Update libarchive README and fix comment

compressed_archive/libarchive/README.md

@@ -0,0 +1,22 @@
+# libarchive Decompression Backend
+
+This backend utilizes the [libarchive](https://www.libarchive.org/) library to support a wide
+variety of compression formats.
+
+This backend is currently only supported on the Linux platform.
+
+## Using
+
+Enabling this backend is achieved by adding the `libarchive` qmake configuration value:
+
+    qmake CONFIG+=libarchive
+
+Upon success, the application can be built as normal.
+
+## Limitations
+
+ * libarchive has a stream-based architecture that does not (currently) offer random access.
+   In practice, this means that you can only seek forward and would have to re-open an archive
+   to read an entry before the current position. This doesn't seem to have a huge performance
+   hit and can be mitigated by creating properly sorted archives.
+ * 7z decompression is slow (but seems to be slightly faster than unarr)

compressed_archive/libarchive/README.txt

@@ -1,24 +0,0 @@
-* Introduction
-
-TODO
-
-* Using
-
-qmake CONFIG+=libarchive
-
-* Supported Archives
-
-TODO
-
-* Limitations
-
-TODO
-
-v4 solid archives, slow 7z, etc.
-
-libarchive is stream based so it reads the archive in linear order. talk about
-advantages of properly ordered archives (1.jpg..10.jpg vs 01.jpg..10.jpg)
-
-* Future Work
-
-fallback to unarr for v4 solid archives?

compressed_archive/libarchive/compressed_archive.cpp

@@ -73,7 +73,7 @@ bool CompressedArchive::archive_seek(quint32 index)
         return true;
     }
 
-    // libarchive uses a streaming architecture so we cannot read files after our current position.
+    // libarchive uses a streaming architecture so we cannot read files before our current position.
     // because of this, when we need to seek to an index before our current position,
     // we must reopen the archive.
     if (idx > index) {