Sandbox

Since we've had many questions on this topic I'm posting a "sticky" thread in this regard.

We're working on a project categorization system that will integtrate with Joomla (i.e. it will be a Joomla Module). Project Admins will categorize their projects such that site visitors/users can navigate by interest/need to select the project(s) of their choice. We will tie it into a "Immediate download from File Release" system is that's all that is needed (so if all you need is the app, you can pull it down from the project categorization system without actually having to enter the project and then navigate to the File Rleases app).

Col

One of the problems with self-categorization in the past has been the obvious mis-categorization done for marketing purposes.
Such pollution devalues the usefulness of the categories and the forge itself.
Categorization cannot be left to the project admins alone.
There needs to be some oversight and management of the projects.
Moderators for the forge.

One thing the Miro monster has done right is delete ~700 empty projects from MamboForge.
The noise-to-signal ratio in MamboForge was deafening.
Please don't give me any crap about "what am I doing about it" - I offered twice to clean-up MamboForge, both ignored.
How many other people requested the MamboForge be cleaned-up? I saw plenty.

Categories need to be the categories that USERS expect.
Some people saw no need for the Mambo Components, Modules, and Mambots Resources list.
"We have the forge, we have categories, blah, blah, blah."
Just checked - Mambo CMM Resources = 56,414 views.
Some people apparently found the categories and the format useful.
Please correct me if I am wrong, but the CMM Resources and the Doc Resources lists
views climbed faster than any other threads ever (Doc Resources now at 108,421 views).

Moderated categorization - including projects hosted elsewhere.
That would be really useful, and there would be no need for a Joomla CMM Resources list.
Let's fix the limitations and short-comings of MamboForge and have better usability at Joomla forge.

Yeah, it's late, I'm tired and pissy - my apologies.
But, the points are valid.

kenmcd wrote:One of the problems with self-categorization in the past has been the obvious mis-categorization done for marketing purposes.

Agreed when I first went throught the mamboforge ....... it was really difficult for me to use the Category section...... And to find out which were the Good and quality addons....... I did more of googling and forum searching than use the Category section for finding what i wanted..... Solution would be to have a better catogory section .... moderation ..... ability to view by number of downloads and popularity........

Boomzilla wrote:We will tie it into a "Immediate download from File Release" system is that's all that is needed (so if all you need is the app, you can pull it down from the project categorization system without actually having to enter the project and then navigate to the File Rleases app).

Could there also be a download counter here ?  Like 'This file was downloaded XXX times'

All good comments/advise. Thanks. Keep them cards and letters coming.

Col

I think a type of feedback system might be helpful such as "Report to moderator" that would be found in most forums. Then anyone looking at a project could question it's placement or existence.

I agree with crash77.  The new forge is starting to fill up with empty projects that could be interpreted as "marketing" projects.

It annoys me when I see a project that shows it is STABLE but has NO files and when I look at the project, there is no referencing website or any other information. It is wasteful of anyone's time that looks at this project and there should be a way to report this type of project.  I wouldn't be opposed to having a "marketing" type of section strictly for links to "commercial" projects as I as well as many others do search out commercial projects as well and it is sometimes difficult to find them...

Perhaps an option to hide a project from the public by the project admin would be helpful? Once a public release is ready, the project can go "live". Empty projects (that don't want to recruit new members) can be wasteful of other folks time. Good points Aaron.

Also, is there any particular way to format packages > releases > files.

I have one project, with stable files, templates

But, would like to add some components and related modules later, should I start a new project or just add into this one?

It would be great to be able to link separate files to different categories. This would allow a visitor to be able to find a template for instance that a developer is working on but once on their project page, they would also see the module and component being worked on. If you can only categorize the "project" then my advice would be to create a new project for every initiative.

There are few different problems to getting good USER oriented categories, and more useful content in the forge.

Some potential solutions:

- Projects with no files do not show up to average users.
Vaporware, marketing BS, un-completed/never-completed - all are just noise to the USERS.
Perhaps there should be a time limit - no files after one month - project disappears from public view.
MamboForge deleted over 700 empty projects.
No one is going to miss this noise.
Search results in MamboForge were 50% to 90% useless noise.

- The obvious marketing fluff template projects should all be put in a "marketing fluff template" category.
That way users would not have to sift thru all this to find real CMMs.
Some templates are real quality open source contributions.
Many are just low-quality bait to drive traffic to a web site to sell the quality stuff.
Regardless - templates should be in a separate category from CMMs.
Again this is where forge moderation is needed.

- One CMM per project, or a way to place the separate sub-projects (releases) in the correct categories.
Categorization at the release level is needed.
Many developers are not very good at categorizing, or even describing, their projects.
While making the CMM Resources list, there were times I had no idea of what some things did.
Moderation is needed to make the forge more usable.


Moderation makes the forum more useful for everyone.
Why not the forge?

kenmcd wrote:
- The obvious marketing fluff template projects should all be put in a "marketing fluff template" category.
That way users would not have to sift thru all this to find real CMMs.
Some templates are real quality open source contributions.
Many are just low-quality bait to drive traffic to a web site to sell the quality stuff.
Regardless - templates should be in a separate category from CMMs.
Again this is where forge moderation is needed.

