**** BEGIN LOGGING AT Tue Mar 12 22:59:59 2002
Mar 12 22:59:59 --- Topic for #ampu is Coding a better government. Join us...
Mar 12 22:59:59 --- Topic for #ampu set by ChanServ at Fri Mar 8 22:54:13
Mar 12 22:59:59 -ChanServ- [#ampu] Welcome to AMPU. See http://www.freesoftware.fsf.org/ampu/ for details.
Mar 12 23:00:08 <jel> hi guys =)
Mar 12 23:00:15 <adiffer> hello
Mar 12 23:00:21 <speek> hi
Mar 12 23:01:13 <jel> I was telling steve earlier, they ran an interesting story today on our national news, about e-government etc.
Mar 12 23:01:59 <jel> The reporter didn't even get someone who was opposed to counter. Reporters must be getting bored with the status quo, too =)
Mar 12 23:02:28 <speek> on the web somewhere/
Mar 12 23:02:32 <speek> ?
Mar 12 23:02:48 <adiffer> That means we aren't seen as all that controversial.
Mar 12 23:03:46 <jel> nope, it's a very good sign. This was mainstream news, and the politician they invited to the show was all in favor. Actually, he argued quite well for it =)
Mar 12 23:04:05 <jel> http://www.channel4.com/news/home/20020312/Story07.htm
Mar 12 23:05:05 <jel> Of course, the government's solution to voter apathy is useless.. they want to model themselves after a show wherein people vote for their favorite singer!
Mar 12 23:05:42 <jel> Their argument is that the people voting for the singers are the target audience they need to get voting, so that must be the way. lol.
Mar 12 23:05:53 <speek> there's a big difference between electronically voting for representatives, and voting directly on issues
Mar 12 23:06:05 <speek> lol
Mar 12 23:06:28 <adiffer> politicians understand polling and public opinion. As long as we can gather opinion we have them covered.
Mar 12 23:07:01 <speek> i'm just saying they won't ever want to give up their control
Mar 12 23:07:02 <jel> Yes.. that was the argument in the story. That people won't come back to voting until they feel that their vote can achieve something.
Mar 12 23:08:33 <jel> So anyway.. anyone got new ideas they want to suggest? Or complaints?
Mar 12 23:09:43 <jel> No? That's cool. =) Everyone can manage this time then, without trouble?
Mar 12 23:10:09 <adiffer> This time works OK. I might get called away on a work issue, but it is unlikely.
Mar 12 23:10:33 <speek> this time works for me
Mar 12 23:10:36 <adiffer> I did manage to get an anonymous checout of the html files. I cant get the ssh type checkout to work yet.
Mar 12 23:10:42 <jel> np. We'll adapt to unusual circumstances. Just want to avoid continual probs. =)
Mar 12 23:11:45 <jel> adiffer: Oh, OK. I didn't realise you were having trouble. You followed the CVS docs on Savannah, right? On the project's CVS page?
Mar 12 23:12:18 <jel> Wow.. it's orange =)
Mar 12 23:12:26 <speek> yes, noticed that :-)
Mar 12 23:12:31 <adiffer> Yup. I generated a key and put it up yesterday.
Mar 12 23:12:56 <adiffer> The color change makes it look like the Open Souce Directory now.
Mar 12 23:13:25 <jel> Could be just taking a little longer than advertised. Any useful error messages?
Mar 12 23:14:27 <adiffer> I'm getting asked for a password when I shouldn't be I think. I made the key without a passphrase. After that its just a 'permission denied.'
Mar 12 23:16:31 <jel> democritus2: you have cvs access ok, right?
Mar 12 23:17:08 <jel> oh, steve's not here =)
Mar 12 23:18:25 <jel> hmm,,, his AOL away message says he's doing the dishes, but he always says that =)
Mar 12 23:18:29 <speek> i was planning on trying tonight
Mar 12 23:18:41 <speek> eats a lot
Mar 12 23:19:26 <speek> :-)
Mar 12 23:19:45 <adiffer> 8)
Mar 12 23:20:19 <jel> adiffer: what command are you using for cvs? It should be "cvs -dadiffer@subversions.gnu.org:/webcvs/ co non-gnu/ampu" or similar
Mar 12 23:21:03 <adiffer> I tend to set my CVSROOT variable to save all the typing, but the command is essentially the same.
