Media on URL

Robert Helling helling at atdotde.de
Wed Mar 30 04:45:33 PDT 2022


Hi Mark,

thank you very much for your comments. I agree that when it comes to linking media from the internets, there is a lot room for improvement. One reason that has not been improved, yet, is that it turns out that getting this right is not so easy...

> On 30. Mar 2022, at 00:56, Mark Stiebel via subsurface <subsurface at subsurface-divelog.org> wrote:
> 
> 
> For my case it would be great to integrate S3 as a media repository - but I don't agree with the philosophy of tight technology integration. Where does it stop?

Honestly, I don’t see that happen (even though, last week I bought a subscription to an S3 compatible cloud storage solution…) as I don’t think many of our users have that itch to scratch and getting the details right probably requires quite some work.

Let’s restrict for the time being to media files that can be fetched by a simple http(s) GET request.

My original idea was that people would have their pictures (and then in extension other media files) in some web based galleries and one could give Subsurface the URL to an html document which would then be parsed and all the image files referred to in the html would be downloaded. That turned out to be harder than expected since most picture hosting sites don’t have simple <img src=…> tags but rather have some elaborate java script around this that needs to be run to actually access the images.

Maybe, once we solved the issues with WebEngine for printing dive log pages, one could try to actually write the image source extraction in java script...

> 
> So instead..
> 
> - Ability to add multiple URLs before hitting the import button - avoids the time sync dialog for every image

That should be very easy, making those comma or newline separated.

> 
> - Dialog text is misleading. Both right-click menu and file input dialog imply you can add multiple images via URL.

Changing the menu text sounds cheap with respect to the previous point.

> 
> - Ways to bulk add images from a URL?
>     - import a text file with multiple image URLS to import as a batch?
>     - parse a directory listing HTML file from base URL to do same?
>     - command line option or tool to adding images to a dive from the command line?

The text file should be similar making the dialog accept more than one URL (possibly using copy and  paste). I would avoid the further indirection via reading an extra file.

> 
> - Image viewer for remote images appears to be broken. Thumbnails in the media tab and on the dive profile work, but double clicking doesn't appear to retrieve the file properly from the server (will need to dig further into this one)

Yes, this needs to be done.

> 
> - Mobile app doesn't show media. Makes sense for local media, but would be good to be able to show media if its stored on a publicly accessible URL

Local storage on mobile devices does not sound too promising for large picture libraries. And using the on device image storages is also quite OS dependent (read: at least hard if not impossible using just Qt). Also this sounds terribly hard to work across devices (and I don’t want to compute hashes of all media on my phone…).

> 
> - Printing. Currently this doesn't appear to support images. Is it something that's never been a consideration, or something that's in the not worth the effort bucket?

Really, I would have thought images on the profile would be printed with the profile. Need to check.

> 
> - Maybe I just need to rethink my workflow, and instead look at exporting dives to HTML - and customising the HTML to my needs. Is it worthwhile templating the HTML export like the print template is?

Hmm, that sounds like a somewhat bigger project but not impossible. Also adding images to the templating shouldn’t be too hard.

>    - Command line option / tool to trigger an export?
>    - Or do I just read the XML divelog, and parse it myself to automate any static HTML generation and uploading to a hosting site?

Again that sounds a bit indirect (and too involved for most of our users).

> 
> - What about video? Is it a consideration to somehow support videos hosted on the common video hosting sites? (YouTube, Vimeo?)

How would you envision that? Having links associated to dives that lead to these video hosters? Or retrieving thumbnails? Or some closer integration (I don’t see a video player widget showing youtube videos embedded in Subsurface in any realistic future…). Anything more than a simple url requires authentification to these services which can be brittle (we removed our integrated posting to facebook for this reason). So, that again is a more ambitious project...

> 
> .. and there was my hopefully at least somewhat coherent thought dump.

Thanks anyway. Will look into the easy things first.

Best
Robert

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20220330/94ead914/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Message signed with OpenPGP
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20220330/94ead914/attachment-0001.sig>


More information about the subsurface mailing list