[technology-pmc] Review and recommendations for the ALF project

[Wayne Beaton <wayne@xxxxxxxxxxx>]

I've done a review of the ALF project.

By way of background for new PMC members, the ALF committers were 
approached some months ago [0] to rationalize continuation of ALF as an 
Eclipse project. At that time, the Technology PMC expressed concern that 
ALF was had not yet managed to develop a community around the project 
and, it seemed, had seemed to have given up any attempt to develop 
community. Their initial response was to defer the discussion. It seems 
that, with passing time, they have chosen to ignore the PMC's request 
for discussion (IMHO, the onus should be on them to follow up, not us).

I have summarized some of my own findings below. My take-away from the 
information is that ALF has not done enough to encourage community 
involvement. While it is relatively easy to determine what the project 
provides, it is difficult to obtain the code, engage the committers, and 
get directly involved in any meaningful way.

According to the Foundation database, their first committers were given 
status in August 2005. This means that ALF has been a Technology project 
for more than three years. I humbly suggest to the Technology PMC that 
this has been long enough for ALF to establish itself and that it is 
time for ALF to establish an end-game plan. More specifically, ALF needs 
to either (a) move to another Eclipse top-level project, (b) become a 
top-level project, or (c) be retired as an Eclipse project (archival).

Upon approval of the PMC, I intend to announce this requirement for ALF 
to establish an end-game on the alf-dev list. If ALF intends to pursue 
either of option (a) or (b), then they must produce a detailed plan for 
the move. I intend to give them two weeks to respond after which time an 
archival review will be scheduled on their behalf.

Thoughts?

Wayne

Here are some of my findings:

--
I apologise if these comments seem overly negative. There is many 
positive aspects to this project; I have not focused on these as much as 
I probably should.

Description on home page is informative. However, there is no link to 
the "Project Info" page from their web site. The link off their main 
page titled "Requirements" points to a wiki page in *edit* mode.

ALF Blog [1] has been inactive since March 2006.

Project planning page [2] has not been updated since 2006. They have not 
provided a project plan in standard form.

"ALF Team" page [3] has some layout issues. George Stanchev is not 
included on this page (he is a committer according to the Foundation 
database).

Latest downloads are dated December 2008. The "ALF Downloads" page does 
include a reference to CVS. The organization in CVS is not easily 
understood. After some searching, it seems that they've nested Eclipse 
projects inside an Eclipse project. Ultimately, I had to download the 
latest build to determine what I should be looking for in the 
repository. It'd be nice to see a link to a Team Project Set or 
Buckminster scripts to make it much easier to add the code to an Eclipse 
workspace.

There have been 503 downloads of the latest milestone build 
(ALF1_0M8-2007-12-04-0018-incubation.zip). 15 of those downloads 
occurred this past Saturday. This strikes me as a little curious. Any 
idea why this might be?

Activity in the mailing list is slow. For the past several months, most 
of the conversation has been around the topic "Is ALF still alive?". 
Granted that much conversation was initiated by the Technology PMC, but 
at least two community members has weighed in ([4], and [5]). Response 
from the committers has been muted.

The project summary page [6] does not layout well. It seems as though 
they're just linking their landing page as the description.

The newsgroup, eclipse.alf, has had no activity since August 2008. 
Activity on the newsgroup has been very light and seems to be almost 
entirely communication from committers. There seems to be very little 
evidence of community involvement on the newsgroup.

The project plan [7] is a PDF rendering of a Microsoft Project document. 
The use of Microsoft Project is not, in and of itself, a problem. 
However, the document itself is intimidating. It gives the impression of 
an entirely closed shop, leaving no room for others to contribute. It's 
also out of date. The release date indicated on the project summary page 
is in the past and needs to be updated along with the plan.

Almost every one of the 222 bugs raised against ALF was created by a 
committer. Despite the fact that the 1.0 release has had milestone 
releases for 10 months, there is no 1.0 version provided in Bugzilla (so 
that adopters can raise bugs against that version).

[0] http://dev.eclipse.org/mhonarc/lists/alf-dev/msg00722.html
[1] http://eclipsealf.blogspot.com/
[2] http://www.eclipse.org/alf/alfdates.php
[3] http://www.eclipse.org/alf/alfteam.php
[4] http://dev.eclipse.org/mhonarc/lists/alf-dev/msg00734.html
[5] http://dev.eclipse.org/mhonarc/lists/alf-dev/msg00736.html
[6] 
http://www.eclipse.org/projects/project_summary.php?projectid=technology.alf
[7] http://www.eclipse.org/alf/includes/ALF_Project_Plan.pdf