Mar 12 23:21:44 <jel> sure. OK, that's weird then. Has it been 24 hours since you registered your key?
Mar 12 23:22:29 <adiffer> I get a 'connection refused' from the command line. It should be very near 24 hours now.
Mar 12 23:24:51 <adiffer> oh...There is a use case I forgot to mention. There is usually one high-level one that acts as a context setting.
Mar 12 23:25:38 <jel> oh? can you give me an example?
Mar 12 23:25:57 <jel> working on cvs. It's doing weird things for me too, at the moment.
Mar 12 23:26:08 <adiffer> I'm looking through your use cases right now. Look at your diagrams that have multiple processes in them. There is a good chance each bubble is a use case.
Mar 12 23:26:25 <adiffer> We may need more than one level of context.
Mar 12 23:27:02 <jel> yes, I'd thought that. In fact, many are includes of other use cases alread
Mar 12 23:27:03 <jel> y
Mar 12 23:27:27 <adiffer> Real projects like this could have dozens of them that partially collapse together later.
Mar 12 23:27:44 <jel> nope, cvs works here. Just forgot case on login. What version are you using?
Mar 12 23:28:19 <jel> collapse together? That's interesting. I would have expected them to break into more?
Mar 12 23:28:24 <adiffer> WinCVS 1.0.6
Mar 12 23:28:52 <adiffer> They do initially. As you focus them, though, you will probably notice some are variations on a theme.
Mar 12 23:28:55 <jel> ah, pattern here. Mike was having trouble with WinCVS, too.
Mar 12 23:29:19 <speek> well, i hadn't installed cygwin
Mar 12 23:29:27 <adiffer> I get similar problems from the command line where I bypass WinCVS.
Mar 12 23:30:22 <jel> That makes sense -- about variations on a theme.
Mar 12 23:31:10 <jel> Sorry, let me read back a second.. Japanese drums just broke my concentration =)
Mar 12 23:31:10 <adiffer> I'll sweep through them tonight and tomorrow and let you know how I see them.
Mar 12 23:32:14 <jel> Yes.. so when do we declare use cases focused? When the requirements are finalised
Mar 12 23:32:14 <jel> ?
Mar 12 23:32:56 <adiffer> Facade implies a bunch of place holders to represent stuff even if there are duplications. Most fields are empty.
Mar 12 23:33:14 <adiffer> Filled implies fields filled in and some supporting documents started.
Mar 12 23:33:37 <jel> That would be great. You say the use cases need broken down more. Does that mean you expected more detail?
Mar 12 23:33:47 <adiffer> Focused implies some trimming to control scope and duplication of coding.
Mar 12 23:34:17 <adiffer> Finished implies the creation of non-standard stereotypes and a hand-off to the design team.
Mar 12 23:34:30 <jel> Ah, OK.
Mar 12 23:35:01 <jel> Can you explain these levels of context?
Mar 12 23:35:05 <adiffer> The facade level can actually take a while. Finished level is one of those 'attention to details jobs' and not all that fun. 8)
Mar 12 23:35:50 <adiffer> If you are tempted to break a use case into several subprocesses, you probably have a high-level use case that provides 'context' for the lower ones.
Mar 12 23:36:12 <jel> But by facades, you mean just working out which use cases should be there, and essentially throwing in empty templates?
Mar 12 23:36:20 <adiffer> You write the high-level one and it includes the lower ones.
Mar 12 23:36:57 <adiffer> Most of the facade level ones are just names and summaries. Don't worry about the course of events, triggers, and all that. We get that later.
Mar 12 23:38:20 <speek> does that mean facade, filled, etc refer to different types of use cases, not just use-cases at different stages of completion?
Mar 12 23:38:58 <adiffer> No. They are at different stages of completion. The level is just a status field.
