Closed (fixed)
Project:
Views TimelineJS integration
Version:
7.x-1.x-dev
Component:
Code
Priority:
Normal
Category:
Feature request
Assigned:
Issue tags:
Reporter:
Created:
30 Nov 2012 at 22:55 UTC
Updated:
4 Jan 2014 at 02:42 UTC
Jump to comment: Most recent, Most recent file
Comments
Comment #1
fluxsauce commentedI had to fix #1855560: Define options for view plugin which will allow view to be exported before I was able to export the feature.
In order for this to work out of the box on a fresh site install:
This will create a page at /timeline , but will not create a menu item.
Comment #2
operinko commentedThe feature that FluxSauce supplied seems to do the trick more than fine.
I don't see a reason to provide a hard-coded content type in the module, but packaging a feature submodule seems like a sensible idea.
Comment #3
fluxsauce commentedWell, to be honest, that's what I would prefer; that way, if there are any customizations, users don't end up with conflicting features. This is "good enough" for now in my opinion; I'll revisit a formal content type later. Thanks!
Comment #4
operinko commentedWell, my main reasoning is that this is supposed to be an abstract addon to views that can be used by practically any content type.
Packaging a new content type with this would just serve to generate overhead.
If someone wants to use the currently packaged feature, they can enable it to create the content type.
If they, at some point, want to customize and override the feature, they can disable that feature and export their own configuration anyway.
If you have some ideas on why a forced content type would be better, by all means let me know :)
Comment #5
fluxsauce commentedI disagree; by providing an optional content type as a submodule, much like the feature, a site admin or developer can chose to use the guaranteed working content type that's configured out of the box.
Software development isn't just making it work, it's also about usability and making a compelling, easy-to-use product that others can easily use. Steve Krug - "Don't Make Me Think".
Unfortunately, it's not that simple. To demonstrate, I'll use http://drupal.org/files/1855590_make.txt because it has all the dependencies.
Notice that I disabled the existing feature; if I didn't, I wouldn't be able to add the content type to a new feature.
This "works", but it's sloppy and it's not best practice. By creating a true content type in an optional submodule, it can be featurized without conflicts.
Comment #6
operinko commentedAh, good points :)
I'll add that to my TODO list.
Basically, it should be an optional sub-module that provides a content type with the basic fields.
I'm just wondering if I should put a Link field to it or not, since it's not a dependency on the main module.
For some reason it would seem a bit harsh to add extra dependencies on the sub-module (that is, it should have the bare minimum amount of fields to make it work, right?)
Comment #7
operinko commentedComment #8
fluxsauce commentedI would say it's not a problem. Dependencies are pretty normal, it prevents duplication of effort. Additionally, http://drupal.org/project/link has 182202 has sites currently report using this module, it's pretty ubiquitous.
Actually, I can do both the content type and the view without a feature, which will (ironically?) remove the dependency on features.
Comment #9
juhaniemi commentedIMO there are very few reasons to provide any hard coded content type in Drupal 7 era and up. Drupal is elastic and flexible, content types are fieldable, even content created by this module shouldn't be removed if this module is uninstalled and replaced by a similar module. That's why I would absolutely vote for feature submodule if we want to provide a simpler way to start kicking with this module. Or are we talking about the same thing?
Comment #10
fluxsauce commentedI outlined the problem with using features in http://drupal.org/node/1855528#comment-6837450 - it works great until you try to adapt it.
If you provide a content type in code, it can be featurized without conflict, and content wouldn't be removed (that would have to be explicitly implemented in hook_uninstall) upon removal.
Another advantage of hard coded entities is speed and flexibility, which I will freely admit is not needed in this instance.
As it stands, the feature I wrote is now committed and works. The discussion was whether featurizing was optimal, or just "good enough." I feel it's "good enough" and I can live with it, as you feel it's optimal over a content type.