Forum OpenACS Q&A: I want my, I want my, I want my web/db

Hey folks,

Thanks to everyone working to scavenge aD.  Let me apologize for not
being able to follow up as I had agreed to, I'm in the middle of some
ugly personal stuff that I need to get through.

That said, with the demise of aD, I mourn the loss of our high signal
to noise ratio web/db forum.  I would like to suggest that we create
another forum here, or perhaps formally relax the suggestion,
inherent in the name, that this forum pertain mainly to OpenACS issues.

I want my web/db!  (Of course, I'd also like some money for nothin'
and chicks for free.)

How about it?

Collapse
Posted by Cathy Sarisky on
I want my web/db! (Of course, I'd also like some money for nothin' and chicks for free.)

Actually, that should be clicks for free...

Sorry, it's Monday, and I just couldn't resist.

Collapse
Posted by James Thornton on
Is it too late to get the aD bboard data? Is anyone trying?
Collapse
Posted by Don Baccus on
I also thought about maybe putting up a web/db board here for non-OpenACS specific stuff.  How do others feel?

Importing the old data wouldn't be that hard for someone with time on their hands.  Anyone have time on their hands?

Collapse
Posted by James Thornton on
If you can get it...
Collapse
Posted by Talli Somekh on
I think reviving web/db here is a good idea, although I think a question that should be asked is whether it will become the de factor place for ACS Java questions. That community seems to be without a home now, and I can see it following web/db over here. I'm not sure if such a move will end up being noise or not. But all in all I think it's a great idea!

That being said, I nominate Todd Gillespie to migrate the web/db and OACS bboard stuff, if for no other reason than he will bitch so eloquently upon learning of his nomination.

talli

PS
Is that content under any kind of license? Do we have to worry about people's usernames and IDs being moved to new ownership without authorization?

Collapse
Posted by Mike Linksvayer on
You could always create an "acs/java refugee support" forum for those abandoned by aD. The irony of it: OpenACS community lends a hand to those left high and dry by dead company that not long ago more or less told said community to shove it. :)

OTOH who knows, perhaps RedHat will continue developing acs/java or find it a good home. AFAIK none of RedHat's plans for aD's assets are known.

Collapse
Posted by Don Baccus on
Well, Mike, that's exactly what we did with ACS Tcl/Oracle users, gave them a new home, and look at us now, supporting Oracle and PostgreSQL with a single code base!
<p>
For our next trick, we'll support Java and Tcl versions with a single code base!  Yeah!
<p>
<b>I AM MAKING A HORRIBLY BAD JOKE. IGNORE ME</b>
<p>
I'd have no problem with our having an "ACS Java Refugee Q & A" forum here, though.  How do others feel?
Collapse
Posted by Michael A. Cleverly on
I have uploaded the web/db messages threads into my personal file storage area. I had to split it up into two files since it just barely exceeded the 5mb limit as a single file (~5.2mb). If someone can get an export in SQL format from aD/RH that'd make the conversion go easier. Otherwise, if we're stuck parsing the HTML threads, I'd be happy to volunteer...
Collapse
Posted by Don Baccus on
AFAIK we're stuck parsing the HTML ...
Collapse
Posted by Mike Linksvayer on
From what I understand ACS4/Tcl (i.e. Oracle) users have a new home here (not to mention a non-EOL'd product) as a side effect of porting ACS4/Tcl w/added multi-db support. Presumably OpenACS gained as much from aD's ACS4/Tcl work as ACS4/Tcl users did from having OpenACS support Oracle, if such a comparison could be made.

In constrast, offering ACS/Java users a gathering place would be purely a guesture of kindness and goodwill. I most definitely wan't suggesting that OpenACS development encompass Tcl and Java!

Collapse
12: I can help with this... (response to 1)
Posted by Cathy Sarisky on
I'd be willing to work on getting Michael's scraped HTML files parsed.  Thanks to the holiday, I have 2 extra days this weekend.  It appears it would not be TOO difficult to parse that HTML back into entries in the database.
<p>
It seems to me that we need to think about how to handle imported users.  Should all the web/db posters be imported into openacs.org?  If so, we need to match web/db posters with existing oACS posters, and then insert all unmatched names.  Of course, we don't have email addresses to go with those unmatched names.  Or password choices, or...  I'm thinking made-up email addresses (since the field is required) and randomized passwords, with extra work for the openacs.org admins if/when/as the posters migrate over to openacs.org and want to have their accounts merged.
<p>
Of course, if someone can get the actual SQL data (not just a strip of the site), then an import is easier.
Collapse
Posted by Tilmann Singer on
I'd be willing to help with the html parsing of the web/db boards too if there is no sql dump available. I still have the "split a bboard thread in messages" code lying around somewhere from the work on the 3.x -> 4 links migration script.

About the poster's identities: wouldn't it be possible to add them as persons (the acs-object with that name), not users? From a short look at the datamodels it seems that nothing should stop us from inserting postings from persons that have no user account. This seems cleaner then generating fake email-adresses IMHO.

