Forum OpenACS Q&A: OpenACS 3.x or 4.x by year end?
your opinions on whether I should use OpenACS 3.x or 4.x and why. You
have my thanks in advance and I'll try to give back through testing
documentation or new module porting/creation.
The website will be a New Mexico state travel website that is a
revamp of an existing site that currently receives 50,000 hits per
month. It is a state of neglect so I hope to double that.
I will be using PostgreSql on a co-hosted Linux box because I like
Open Source and I am poor. I will be the only developer resource
(also because I am cash poor).
I have experience compiling AOLServer and creating C modules for it
but I have no experience with ACS or OpenACS. The core of the system
will be magazine type articles with photographs and I will need a way
to rotate the articles from the top page to a section home page to an
archive. The pages will be static but I will be creating a database
to hold the content (if 4.x I would like to build upon CR).
Sites like I envision this one: gorp.com and away.com (but only state-
wide coverage), photo.net's travel section, bird.net.
If I use 3.x I will be using: adserver, banner ideas, calendar,
classifieds, clickthrough, bboard, contest, FAQ, Neighbor-to-
Neighbor, Survey, Ticket (feedback feature only), general comments
and general links. I will probably cook up a site-wide search using
PLS. Not all the functionality needs to be done by year end, but I
must make the 3.x vs. 4.x choice and have the article system and
adserver production ready.
4.x features I would like to add are site-wide search, photo album
and postcard. I am having a hard time matching up 3.x functionality
such as adserver with 4.x.
I am presently partial to 3.x because it appears to have 80% of what
I need, is time tested, has better documentation (at present), and
has more ArsDigita problem sets and "The Book" to guide me.
Any advice would be most appreciated.
How much do you value the ability to *easily* add new modules to the site that have been created by third parties? Do you have lots of content that needs the new version 4 permissions? Do you need the new version 4 templates? How accepting of risk are you? Very? Moderate? If yes to many of these questions, your way is clear my son.
If no, then you may be happy with Version 3 which as you say, has been tested by time....
By the way, I haven't tried it, but I believe Don says that OACS 4 should support at some point ACS 4 modules. If so, then there is an adserver for 4. I have a version of it for ACS 4.0.1 that folks tell me broke for some reason in 4.2. I dunno. Furfly probably has a fresher version.
BTW - my comment about ACS 4 compatibility only applies to the *Oracle* version of OpenACS4. This guy's using Postgres so there's no avoiding the porting issue.
OK - back on topic. The strongest argument in favor of OpenACS 4 vs. 3 would be the Content Management tools. For your use, Musea's "Edit-this-page" may well suffice. It's relatively easy to build instances that look a lot like the traditional FAQ and News packages as well as more vanilla article scheduling/processing stuff.
The strongest argument against is that not all packages available in the ACS 3 world are available in the [Open]ACS 4 world. aD never completed porting stuff over. We'll be seeing packages with similar functionality appearing over the next few months but will it be in time to help you?
In my opinion use 3.x if:
<li>Site has many users and requires good performance. Since 4.x permissions may not yet scale well.
<li>You need a module has is not yet present in 4.x and is very hard to port or no 4.x modules may substitute for it.
<li>You are sure that the site will not use 4.x features now and the future.
<li>There is little chance of upgrading to 4.x
Use 4.x if:
<li>site needs or greatly benefits from the new features of 4.x like complex permissions, CMS/CR, templating, etc.
<li>site may need or outgrow what 3.x offers.
You may also want to talk to your client regarding this. Well you have to put into lay man terms the problems and scenarios. For me I picked 4.x on site since it was relatively light load and deadlines can be moved. I picked 3.x on another site since I need to make it in 1 week and did not require any 4.x specific features. Look at the short term plus and minus if its going to be a short term project otherwise look at the long term plus and minus. OpenACS 4.x is not in a bad shape, infact its in a good shape considering its on alpha release. Also bear in mind that OpenACS 4.x was based on ACS 4.2. Good luck in your work.