Moderation makes the forum more useful for everyone.
Why not the forge?

Templates should definatly be in a seperate catogory... I sick of seeing templates everywhere. also I have to download and install them to see if they are good or not. Isnt there any way to force these people to add a preview before people download it. I think steps should be taken before the forge becomes a junkyard. I Agree some degree of moderation is required . But we need to be very diplomatic. It would not be nice to delete somebodys project just because they didnt update it. The best thing is to hide them from view based on the following criterias
1. Popularity.
2. number of downloads.
3. Project updated time.

Popularity
I would never recommend hiding a project based on popularity or number of downloads.
Some excellent projects are saddled with poor descriptions or poor "marketing."
Moderators could help make useful projects more visable.
Some VERY good CMMs are lost in obscurity because the developers are GREAT coders, but are also poor marketers.
Gee, this is news.

Freshness
Some critera related to how current a project is would be useful.
From a search usability wishlist standpoint, freshness would be incorporated into the search.
The same freshness criteria would be very useful in the browsing of categories.
In the CMM Resources list, I, very subjectively, tried to list newer more recently updated projects at the top of a category.
Projects which are updated more often are reasonably expected to be more useful to users.
Freshness would be a very powerful addition to the JoomlaForge search.

kenmcd wrote:There are few different problems to getting good USER oriented categories, and more useful content in the forge.

Some potential solutions:

- Projects with no files do not show up to average users.
Vaporware, marketing BS, un-completed/never-completed - all are just noise to the USERS.
Perhaps there should be a time limit - no files after one month - project disappears from public view.
MamboForge deleted over 700 empty projects.
No one is going to miss this noise.
Search results in MamboForge were 50% to 90% useless noise.

- The obvious marketing fluff template projects should all be put in a "marketing fluff template" category.
That way users would not have to sift thru all this to find real CMMs.
Some templates are real quality open source contributions.
Many are just low-quality bait to drive traffic to a web site to sell the quality stuff.
Regardless - templates should be in a separate category from CMMs.
Again this is where forge moderation is needed.

- One CMM per project, or a way to place the separate sub-projects (releases) in the correct categories.
Categorization at the release level is needed.
Many developers are not very good at categorizing, or even describing, their projects.
While making the CMM Resources list, there were times I had no idea of what some things did.
Moderation is needed to make the forge more usable.


Moderation makes the forum more useful for everyone.
Why not the forge?

I agree!! However... restricting projects that have not released files is not a good option because I for one would love to be able to track some of those projects for when the files are released.  I like to see what is on the horizon! I do believe those projects should be "flagged" in some way so a forge visitor can easily know if that project has files or not. None the less, it is not as bad as the miscategorizations that do occur and even more so the purposeful miscategorizations to mislead someone.

Not for nothing but I think if we don't get developer.joomla.org organized soon it will be a real thorn in our side. I have to browse through 280 projects right now with NO idea of what kind of files I might or might not find...

crash777 wrote:Not for nothing but I think if we don't get developer.joomla.org organized soon it will be a real thorn in our side. I have to browse through 280 projects right now with NO idea of what kind of files I might or might not find...

I agree..... I was reading a few other comments and articles .... about joomla many people are dismayed by the complexity of the forge. But I guess everthing takes time just hope it wont take too long.

Thanks for the posts with input, ideas, recommendations etc.
The implementation of project categorization will be keyed to project activity. No activity/empty projects will drop to the bottom of the list + we'll support a "hide" capability such that projects with less than X activity will not show up in the categorization list (modal).

I think we're seeing a large number of empty project as this forge is quite new and project teams have (a) initially grabbed the project name they had on mambo Forge and (b) are working on migration of project data, code etc. I expect to see more content over the next few weeks.

Col

Some of the abuse I think we all are referring to are projects like this:  http://developer.joomla.org/sf/projects/joomla_test

Regards

Don't discount projects like that. Sometimes they are new project leaders trying to get a feel for the site. I wouldn't necessarily call it abuse. Those type of projects at least explain what they are in the title and I don't have to waste time clicking further. I am more annoyed with the projects that STATE they have stable file releases and then when you click on them, they refer you to their website where you are bombarded with adsense, banners and NO stable release files... or a promise that one is coming. Now THAT is a total waste of time and ABUSE in my book.

I partially agree with you on the missing files for projects.  I also stand by my point that a vague description like this is probably someone registering a project (abuse may be harsh but it's probably not a valid project) just to see what the process is like on the forge site.

hmmm okay, I will submit to that view

I see what you are saying. Yes, abuse is harsh but I guess it is technically correct. They should not be using a live server to "test".

crash777 wrote:I see what you are saying. Yes, abuse is harsh but I guess it is technically correct. They should not be using a live server to "test".

Ok, so find me a "test server" and I'll go try SourceForge on that one  .

