Forum OpenACS Development: Schedule
In the meantime, we hope we can get *everyone* to participate in commenting, contributing thoughts and ideas, and kicking our butts (specifically mine) every now and then to make sure we are still keeping everyone involved.
Is there anything else that needs to be done to the driver besides adding bind variable emulation?
I'm not a software engineer by training and my technical skills are minimal. I've been using OpenACS as a learning tool along with ACS's psets. Knowing that whatever I'm learning in OpenACS 3.x will be transportable to OpenACS 4.x is comforting.
I can't even begin to fathom the amount of time and effort it'll take Ben Adida and others to port ACS4 to the OpenACS form. All I can say is a big thank you for the effort from everyone.
I hope I can contribute to OpenACS one of these days.
I had previously installed ACS4/Oracle on my machine, but since I cannot afford to pay for Oracle I decided to remove that again.
Should I reinstall ACS/Oracle on my development machine? ACS 3.x is (at least to me) rather self-explanatory code wise. But I had a hard time finding the ACS4 entry points... what is a good place to start?
[yes, I am aware of the boot-camp-self-study things]
IF you have a minimum of 192MB RAM. I bumped my laptop up because it was swapping horribly with 128MB RAM when I had Oracle up along with KDE+Konqueror or Mozilla. Makes the battery run down fast ...
In contrast, OpenACS (PG vs. Oracle, that is) runs like a dream in 128MB.
As far as I can tell, we didn't release the OpenACS4 core in alpha on 01 June, nor in beta on 30 June, making the listed 31 July date for a final release' of the core look decidedly unlikely.
I'm actually impressed at how much progress has been made with the port -- well done to all who have worked on it! But for outsiders looking in, it would probably help the 'image' of OpenACS as a whole if the schedule was kept a bit more up to date.
On a related note, in the same text box on the http://openacs.org/4/ page which links to the schedule, there is a title of "User Guide (June 2001)" which should probably be updated to August 2001?
<a href="http://openacs.org/4/schedule">here</a>. This "schedule" is merely a ramble about some dates I think might be possible, and is subject to change, overhaul based on your bronx cheers, laughter, and other feedback.
Take a look and fire away. My major motivation in updating it is to get feedback. I'll be asking for specific feedback from package porters soon and will use their comments about their confidence in making a release happen to make a more precise list of what we might be able to release and when, hopefully within a framework similar to the one laid out in this rough "schedule" update.
The revised schedule is at http://openacs.org/4/schedule.
Other than that, two minor nits: (i) it will be safer to change the "as of today" phrasing to a link to a specific version of the status document, so we don't lose track of which modules were in what status as of July 29th; (ii) 'documenation' should be 'documentation', there are two occurrences of this.
One more thought: it would be easier to assess how realistic these dates are if we had all or most of the 'non-porting' tasks visibly listed in the status document ... especially if none are assigned yet, some of these tasks might become critical path items otherwise.
Roberto: Any progress on a documentation task list or similar?
And, yes, I forgot to spec HTML when I posted and this paleolithic version of bboard doesn't offer me the choice of switching presentation nor does the admin's "edit post" page work. Hurry up, Musea, save us from ourselves! :)
Roberto is indeed busy putting together a page on documentation stuff, once he's got that done and is comfortable with it we can talk about putting parts of it into the status sheet. He's been talking to folks who are interested in helping with documentation so I think we're going to be in pretty good shape there.
Testing's the big bugaboo at the moment, I'd love it if someone would step up to organize testing ala Roberto's work on documentation. Failing that, it will be next weekend before I can get serious about talking to folks and trying to get some organizational stuff going myself.
Pointing to "the state of the status page as of today" seems to me to be a somewhat unstable reference. Even "as of 29th July 2001" is imperfect, because the status page will be ever changing...
Also the date at the top of the schedule (which I did see!) could change, say if you revise the schedule slightly in a week or so based on feedback from porters, and then the 'today' reference would accidentally point to the wrong version of the status page if it were left in the revised schedule.
If instead you link from the schedule to a specific version of the status page, then it is (a) clear and (b) easy for people to follow the reference and find out exactly which modules will (or should) be in which release.
Sorry if I'm being overly picky here... at least it is feedback
I'm extremely busy with client work this week, but will be trying to contact folks to double-check with them that this date's fine. I've already heard from Roberto, for instance, that he exects to have the CMS package (currently marked "DATAMODELS", I think - certainly not "PORTED") ported by the end of next week at the latest, squeezing in under my hand-waving August 15th goal by a few days. I need to get similar "yay" or "nay" statements from others soon.
I meant my posting of the revised schedule to be a basis for discussion of its practicality and didn't want to firmly put anyone on the spot, in other words.
As I get a better handle on what is and is not practical, I'll be able to be more precise as to what to expect.
first alpha (pre-)release date to September 1st. This will give folks
working on a couple of the trickier packages time to finish and adequately test things, give us time to finish putting together our plan for more formal testing, etc.