Mar 12 23:39:11 <jel> Ah, OK. We can do that. I hadn't bothered with that level, since most of the events essentially are chosen directly from the UI, so it's fairly obvious, right? No reason we can't do it, even so.
Mar 12 23:39:28 <speek> ok
Mar 12 23:39:51 <adiffer> There are going to be plenty on the server side too. Think about data validation, user validation, and all that.
Mar 12 23:40:38 <jel> speek: he's just saying (if I understand correctly, adiffer?) that we should create use-cases for the higher-level "calling code" too (although, it'll be in terms of the user calling it, not the code itself)
Mar 12 23:41:43 <adiffer> yup. Anything the system must be able to do gets represented in a use case or a non-functional requirement. Use cases cover on the functional stuff.
Mar 12 23:41:49 <speek> yeah, i would have expected all the use cases at this point to be user-centric
Mar 12 23:42:30 <adiffer> Users are the primary actors. There shouldn't be many other actors around.
Mar 12 23:42:39 <jel> adiffer: yup, that's a whole pile of worms I haven't gotten into yet. Most of the use cases already up will be affected by network distribution & some validation.
Mar 12 23:42:40 <speek> logging in, for example, would be a use-case, no?
Mar 12 23:42:59 <adiffer> Logging in is definitely a use case.
Mar 12 23:43:24 <speek> when we try to specify our application, we often talk of "work-flows"
Mar 12 23:43:39 <adiffer> When the system decided the voting deadline has passed and acts, there is another use case for an actor called 'Clock'
Mar 12 23:43:42 <jel> I switched at some point to an event-driven architecture (although it'd been in my head for a while), so that all posted info can be "caught" and distributed, but I thing some work needs done there still.
Mar 12 23:44:38 <jel> adiffer: I figured login would go like "Anonymous: [login]->[get promoted to citizen]" and then Citizen would logout again, would that be OK?
Mar 12 23:44:50 <speek> actor->'Clock'-> i like that
Mar 12 23:45:20 <adiffer> Thinking in terms of events is necessary. Most user driven activity is that way.
Mar 12 23:45:59 <adiffer> I suspect there will be more than one type of login case or some extension points to cover for different actors.
Mar 12 23:46:55 <adiffer> Now that I've had a glance at the current use cases, I would estimate we will wind up around 5 dozen at the facade level.
Mar 12 23:47:17 <jel> adiffer: sure.. I was referring to the aliasing of anonymous->citizen->anonymous again, but I'm sure it's OK, now that I think about it.
Mar 12 23:47:38 <jel> 5 dozen? probably right. But don't say that! =)
Mar 12 23:48:13 <adiffer> The better they are written, the easier time you will have explaining to people that want to help how to help. 8)
Mar 12 23:48:30 <jel> Yup, it's gonna be complex enough, without having bad docs.
Mar 12 23:48:48 <jel> So is 5 dozen bad? =)
Mar 12 23:49:15 <adiffer> 5 dozen is probably a small start considering we are planning to take over the world.
Mar 12 23:49:21 <speek> how do you represent an event's trigger logic?
Mar 12 23:49:28 <jel> hehheh.. nicely put =)
Mar 12 23:50:04 <adiffer> The trigger is described in words within each use case. It is usually a one liner. If it is more, there is a good chance the use case should be split.
Mar 12 23:50:12 <jel> speek - not sure what you're asking. I'd figured event queues, both system-wide and local, with listeners
Mar 12 23:50:28 <jel> ah, UML =)
Mar 12 23:50:31 <speek> ok, but what is the actor if the logic is something other than time?
Mar 12 23:51:30 <adiffer> ? If I post a response to a debate topic, I am the actor that triggers 'Post a Response'.
Mar 12 23:52:05 <jel> speek: there is a "System Automaton" actor which does most general housekeeping. Other actors will just listen for events themselves, I guess, so will be whatever is appropriate to their overall task - not "Event Listener", for example.
Mar 12 23:52:27 <speek> ok, so 'system' is an actor too
Mar 12 23:53:02 <adiffer> System is the one actor that gets fully described by the documentation. The other actors are assumed to be atomic.
