What is the status of a fix for this issue? Did SirPavlova's analysis bring us closer to a fix?
From the examples in:
https://orionfeedback.org/d/8598-pdf-viewer-is-changing-the-name-of-the-pdf/4
I notice that this issue is still present in Orion 1.0.4 with the example provided by JSiuDev in April 2025 though it is fixed for the examples I provided in 2024.
I still have the same issue with files shared with me on a regular basis that I unfortunately I can't post here for bug reproduction purposes -- they come from a repository hosted on AWS accessed from a link distributed in a personalized e-mail. Though Orion is my default browser on my laptop, I now copy and paste the links in that e-mail over to Brave to download those files with the correct filenames. Ideally, I could just let Orion open any link I trust in the e-mail without having to guess whether the filenames of any documents available for download will be preserved.
If it is not possible to fully fix this issue due to a WebKit and/or pdf.js issue, can someone highlight what is broken about them so the issue can be reported to them?
And/or create a workaround by offering Save from the context menu (i.e. right-click on the canvas) that follows all the RFC's for Content Disposition headers, URI encoding, etc to replace pdf.js's Download button. I see that Orion already offers Summarize Page and Translate with Kagi in that menu so it should be possible and presumably relatively simple compared to those to add an option to just save the file/PDF.