basic coding knowledge, otherwise you might mess things up. Even if you aren't brave enough to use AI yourself: I would guess, that prices for a custom module will come down a lot. @datakick what is your experience with AI these days? I would say it has speed up my developing work about 3-5 times. It's hard to tell, but it's for sure huge. The first time I have the feeling, that my todo-list may become shorter 🫣

Wtf. If this statement is true... 😵 Probably they are afraid of lawsuit against them.

Is this all information they gave? No technical details?

It's long ago, that I asked for this. But a few days ago, I finally used this solution and it works as described by @datakick 😏

I am not sure... Is this a bug report or an explanation? Do you have uploaded a thumb image for packets? Cause you are right, these are exactly the images we are talking about "thumb".

In categories you have two different types of images: \thumb is the preview image.

What does this mean? Unfortunately I didn't find time to participate a lot on github recently. Hopefully I can change it in 2026. I have so much open Todos 😑 But AI is helping to get faster at solving stuff with code. 😊

Yeah we received a few orders from chatgpt in the past. But the sales don't happen directly there. We just se the utm parameter when a customer clicked there.

My search console is still full of entries, cause I block the urls with robots.txt. I was now looking into the source code and I believe, there is a simple solution. Just install the module "tbdetectcrawler". It will stop adding this trigger urls to crawlers.

Either you haven't show us the important results or your problem is something different. Aren't there any numbers in red with high "ms" or "Mb"? Cause from the screenshot you show, nothing is slow.

You should go to the performance tab and switch profiling on. Then refresh the slow backoffice page and scroll down. There you will have the information what is loading slowly. You can also post a screenshot here, if you have difficulties to interpretate the information. IMPORTANT: after quick test you need to deactivate this setting asap. Cause this profiling information is also loading in your FrontOffice shop. So your customer would see it as well.

Thanks! Github search sucks. I search there for "view" it doesn't show the "blockviewd" repository... I looked only 5 minutes in the code, but it looks sensible. I see, that tb is reading the cookie and afterwards loading the needed products.

The thread in prestashop forum is quite interesting. I am now with 50k of products and notice here and there some performance issues. But these are almost always issues on my own code. With 3.4 million products you will find even much more such stuff. But these issues are often not relevant, if you have only 1k products. In my expierence the perfomance issues came to 90% from modules and maybe 10% from bad hardware or missconfiguration. What is the name of the corresponding thirtybees module?

The core is far away of handling this in a clean way. It's not even clear when the invoice will be sent, as this is coupled with order states. In addition "invoice" isn't exactly the same in each country. That makes it very hard for the core to find a "fits-all-solution". For a single shop it's much easier to find a solution with a few overrides. But if you need such customizations you need php knowledge. An additional column with a dirty Mail::send() override might be enough.

Are the files on your server in the folder? I would surely try deleting all caches. But the fact that /themes/niara/js/vendor/picturefill.min.js is not found, sounds like the file is missing.

Sorry, I didn't write my question clear enough. With hypothetical I meant something different. I wanted to ask, if you have some specific Dev in mind? Cause for me it's not so easy to find someone similair to @datakick. Of course there are some Prestashop devs out there, who would like to get the job. But I need someone who is really a powerful dev, that writes clean, maintanable and secure code.

This is just hypothetical, right? Or do you have any powerful prestashop/thirtybees in mind? Cause my company is steadily growing and I am considering hiring a full time dev. If I can find a suitable person. He would ofc also contribute to the core.

I believe the project is as death as the last 5 times, this was asked 😉 But you are right, @datakick is absent for some time now. I guess it's just holidays/recovery 😊