Skip to content
Browse files

[#88] Improved Attachments API

Can have multiple attachments, using Paperclip rather than our custom hand rolled version.
  • Loading branch information...
1 parent 6c65853 commit 801447afd68d3fc9c8dc13ef47c35eeef4623091 @peakpg peakpg committed May 9, 2012
Showing with 30 additions and 14 deletions.
  1. +20 −0 doc/release_notes.md
  2. +10 −14 todo_list.markdown
View
20 doc/release_notes.md
@@ -1,6 +1,26 @@
v3.5.0
======
+This release includes a number of new features, including:
+
+* Improved Attachments
+* Mobile Friendly templates
+* Rail 3.2 compatibility
+
+See the upgrade instructions here for existing projects:
+
+Improved Attachments
+--------------------
+
+Attachments have been completely reworked to use Paperclip (https://github.com/thoughtbot/paperclip).
+
+* Each block can now have multiple attachments using to different styles.
+* Attachments can be defined as one to one (has_attachment :image) or be stored as a collection (has_many_attachments).
+* Upgrade migrations are provide to migrate file and data for older projects to the new attachment structure.
+* New generators have been provided to create content blocks with the new attachment styles.
+
+See this Attachments API guide for more details: https://github.com/browsermedia/browsercms/wiki/Attachments-API
+
Mobile
------
View
24 todo_list.markdown
@@ -1,10 +1,17 @@
Tasks:
-* Paperclip Based Assets - Replace the existing 'custom' file upload behavior with one that uses Paperclip. Allow blocks to have more than one attachment.
+* Heroku - Make BrowserCMS work on the platform.
+
+# 3.5.x Goals
-- [Documentation] Write a quick how to guide about Attachments API.
-- [IMPROVE] The error message for a duplicate path for Fileblocks is not clear.
+* Faster Files - Take advantage of X-Sendfile (if possible) to speed up sending uploaded CMS files.
+* [IMPROVE] The error message for a duplicate path for Fileblocks is not clear.
+* Browser Compatibility Testing - Ensure compatibility with latest versions of Chrome/IE 9/Firefox/Safari. IE 9 probably needs the most works.
+- [TEST] Upgrade script from 3.4.x and 3.3.x
+- [BUG] Updating a page throws 'path already used' error? Created a public section.
+- [BUG] Reverting blocks with multiple attachments doesn't really work. It stays associated with the latest attachment, even if you had previous deleted attachments.
+- [BUG] Minor - In development mode, need to restart if core CMS code is changed (loses definations for custom blocks with attachments)
## Publishing/Versioning Improvements:
- The mess that is publishing/saving is coming back again when trying to interact with blocks with associated attachments.
@@ -20,16 +27,6 @@ When almost certainly should be.
- Migrations are now generated with .cms. Will this cause problems during upgrades? (Write upgrade scenarios)
- rails -h only provides generate | destroy methods with engine on a new project. Why? It sucks to have to cd into test/dummy to run tests.
-# 3.5.x Goals
-
-* Heroku - Make BrowserCMS work on the platform.
-* Faster Files - Take advantage of X-Sendfile (if possible) to speed up sending uploaded CMS files.
-* Browser Compatibility Testing - Ensure compatibility with latest versions of Chrome/IE 9/Firefox/Safari. IE 9 probably needs the most works.
-
-- [TEST] Upgrade script from 3.4.x and 3.3.x
-- [BUG] Updating a page throws 'path already used' error? Created a public section.
-- [BUG] Reverting blocks with multiple attachments doesn't really work. It stays associated with the latest attachment, even if you had previous deleted attachments.
-- [BUG] Minor - In development mode, need to restart if core CMS code is changed (loses definations for custom blocks with attachments)
# 3.4.x
@@ -38,7 +35,6 @@ When almost certainly should be.
* Refactor Cucumber steps to add seed data once as part of the env.rb file, then use truncation to leave it there.
* If user's try to add to a Rails 3.2 project, it will blow up midway through (i.e. the jquery-rails dependency will fail since R3.2 require jquery-rails-2.0. A better plan would be to fail fast.
-
# Future
* How much value is there to allow users to pick the table prefix (as compared to the complexity it brings). Would it be better to just force everything to cms_?

0 comments on commit 801447a

Please sign in to comment.
Something went wrong with that request. Please try again.