KDE.org/New Site: Difference between revisions

Revision as of 17:34, 7 September 2016

We need to get things done.

There were discussions about what to use. At the end, 3 solutions where retained:

Type	Getting things done	Maintainability	Custom themes	Localization
Wordpress	Very fast for basic stuff, more difficult for specific needs	Very good if we don't need to develop new plugins	Not sure if possible to have several presentations of posts	Seems to be possible with extra plugins (POT?)
Drupal	A little more difficult for basic stuff but constant learning curve	Very good if we don't need to develop new plugins, a lot can be done through web interface, configuration can be exported/imported as xml	A theme should be written from scratch / expended. Almost everything possible	Feature present in core. POT files should be exportable with plugins
Custom solution	Very specific, unexpected issues can appear because of corner cases	With a good framework (Django, Sf2, Laravel), there are coding rules so it should not go everywhere but still need to maintain that. What about in 5 years?	Everything possible, should be written.	Whatever we want

Conclusion:

Wordpress seems too simplistic for what we need.
Drupal will need a custom theme and certainly some extra modules (specially for the applications page + announcements) but the tests are very promising.
I (ochurlaud) am afraid that we create a monster like Capacity again and that it's not maintainable on the long run. Also it means that we need to maintain it for security holes etc...

We (Lydia, Olivier and Ken) decided to go with Drupal. The choice is supported by (some) other members.

Steps

Might need to be reorganized...

Step 1

Choose solution. Solution: ......

Step 2

Get a test host by the sysadmin [in progress]
build the solution and test things
Create palette and style

Step 3

Create neverland-based theme.

Step N

replace current kde.org

Specifications

Type of content

Pages: normal pages with text/picture content; Difficulty: easy

Applications: Grid with applications, on click, show details; Idea: use e-commerce plugin and remove prices (can we subclass a plugin?)

Blogs, tweets and other feeds: Feeds plugin,

Theme

The theme must be usable (in term of style and colors) by all the KDE websites and represent its identity.

Other features

Access: Need a ldap/phabricator connection; Need to know who is allowed to do what

@@ Line 2: / Line 2: @@
 There were discussions about what to use. At the end, 3 solutions where retained:
-* custom CMS
+{|class="wikitable" style="border: 1px solid grey;"
-** pro: does exactly what we need
+! Type
-** con: not enough manpower, need to maintain in the future
+! Getting things done
-* wordpress
+! Maintainability
-** pro: very easy to deal with, should be easy to maintain (updates)
+! Custom themes
-** con: maybe not flexible enough, might need to create/edit plugins
+! Localization
-* drupal
+|-
-** pro: easy to maintain (updates) and for users to edit
+! Wordpress
-** con: more complex than wordpress
+| Very fast for basic stuff, more difficult for specific needs
+|| Very good if we don't need to develop new plugins
+|| Not sure if possible to have several presentations of posts
+|| Seems to be possible with extra plugins (POT?)
+|-
+! Drupal
+| A little more difficult for basic stuff but constant learning curve
+|| Very good if we don't need to develop new plugins, a lot can be done through web interface, configuration can be exported/imported as xml
+|| A theme should be written from scratch / expended. Almost everything possible
+|| Feature present in core. POT files should be exportable with plugins
+|-
+! Custom solution
+| Very specific, unexpected issues can appear because of corner cases
+|| With a good framework (Django, Sf2, Laravel), there are coding rules so it should not go everywhere but still need to maintain that. What about in 5 years?
+|| Everything possible, should be written.
+|| Whatever we want
+|-
+|}
+'''Conclusion:'''
+* Wordpress seems too simplistic for what we need.
+* Drupal will need a custom theme and certainly some extra modules (specially for the applications page + announcements) but the tests are very promising.
+* I (ochurlaud) am afraid that we create a monster like Capacity again and that it's not maintainable on the long run. Also it means that we need to maintain it for security holes etc...
+We (Lydia, Olivier and Ken) decided to go with Drupal. The choice is supported by (some) other members.
 == Steps ==