Mar 12 23:53:14 <jel> Sure. We'll try to break it down into individual roles that the system performs (database manager, network corresponder, whatever)
Mar 12 23:54:40 <jel> adiffer: oh? Didn't know that. What would System docs look like?
Mar 12 23:54:41 <adiffer> For my other project I have an actor known as 'Notification Tool'. It handles all outbound communcations handled by other applications. The System formats messages and then gets out of the way.
Mar 12 23:55:31 <adiffer> We might be using 'System' in different ways. The System is everything we are building. There are actors external to it.
Mar 12 23:55:38 <jel> adiffer: yes, that's much like I'd planned, too. Post events, let the event scheduler fire them off, and maybe have custom event listeners route them to remote servers
Mar 12 23:55:39 <speek> sounds like we end up describing a services oriented framework
Mar 12 23:56:40 <adiffer> For actors it is all about services. The system sees it as events.
Mar 12 23:56:45 <speek> (which is good, i think)
Mar 12 23:57:01 <jel> the whole system is essentially a hierarchical database, with events, and a client to track and generate those events, as I currently see it.
Mar 12 23:57:24 <adiffer> sounds good to me.
Mar 12 23:58:10 <jel> Seems like the cleanest (most minimalistic) way to do things. Maybe we'll need to remodel at some point, though =)
Mar 12 23:58:55 <adiffer> Your model will probably jump out at you after the requirements are fully written. Mine did.
Mar 12 23:59:02 <speek> maybe we should brainstorm those 5 dozen facades?
Mar 12 23:59:36 <jel> So no one has noticed holes in the use cases done so far? They seem to cover what's required of them?
Mar 12 23:59:37 <adiffer> I agree. One or two line summaries help clinch meanings.
Mar 13 00:00:10 <speek> i agree with adiffer in that those bubbles will need further breakdown
Mar 13 00:00:18 <jel> ok.. let me try and think what hasn't been done yet...
Mar 13 00:00:38 <speek> we could do that, or try and complete the high-level contexts first
Mar 13 00:01:02 <jel> There is a whole network architecture missing, but maybe that should be left to the end, since it's possible to build upon a standalone server later? Maybe even leave it until version 2?
Mar 13 00:02:10 <speek> isn't that an implementation issue?
Mar 13 00:02:51 <jel> well, to a point, yes. That's what I'm saying. *Unless* there is some need to think about it's interoperation now.
Mar 13 00:02:51 <speek> how about "join a forum"?
Mar 13 00:02:52 <adiffer> We can write requirements for it and then mark it as out of scope for version 1. That is what the focused state is really for as it finishes up.
Mar 13 00:04:04 <jel> OK, "Subscribe to a Forum", fairly self-explanatory, as is "Unsubscribe from a Forum"
Mar 13 00:04:25 <jel> "Create a Forum"
Mar 13 00:04:26 <speek> ya, it's probably not the same scope as the others
Mar 13 00:04:32 <speek> good one
Mar 13 00:04:59 <speek> "Modify User Preference", might be the larger context that includes "Join a Forum"
Mar 13 00:05:56 <adiffer> Create New User
Mar 13 00:06:03 <jel> "Archive a Forum" - Timeout-actived task to declare a forum outdated if no one uses it (Subscribes/Posts) in a certain period.
Mar 13 00:06:23 <speek> so we want a moderation system? - "rate comment"
Mar 13 00:06:33 <jel> Yup. Births and Deaths, I'd thought of callign them, but maybe that's undiplomatic? =)
Mar 13 00:07:09 <adiffer> I'd like a rating system. It helps we deal with the signal to noise ratio.
Mar 13 00:07:24 <jel> Yes.. probably right.
Mar 13 00:07:58 <speek> i agree
Mar 13 00:08:10 <adiffer> What is the difference between Create Initiative and Create Forum?
Mar 13 00:08:16 <jel> Any reason not to build automatic moderation on top, too.. for things like swearing in underage forums?