And maybe one day there will even be admin pages that allow an administrator to elevate an existing person entry to a user account. This could be of more general interest: imagine a company that stores all it's client contacts as persons in it's OpenACS system, and upon request can create user accounts for their clients by just typing their email address in an admin form. Has anyone done somehting like that already?

Anyway, if a merge is required there is propably not such a big difference between doing it with two user objects or a user and a person object.

Collapse
Posted by C. R. Oldham on
And maybe one day there will even be admin pages that allow an administrator to elevate an existing person entry to a user account. This could be of more general interest: imagine a company that stores all it's client contacts as persons in it's OpenACS system, and upon request can create user accounts for their clients by just typing their email address in an admin form.

We need that functionality. No committment, but that add on might come from us.

Actually, synching up the accounts between arsdigita.com & openacs.org wasn't hard; each post contains a link to /shared/community-member?user_id=###, where the email address is available. Then I cross referened the aD user_id's to oacs user_id's based on email address, and voila.
Collapse
Posted by Todd Gillespie on
I am, as we speak, slurping Michael's files down the DSL. Concurrent to this, I am brewing a masterful plan for revenge against Talli. Ohh... it will be painful.

Cathy, the last time I had to build something that handled imports like this, we used a 'proxy user' system, where data could be owned by what are essentially shells of users -- the person wasn't registered, and the system could ignore the "user" in many places since it wasn't real, but if that user registered, [s]he would inherit any the data registered in his/her name. Does OpenACS ("oaks") have anything that can be munged into acting like that?

Cathy, would you mind working with me? Or I could just write some db manipulation fxns for you... either way.

Something I brought up on the IRC channel - has anyone looked at the legality? Would lawyers over at RH/aD notice this and take offense whilst they are trying to catalogue their acquired IP? Should we wait?

Collapse
Posted by Graeme Walters on
CR Oldham requested a change to allow an administrator to convert persons into users. On a similar theme, I've submitted a patch which automatically converts persons into users when they register. The user-new pages look for a person with the given email address, and load the name into the input fields. The acs_user__new sql function now only creates the person if necessary, otherwise updates the existing person's object type to 'user'. https://openacs.org/sdm/one-patch.tcl?patch_id=144
Collapse
Posted by C. R. Oldham on
Awesome, Graeme.  Thank you!
Collapse
Posted by Jade Rubick on
Graeme, is this patch pretty secure?

For example, if someone is not on the OpenACS website, but they are registered on the web/db board, what would happen if I tried to create an account as them? I guess I probably would just get all their postings linked to my new account, right?

I guess that's not as bad as it seems. :)

Collapse
Posted by Talli Somekh on
I'm not sure that this question was answered yet, but I wanted to get back to Lars' original inquiry whether we should try to improve the current site or start moving to the new one. If we're going to migrate the web/db boards (and I do think that it might be important to find out what aD/Red Hat thinks before we take their content) shouldn't we migrate it to the new site so we don't have to worry about migrating it again?

If that is so, then it's probably time for http://openacs.museatech.net to move over to the OACS server OpenForce purchased. Maybe run it on new.openacs.org or something? It would be more of a task to migrate a site with a large DB twice than just once, no?

talli

Collapse
Posted by Jade Rubick on
Someone should contact aD's lawyer: Richard Buck
Collapse
Posted by Cynthia Kiser on
Richard Buck is not aD's lawyer. He was the head of engineering. Phil Lotaine is (was?) aD's lawyer.
Collapse
Posted by Don Baccus on
The patch really needs to be OpenACS 4-ized before we can accept it into the tree.  That means:

1. query patches to the query files, not .tcl files

2. PG as well as Oracle (I don't want to start drifting apart on minor patches, we'll never get the worlds aligned again if we start down that path!)

Collapse
Posted by Cathy Sarisky on
Sort of summarizing for my own benefit and so you can tell me where I'm confused...

Todd asked if I'd be willing to work with him: sure, if there's work for us to do!

If I'm reading Michael's post correctly, he has already written some code to do the openacs-web/db import (or at least matched user names to emails, something beyond what was in the downloaded html files) in some way, so perhaps this is done or nearly done already.

I was picturing a script to insert the web/db users into the users table (perhaps as deleted users, but with correct emails?).  But Graeme's patch and importing into oacs4 sounds like a better way to go, if the site migration is going to happen soon.  I haven't done anything with oacs4, so I won't really be able to help with that.

The advantage to inserting deleted users in 3.x (rather than patching 3.x) is that the new data could be imported to OACS4 right along with the users and bboard data, without additional scripting.

I agree that someone should take this up with aD.

So where do we go from here?  Seems like some decision from the gatekeepers as to what tack we take?

Collapse
Posted by Don Baccus on
I don't really expect migration to happen really soon unless someone waves a magic resource wand and frees up a couple of people to finish wrapping it up.

Musea took the lead early on but then had to drift off to work for money (boo-hoo for them, eh?)  I fixed a bunch of Postgres-related performance problems that makes the site practical to run (though I'd like to speed up bboard even more).  But now I'm busy juggling contract work with trying to finalize the OpenACS 4 beta (very close, very close!)

So what really needs to happen is for a couple of folks to step forward and offer to push on.