Hi everyone. Hope you are well. BentoPDF recently hit 10k stars on Github in just under 3 months of launch and I am very grateful to the community! ❤️

BentoPDF’s new version has been released. And I had implemented some of the requested feature here such as: Digital Signing of PDFs and Validation along with Email to PDF support and Deskewing of PDF. I have attached the release note link with the post. Moreover the OCR feature now performs on par with OCRMyPDF.

The reason I am making this post is gain feedback on the existing features of Bento, but most importantly Bento is going to have a Desktop version soon. Initially it will be launched for Mac users. Bento is inherently fast, but browsers and wasm have limitations, and this aims to solve it with the use of native libraries and leverage the CPU for faster processing and handling of large files efficiently.

I want to know what is the feature you use the most or is there any feature you’d like to be done that existing PDF softwares don’t do well. I am happy for any feedback! Thank you (:

  • alam@lemmy.worldOP
    link
    fedilink
    English
    arrow-up
    4
    ·
    edit-2
    9 hours ago

    A few clarifications first, because there are some incorrect assumptions here. BentoPDF is not AI written. BentoPDF was built manually over months of active development, with continuous refactoring, and iteration and community contributions. In my previous role, I worked on building AI systems to solve specific problems, and it doesn’t have any direct relation to Bento. Reducing the work of contributors to that is frankly unfair.

    That said, it is a fast moving open source project, and not every feature is finished or polished to the same degree. You see, the bad part about building a software is that bugs are inevitable, even billion dollar companies have it, and I am but a solo dev. But the good part is that they can always be fixed.

    So Bugs and rough edges are expected at this stage, and actionable bug reports are genuinely useful there. So, If you’ve run into specific breakages, opening issues with details, or just telling what’s wrong and what can be improved, is the most effective way to improve things. Broad label like AI written don’t really help move the project forward, especially given the amount of real work behind it.

    But I do appreciate that you want a tool like this to exist and improve. Also, as regards to the Used By People Section, it was already fixed, along with a FOUC issue in simple mode, but it’s in our edge release, which I have mentioned in the very top of our release note. Thanks