summaryrefslogtreecommitdiff
path: root/doc
diff options
context:
space:
mode:
Diffstat (limited to 'doc')
-rw-r--r--doc/todo/for_amazon_s3_pre-gzip-encode_safe_files.mdwn17
1 files changed, 17 insertions, 0 deletions
diff --git a/doc/todo/for_amazon_s3_pre-gzip-encode_safe_files.mdwn b/doc/todo/for_amazon_s3_pre-gzip-encode_safe_files.mdwn
new file mode 100644
index 000000000..c0c0c12ab
--- /dev/null
+++ b/doc/todo/for_amazon_s3_pre-gzip-encode_safe_files.mdwn
@@ -0,0 +1,17 @@
+Regarding the [[Amazon_S3_Plugin|plugins/amazon_s3]]:
+
+Amazon S3 doesn't seem to support automatic GZIP encoding content (such as HTML, JavaScript, and CSS) that might be compressed by a full-capability webserver. (I'll also note that NearlyFreeSpeech.NET doesn't support compressing out-going files on-the-fly). However, Amazon S3 does support setting some response headers, such as Transfer-Encoding and the like.
+
+One possibility of decreasing bandwidth costs/download sizes would be to GZIP all content on the site and set the necessary header... however there are certain browser compatibility issues to be navigated.
+
+Another side item that would be useful potentially would be to have a config option to create a mapping of files that can be gzipped as an alternate name...
+
+For example:
+
+ gzipped_files => {
+ js => "js.gz"
+ }
+
+Would take all js files and gzip them w/ the altered extension. *This* could allow for using JavaScript to customize what other JS/CSS code gets loaded in based on browser-detection JS code.
+
+--[[harningt]]