I'll post a list of packages which we expect to include in this first alpha release later this week or early next.
Things are looking pretty good, actually - just because I'm holding off doesn't mean you shouldn't download from CVS and try it! Just check the project status file for packages that have been ported and play away.
Can you please update it to the new timelines?
I'm holding off even evaluating OpenACS 4x until it's Beta, despite so many insisting I should consider the Alpha version production worthy, and would like to know when that's finally going to be out.
Alpha2 should be out this week, with beta coming out a week or two later and a final version a week or so after that. At least that is the inside word. Whether it's exactly this way or slightly different is, you should see some movement reasonably soon.
That being said, if you are unwilling to download the software and compare the unfinished system that aD abandoned with what the OpenACS community has worked on for the past 6 months to clean up and extend quite impressively, I doubt waiting another week or two will help one way or the other.
As Talli said I'm hoping to roll out a second alpha tarball this week (I'd hoped for Monday but became ill on Thursday and am just now back to being more-or-less normal). Note that the schedule didn't have a slot for an alpha2. I made that (somewhat unilateral) decision based on the fact that I and others fixed a lot of bugs and added a couple of previously unported packages. We should get some more testing in by beta.
I'm hoping for a alpha2+2 weeks schedule for a beta and an alpha2+4 weeks schedule for a final release.
The Oracle port of Edit This Page was not complete, I intend to
complete it before building an alpha2 because I feel it is
*extremely* important that we have this package available for Oracle
in our first release. And that it receive a decent amount of
Neophytos and David Bauer are working on integrating the latest
version of acs-workflow (we started with an earlier version) so it,
too can make it into the alpha2 tarball. They have told me they'll
have it done by Monday.
Those are the two outstanding items holding things up. Last week I
spent a lot of time getting wimpy point to work (the original Oracle
version had a some - uh, significant - problems as did the PG port
that copied them). There's still a lot needed on wimpy point but
I'm not going to hold things up for them. Other people have also
been chasing and fixing significant bugs.
As soon as I get this tarball out I plan to make a list of things
that I feel need doing before we release a beta. It's not a long
list and I'll post it for public comment to see if there's general
Having a fixed task list should make it easier to hold firm to a
schedule for the beta release.
This beta will lead quickly (one to two weeks) to a release as we'll
ask to freeze feature additions/package improvements and *only* fix
important bugs between beta and release.
I know folks are impatient, but emotionally I'm only capable of
accepting a certain amount of brokeness even in a release I label
Do you have an estimate of when you will have the final version? I want to install OpenACS; but I rather wait a few more days and get the new one than to install and have to upgrade a few days later.
major bugs fixed is the Right Thing To Do, for sure. Sounds like
we can hope to have a good, solid final release by New Year.
post here, but can we get an update on the likely milestones
between here and first release, as well as an ETA?
BTW, despite the fact that I entered the word "schedule" when I
searched the forums, and despite the fact that the title of this
thread is "schedule", it actually appeared almost halfway down
the results list. Will the new search package give us better
We've got some progress on the documentation front, especially in regard to outlining the tasks that are critical for a first release. I haven't talked to Roberto yet about details but I've been insisting for some time that the biggest barrier to a beta release is the lack of good, integrated installation documentation and the like (various files uploaded to openacs.org don't count).
We can now create HTML files from the original aD XML files (even I can do that following Roberto's instructions, which were based on aD's original howto). Vinod has provided us tempates for them which use our OpenACS logo rather than the aD logo. Roberto's working through the documentation removing aD references from their .xml sources.
Etc etc etc. Much has happened to make this last barrier look more like a speedbump and less like a roadblock.
But I'm not interested in providing more hand-waving dates until Roberto's ready to commit to some dates, which means getting a few more documentation volunteers.
We could use folks to help in simple ways, i.e. editing for English, spelling and other non-technical issues ...
seemed to require technical expertise. If you need a copy-editor,
though, I can do some of that.
Regarding search, I don't think it's unreasonable for a search
package to consider the subject line and topic labels in addition
to words in body and dates in its relevance ranking, and I don't
(personally) think the inclusion of this should require a specific
choice or intervention on the part of the user.
Regarding the release, am I correct in understanding that
documentation is the main remaining hurdle, i.e., that you don't
expect any more major changes to the code beyond what's in the
current beta? That's really what I'm trying to find out, more than a
release date. I want to know when you think all non-cosmetic
and non-documentation changes for the first release are likely to
be completed or, at least, what non-cosmetic and
non-documentation changes still remain to be made.
Don, I know this must drive you nuts, but couldn't we have a pre-beta after which any data model changes would have to be accompanied by upgrade scripts? And roll the real beta when the docs are done?
For openfts, it's also possible write to write a custom scoring function that would allow more restrictive searches using boolean operators, wildcards, exact matching, etc. I haven't looked at the openfts stuff in a while, so it's possible that new scoring functions are already available.
So now I'm planning on cutting it next weekend.