Our database is hosed+discussion
While I obviously already know enough to be dangerous, I want to know more ;-) Is there a wagn console that I need access to to clean up messes like this?
Nothing built-in, though I often build consoles when Wagn structures get complex enough to warrant that.
Based on what you wrote, the lower-level SQL database is probably fine, it sounds like some re-Wagneering could probably fix things up pretty quickly. I could take some time with you in the next day or so to grok situation and straighten it out. Do you have Skype? That would let us share screens, which would probably help a lot.
--John Abbe.....Mon Aug 23 09:50:23 -0700 2010
Hey John, I do have Skype but have limited bandwidth at home and an ISP "rule" not to Skype. Would love to try and work on this with you though.
--Peter Kindfield.....Mon Aug 23 09:59:45 -0700 2010
So John, should I call you? When?
--Peter Kindfield.....Mon Aug 23 15:58:24 -0700 2010
John, when you say you build consoles, you just mean with cards, right?
--Ethan McCutchen.....Tue Aug 24 09:27:27 -0700 2010
do I understand correctly that you changed "friends" to "any user+friends"? And then what happened? What is the result that you're going after? What are the current problems?
--Ethan McCutchen.....Tue Aug 24 17:33:20 -0700 2010
(should have a chance to look at this tomorrow if you guys want a hand)
--Ethan McCutchen.....Tue Aug 24 17:33:53 -0700 2010
I had a card called friends that was defined as a pointer and included in user cards via user+*type+*content. I changed Friends to be a basic card. But user+friends still acts like the old pointer card. I also tried removing peter_kindfield+friends thinking it would be recreated from Peter_Kindfield and the new friends card but I get a card doesn't exist message.
--Peter Kindfield.....Wed Aug 25 04:46:10 -0700 2010
Well, nothing that you're saying sounds surprising to me, so this must be a case where I'm not understanding what behavior you're expecting because my own expectations have been biased by years of working with Wagn.
First, to explain what is happening:
The "friends" card itself is not a setting card, so the type of "friends" card itself doesn't have any impact on the type of any other cards, including those ending in +friends. Cards ending in +friends are represented by the Set "friends+*right". Those cards are shaped by settings connected to that set (or other applicable sets). The two settings that could impact the *types* of cards in that set are *default and *content, both of which can impact the *initial* type of new cards in that set. But neither of them will change the type of existing cards.
So then it sounds like you deleted Peter Kindfield+friends and were expecting that the card would get automatically regenerated. That doesn't happen, but you should be able to click to create it if "+friends" is on the User card. In fact, if you re-create it, the old content value should still be there in the revision history.
As things currently stand, if you were to delete the Peter Kindfield card, then it would warn you that this could lead to deleting all the cards plussed to it, because they all depend on that card. You can't have A+B without A and B. However, you *can* have A and B without A+B. Does that make sense?
I think there may be further confusion, though, and that's what I'm not understanding yet...
--Ethan McCutchen.....Wed Aug 25 13:39:00 -0700 2010
Ah........... Believe it or not, I just finally got that A+B is a card in the database, not the real-time joining of two cards retrieved from the database !!!!
--Peter Kindfield.....Wed Aug 25 15:35:15 -0700 2010
I should have mentioned yesterday that there are cases where we create A+B dynamically (like relative searches). It's easy to tell which is which by looking at the actual card in open view. If it says "Virtual" in the right corner, then it isn't really in the database.
The idea is that sometimes we have all the information we need to render a card (like in relative searches), so we don't have to create a new one. But if the card actually contains some new information (like most pointers), then there is an actual card in the database to store it.
--Ethan McCutchen.....Thu Aug 26 08:35:00 -0700 2010