Mar 13 00:08:24 <speek> and that way we can have meta-moderation forums
Mar 13 00:08:30 <speek> :-|
Mar 13 00:08:48 <jel> adiffer: Initiative is an issue for decision-making - what leads to a Jury.
Mar 13 00:08:51 <speek> Forums are categories, initiative's are specific "stories"
Mar 13 00:09:28 <adiffer> So you make an initiative from within a forum?
Mar 13 00:09:33 <jel> Yes, but there should be normal conversational messages as well as Initiatives/stories
Mar 13 00:09:39 <speek> automatic moderation doesn't fit with a democracy system, does it?
Mar 13 00:09:59 <adiffer> automatic moderation will lead to cultural biases.
Mar 13 00:10:17 <jel> adiffer: yup. Forums are were the subscribed group of users interested in the Initiatives come from.
Mar 13 00:10:26 <adiffer> oops. Gotta make a call. I'll be back.
Mar 13 00:10:45 <speek> how about "move initiative"
Mar 13 00:10:56 <-- adiffer has quit ()
Mar 13 00:11:08 <jel> Yes, I worried about moderation a little, in all forms. But if it's allowed at all, then it seems sensible to allow it to be done automatically.
Mar 13 00:11:43 <speek> moderation is just for sorting purposes, IMO - not to hide comments
Mar 13 00:11:45 <jel> I mean, if you're going to mod-down everything that contains bad language, then why not have the system look for bad language and mod it down for you?
Mar 13 00:12:03 <jel> ahh, that's a good point speek.
Mar 13 00:12:13 <speek> there's a difference between a user modding down, and the system modding down
Mar 13 00:12:54 <speek> I was joking about the meta-moderation forum, but automatic mod will make that forum explode
Mar 13 00:13:09 --> adiffer (~adiffer@user-2ivfo8p.dialup.mindspring.com) has joined #ampu
Mar 13 00:13:19 <speek> "hell" should be worht -1, "fuck" - 5, etc
Mar 13 00:13:52 <adiffer> Is this thing going to be English specific?
Mar 13 00:13:59 <jel> OK.. then we'll have modding "fold" posts, so you can zoom in to see them? ...
Mar 13 00:14:13 <speek> sure, that's a presentation issue
Mar 13 00:14:27 <speek> i don't see why it'd be language specific
Mar 13 00:14:53 <adiffer> What does fsck get?
Mar 13 00:15:00 <speek> +3
Mar 13 00:15:11 <speek> "joint" gets -100 though
Mar 13 00:15:35 <adiffer> Should be fun talking about joint task forces then. 8)
Mar 13 00:15:37 <jel> Then, should we allow automatic flagging of messages that contain certain keywords (like a spam filter adds *SPAM* to noticed email (so you can tell your email client to delete anything with *SPAM*) ?
Mar 13 00:15:55 <speek> small price to pay to win war on drugs
Mar 13 00:16:27 <adiffer> Letting people mark comments is fair so others can use the flags.
Mar 13 00:16:28 <speek> jel - i'm being sarcastic
Mar 13 00:16:33 <jel> adiffer, not language specific.. thinking of Forums where people can specify that SHIT isn't allowed.
Mar 13 00:16:59 <speek> i can't see that such censorship will help
Mar 13 00:17:12 <speek> simple modding seems to be sufficient
Mar 13 00:17:36 <adiffer> I would suggest that such features come later. They are hard to automate in a consistent way. Humans flagging stuff is easy.
Mar 13 00:17:40 <speek> why are you talking about email?
Mar 13 00:18:19 <jel> Really? OK, I was just trying to add adaptability, but I know where you're coming from, and actually agree in principle, so we'll drop that.
Mar 13 00:18:40 <adiffer> So we have a 'mod comment' use case?
Mar 13 00:18:46 <speek> ya
Mar 13 00:19:10 <jel> yup, "mod comment" =)
Mar 13 00:19:16 <adiffer> I like the folding approach better than the hiding one on K5.
