raminf's dashboard

Came back and re-read my previous note.

In my zeal, I violated one of my own rules which is not to push a specific implementation path when giving feedback. As a developer, I prefer general suggestions so I can decide which way to go with it. So please ignore my rant about XSLT etc. :-)

Your app is helping simplify a process. Another part of the process that can use simplifying is the attribution part (for those wanting to keep in compliance with CC credits). Anything you can do to help reduce the PITA factor is a step forward in my book.

Thanks.

The sidecar would be awesome. As long as the data could be exported in a way that could be easily incorporated into the finished work.

I'd start with what needs to be generated and work my way backwards. Some common use-cases would be making web-pages, Keynote presentations, videos, or maybe app development. In some cases you'd want HTML links, in others a flat text file with userids that is turned into a credit roll at the end of a Final Cut video.

For maximum flexibility I'd suggest letting people choose which fields of metadata they want to export. For example the video people may only want userids but the web-page people would want thumbnail + HTML anchor tags.

Perhaps there could be presets for various representations (including plain text or CSV export) with an XSLT stylesheet to go with each one that those familiar with that world can modify to suit their needs. If you don't want to geek out on XSLT, Matt Gemmell's MGTemplateEngine is great for converting structured data into output in a human-friendly way. So they pick the output format hit 'go' and it's done, ready for use in their project. It doesn't matter what the underlying storage format would be, but I imagine XMP would give maximum flexibility.

Ultimately it would be nice to have 'projects' into which you can save all your material and could come back and browse during the course of the content production process. Once done you could export the metadata in one go. But of course, that's a much more involved feature. Best forget I said anything :-)

BTW, really enjoying using the app. Sometimes I just like to look at pictures to get inspiration for a design. This is totally scratching that itch.

I missed the 'Copy Attribution' button in Download Manager. Perhaps it could stay visible on the download page or even merit its own separate menu item. It didn't occur to me to run my mouse over the item. I just went straight for the download folder.

As it is, the attribution string is a good start but not as useful as it could be.

Here's my typical workflow:

- Let's say I'm building a video, website, or Keynote presentation. I try to find as many items as possible that might be useful for later use, so I enter a keyword. In Flickr, I'd make sure the proper CC flags were set -- in ViewFinder I just have toggle the license in the tool strip. Yay! That alone is a great improvement. Can't tell you how many times I've gone through the process and downloaded images only to realize I'd forgotten to toggle the "Commercial Use" flag in advanced search mode and had to start all over.

- I scan through the various photos and download the images I think I might use later in as large a format as I might need.

- I go about building my content. This is an iterative process and often involves sending the images through Photoshop for scaling, touching up, and cropping. The output may be in a different format too (depending on size requirements) which will most likely strip the IPTC metadata. I may use some of the downloads or I may realize I need more so I go back and search for some more photos.

- What information I save during this 'gathering' process is what will help me get from the picture back to the original Flickr page so I can later figure out who should be credited and how.

- At some point, I'm done and have to circle back and put together the attribution data for everything I've used.

- The process involves looking at the finished material, noticing a picture, then visually matching it against the source and hopefully being able to go from there back to the original attribution.

- After much trial and error, I ended up keeping a parallel folder next to the 'image source' folder. Every time I'd download an image I'd drag-drop the image webloc from the browser into that folder. When the project was done, I'd go back and reopen each and every one of those weblocs and see if it was a picture I actually ended up using.

- An alternate process involves using a Curio worksheet--i.e. something that lets me save text, links, and images. I'd drag-drop the link, the title, the user-id, and a small version of the image. It's tedious, but now I have everything I need.

Ideally, ViewFinder would download a thumbnail, a title, a link to the page, the userid of the photographer and a link back to their profile page for each image downloaded and would keep it somewhere near where the big picture is getting downloaded.

Really ideally would be if there could be a 'library,' 'paste-bin,' or 'lightbox' concept in which all working images for a given project are drag-dropped and from which the credit data could easily be exported in a way that would require no manual intervention (i.e. it could spit out the HTML code for an attribution link).

Hope this all makes sense. I'm enjoying using ViewFinder for my next project. I just hope it helps me streamline the part of the process that currently carries the most headache.

Really like the EventBox UI and concept. Lack of support for multiple accounts is the only thing stopping me from tossing out every other Twitter client and just sticking with this one. So, yes please -- multiple accounts, preferably across all services.