Ririsoft

Tank you for your time and comments here! > 3. We could have `infer::mime_type` be a slice of possible mime types, with the first one being the most recent one...

Hi @paolobarbolini, The situation is clearly improving thanks to your commits. My understanding is that you are favoring [mime-db](https://docs.rs/mime-db) instead of [mime_guess](https://docs.rs/mime_guess). This is totally fine for me, since mime_guess...

> > > mime_guess is in passive maintenance mode. It's a tangent of a tangent of a side-project I haven't touched in years. Thanks a lot for the clarification, it...

Good catch for the image loop size 👍

Fixing this issue could help close #82 as well.

> Builder would be fine, in current situation header cannot be created for attachment, which is probably most common usecase for Content-Disposion. I am totally in line (pun not intended)...

Hello, My point is that the makefile is using a hardcoded path for the installation destination through its usage of apxs. This does not ease the work of distro packagers...

Hello @airween `./configure` works now, thank you very much. However I still cannot use `DESTDIR` or any other variable to customize the installation directory for `make install` in a distinct...

Hi, Using ``` DESTDIR=/foo/bar make install ``` or ``` make DESTDIR=/foo/bar install ``` Produces the same result and fails to install. This is because apxs is used during `make install`...

> All I could do that make a patch which replaces the "-i" in Makefile.am to "-c", then apxs doesn't want > to install, and debuild script can install to...