Seriously, I've registered 3 projects on the J-Forge to date.  None of them have files (yet).  It will probably take me a month or more to get to all of them.  I agree that empty projects can create confusion, annoyance, and clutter, but I think that we need to set some sort of reasonable guidelines.  Like maybe we can automatically delete all empty projects after 1 year (with advanced notice to the project administrator to correct the situation).  For example, I have a project on the old MamboForge that I was going to do but never got to.  It's been empty for almost a year.  Deleting it (with advanced notice) at this point wouldn't bother me.
   

Sorry, I wish I could but I can't help you with a test forge server.

I have no problem with someone registering a project that will contain something but to register "joomla test" (or whatever it was) seems counterproductive. There have been times where I HAVE to make changes to live servers (most of my clients as a matter of fact do not want "test servers") but I would never make a change labeled "test". In other words, if I were to start a component project on the forge, I would not label it as "test". Even though I am "testing" (more accurately, LEARNING) the server, one day it will have something useful for the community. The name should reflect what it will be. This leads right into proper categorization which is a responsibility of the admin of the server to setup and the moderators and project leads to keep in check. I think it would be a great idea to have "test" areas of the server so people can learn! I don't think anyone was specifically targeting you... it was more on line with the need to properly categorize projects that will not be immediately useful to someone looking for files to download. I hope I got my point across.

crash777 wrote:Sorry, I wish I could but I can't help you with a test forge server.

I have no problem with someone registering a project that will contain something but to register "joomla test" (or whatever it was) seems counterproductive. There have been times where I HAVE to make changes to live servers (most of my clients as a matter of fact do not want "test servers") but I would never make a change labeled "test". In other words, if I were to start a component project on the forge, I would not label it as "test". Even though I am "testing" (more accurately, LEARNING) the server, one day it will have something useful for the community. The name should reflect what it will be. This leads right into proper categorization which is a responsibility of the admin of the server to setup and the moderators and project leads to keep in check. I think it would be a great idea to have "test" areas of the server so people can learn! I don't think anyone was specifically targeting you... it was more on line with the need to properly categorize projects that will not be immediately useful to someone looking for files to download. I hope I got my point across.

Ok,I must not have ready the entire thread.  I didn't realize that someone had registered a project called "test" .  I agree, that has no place on J-Forge.  So I guess we're actually in agreement  .

Kewl.. We all agree to agree... LOL

So, couple of comments.

1. If I (or other site admins that approve projects) see a project with a name and/or description that is meaningless/confusing I reject them with (what I think is) a polite request to resubmit with a better name ("better" in this case being in the eye of the beholder...)

2. Yes - we will get cruft growing over time. We will start purging projects in a while (length of time defined by "a while" still TBD). Process will be one of notification of removal of project to project admin(s), wait for reasonable amount of time (see definition of "a while), and then either removal or offer the projectt up for adoption (for a while).

Oh - and the projec categorization system will help promote active projects and relagate inactive projects to the netherworld.

Col

Boomzilla wrote:1. If I (or other site admins that approve projects) see a project with a name and/or description that is meaningless/confusing I reject them with (what I think is) a polite request to resubmit with a better name ("better" in this case being in the eye of the beholder...)

Perfect! This will keep the "riff raff" out ...

Boomzilla wrote:2. Yes - we will get cruft growing over time. We will start purging projects in a while (length of time defined by "a while" still TBD). Process will be one of notification of removal of project to project admin(s), wait for reasonable amount of time (see definition of "a while), and then either removal or offer the projectt up for adoption (for a while).

Well the other recommendation about being able to filter projects with "x" amount of activity will help the visitor to filter out the cruft on their own. If the projects get purged after a certain amount of time, that will help trememdously (we shouldn't have to filter through "Please remove this project" - there should be a way for a project owner to either remove their own project, or mark their project as "dead"

Boomzilla wrote:Oh - and the projec categorization system will help promote active projects and relagate inactive projects to the netherworld.

Thanks for listening to us! It is very much appreciated!

A project rating system may help as well.

Boomzilla wrote:Thanks for the posts with input, ideas, recommendations etc.
The implementation of project categorization will be keyed to project activity. No activity/empty projects will drop to the bottom of the list + we'll support a "hide" capability such that projects with less than X activity will not show up in the categorization list (modal).

I think we're seeing a large number of empty project as this forge is quite new and project teams have (a) initially grabbed the project name they had on mambo Forge and (b) are working on migration of project data, code etc. I expect to see more content over the next few weeks.

Col

Good news.

Another problem I ran into with the CMM Resources list was projects which belonged in multiple categories.
Projects such as Community Builder belonged in User Management, Directories - People, etc.
Xaneon Extensions belonged in SEF/SEO and in Multi-sites.
phpShop - shopping cart or stand-alone catalog.
There are many others.

Some components have alternative uses which are not immediately evident.
There were components which were well known in one category that had other uses which were almost unknown.
Usually these were discovered by a random forum post which said "I used this component for that, look here."
(some creative uses by just setting configuration)

The ability to include releases in multiple categories would be very useful. 
(this would definitely need some oversight to prevent abuse)