Mar 13 00:19:37 <speek> and "move initiative" from one forum to another
Mar 13 00:20:15 <jel> Yea.. that one's speeks idea. Makes sense. Just because it's not the best comment, doesn't mean it should disappear off the face of the planet =)
Mar 13 00:20:20 <adiffer> copy initiative?
Mar 13 00:20:24 <jel> I'd love to read the old stories, too.
Mar 13 00:20:33 <speek> both actually
Mar 13 00:20:42 <speek> sometimes, other forums may be interested
Mar 13 00:20:56 <speek> sometimes an initiative belongs elsewhere
Mar 13 00:21:15 <adiffer> withdraw initiative?
Mar 13 00:21:22 <speek> cancel initiative is already there
Mar 13 00:21:22 <jel> adiffer: I'd thought more of just having an import option during initiative creation. I guess it would still be copied in though?
Mar 13 00:21:53 <speek> you wouldn't want a copy, you'd want a reference, right?
Mar 13 00:22:12 <adiffer> jel: There you are. That is exactly what would pop up when focusing all these. The common threads would come together and use cases would collapse into each other.
Mar 13 00:22:14 <speek> so that discussions from both forums were visible to both
Mar 13 00:22:49 <adiffer> So an initiative could be running in more than one forum?
Mar 13 00:22:55 <speek> right
Mar 13 00:22:56 <jel> speek: don't know. Most likely copy some non-unique fields, bring up an editor for the others. Depends what exactly is in an initiative though.
Mar 13 00:23:17 <speek> oh, you're talking using an existing initiative as a template for a new one
Mar 13 00:24:21 <jel> speek: yes. I think different initiatives, although following the same solution model, would be too dissimilar to mix up in conversation.
Mar 13 00:24:42 <speek> it would be nice to be able to "merge" initiatives - to avoid unnecessary duplication
Mar 13 00:24:57 <speek> if a user notices two initiatives are basically the same, the discussions and juries should be merged
Mar 13 00:25:11 <jel> adiffer: that's possible, but a different issue...
Mar 13 00:25:21 <jel> speek: let me think of an example..
Mar 13 00:25:35 <jel> I'm thinking quite abstractly here..
Mar 13 00:25:41 <adiffer> To merge initiatives, would there have to be a subiniative for voting purposes?
Mar 13 00:25:53 <speek> i would think yes
Mar 13 00:26:07 <jel> say you've got a problem, of managing funds within your large organisation. You're stuck.
Mar 13 00:26:08 <speek> just about everything any user asks the system to do, would require an initiative
Mar 13 00:27:11 <speek> which brings up a problem - the system could get overwhelmingly cumbersome
Mar 13 00:27:13 <jel> then you realise that someone managed their tool maintenance in a large workshop using a method that might work. You could import the basic structure of the solution, but change the details.
Mar 13 00:27:40 <speek> right, use it as a template for your intiative and the solutions
Mar 13 00:28:03 <speek> but, it's got no connection to the original initiative as far as the system is concerned
Mar 13 00:28:21 <jel> Yes, but then.. you wouldn't want to discuss your fund management in the same conversation as the tool maintenance.
Mar 13 00:28:50 <speek> no, exactly - but what your describing isn't what i was talking about with "move" and adiffer with "copy"
Mar 13 00:29:09 <speek> it's just a convenience tool for introducing a new initiative
Mar 13 00:29:21 <jel> Yes, adiffer is thinking along the lines of usenet cross-posts ?
Mar 13 00:29:26 <speek> yes
Mar 13 00:29:48 <adiffer> Keystroke savers aren't a big worry. We can write them as we think of them.
Mar 13 00:29:52 <speek> sometimes, an initiative will be posted to multiple forums initially
Mar 13 00:30:11 <adiffer> Cross-posts that keep active in both forums are another thing. They lead to joint juries.
Mar 13 00:30:27 <speek> sometimes, users will want them multi-listed later, after they were created
Mar 13 00:30:28 <jel> "Create an Initiative" mentions cross-posting as an extension. All quite vague, though.
