Let user edit gif once recorded

[quentin-st]

As said in #3, I'm using ScreenToGif on a windows environment. It lets one edit the gif once recorded:

It is a really cool feature since you can delete the X first/last frames (we could even eventually drop the "delay in seconds before recording starts" setting), add transitions, crop the gif...

Transitions & gif cropping may be useless, but frames deleting is a must have. What do you think?

[lode]

I'm used to SilentCast (https://github.com/colinkeenan/silentcast) which allows the same although with a more lean interface: you get a Nautilus window with the images and you can i.e. delete files before they are converted to gif.

(I don't like the messy interface of SilentCast though, Peek is way nicer! 👍)

[quentin-st]

Nice one ideed! (but messy indeed, just like you said)
I'd rather have a ScreenToGif interface though - an horizontal list of frames: when scrolling, you get a nice preview of what the gif will look like. Plus, you can fit contextual information (frame duration & index) easier

[phw]

Nice idea, like it. The trick will be to make that as lean and simple as the rest of the interface. But I definitely think that would be a nice enhancement.

I will however not be able to tackle this anytime soon, a few other things that need to be taken care of first.

[lode]

Personally I wouldn't find it too messy to have a button to open Nautilus as a first step. The interface would stay clean and you get another known interface. Indeed as you say, it takes quite some time before you have such an interface remade in Peek itself (to remove, copy, reorder, etc.)

[phw]

I also like the Nautilus solution for its simplicity. One thing to keep in mind us that currently the Gif is created by imagemagick, which does remove redundant information from the frames. So the editing view should be shown before the final Gif is generated.

It would be best to tackle #1 before dealing with this.

[phw]

The upcoming Peek 1.2 will extract all frames as PNG before processing it. This makes it easier to hook into this and implement this functionality. Probably it could just be added as a special post processor (which opens a new view).