Mar 13 00:31:01 <speek> and sometimes, the initiative is just in the wrong place, and needs to be "moved"
Mar 13 00:31:23 <adiffer> hmmm.. How about creating a new forum that contains two others so the new iniative only lives in the top level one?
Mar 13 00:31:53 <speek> what's hard about linking to the same initiative from two different forums?
Mar 13 00:32:33 <jel> adiffer: it's probably going to be as easy to do with normal cross-posting, but here's what I said in the use case :
Mar 13 00:32:33 <speek> the system could handle it that internally, but you wouldn't want to bother the users with that
Mar 13 00:32:34 <adiffer> This whole issue breaks the hierarchy of the forums either way... 8/
Mar 13 00:32:38 <jel> * Cross-posting should be possible at [Step 3], but some thought is required regarding the overall structure of the system in such cases -- for example, should Juries be selected from a cross-section of all Forums involved? Although relevant enough to hear about the issue, is every Forum relevant enough to partipate in the decision? Another solution might be to have one main Forum for the Initiative, but announcement to Forums considered rele
Mar 13 00:32:38 <jel> vant.
Mar 13 00:34:00 <speek> we'll end up with every initiative in it's own forum, and initiatives beyond "advertised" to different groups to participate
Mar 13 00:34:05 <speek> which might be better anyway
Mar 13 00:34:31 <adiffer> hmmm.... More thought is needed. Let us cogitate a while.
Mar 13 00:34:32 <speek> after all, just cause I'm in the IP Law forum doesn't mean I want to participate in every initiative that goes there
Mar 13 00:34:50 <speek> s/beyond/being
Mar 13 00:35:48 <speek> how about this: users subscribe to groups
Mar 13 00:35:53 <jel> Essentially, everything is just nodes in a hierarchy, as I see it. Initiatives are just extended top-level messages posted to a forum. Responses to an initiative are very similar to replies to a message, etc. So Forum vs Initiative isn't really an issue. Forums are just a category for grouping messages, and allowing people to register an interest in them by subscribing.
Mar 13 00:36:02 <speek> initiatives generate forums to live in
Mar 13 00:36:25 <speek> initiatives are advertised to users of certain groups who can then join the initiative's forum, or not
Mar 13 00:36:44 <jel> speek: group == Forum
Mar 13 00:37:09 <speek> speek breaks (group == forum)
Mar 13 00:37:18 <speek> suggests otherwise
Mar 13 00:37:30 <jel> adiffer: you seemed to have concern earlier about Forums vs Initiatives, too. How come?
Mar 13 00:38:15 <speek> the thing is, a jury belongs to an initiative, not a forum
Mar 13 00:38:19 <adiffer> Just term confusion on my part. I was trying to clarify definitions. I should read the dictionary. 8)
Mar 13 00:38:50 <speek> but, juries are selected from forums
Mar 13 00:39:26 <speek> i'm not explaining this well
Mar 13 00:39:37 <speek> wish i could draw a picture
Mar 13 00:40:39 <adiffer> 8)
Mar 13 00:41:11 <speek> i guess i'm saying the system would look simpler if we kept a rigid hierarchy of forum->initiative->jury
Mar 13 00:41:25 <speek> disallow an initiative from belonging to multiple forums
Mar 13 00:41:36 <adiffer> speek: I tend to agree.
Mar 13 00:41:43 <speek> disallow juries from being selected from multiple forums, but then belonging to only one initiative
Mar 13 00:41:46 <speek> that complicates things
Mar 13 00:41:56 <speek> users are in groups, and that separate
Mar 13 00:42:18 <speek> when an initiative is created, the creator selects groups to "advertise" the new initiative to
Mar 13 00:42:35 <jel> speek: hmmm at the moment, Initiatives, Juries, and Taskforces all "belong" to the Citizens in the Forum which begins the process. I thought of one problem with that (not a big one), but I've forgotten right now.
Mar 13 00:42:37 <speek> a new forum is created, and the initiative goes in the forum
Mar 13 00:43:00 <speek> i don't see how these things belong to citizens
Mar 13 00:43:30 <speek> users can then join the forum or not, and thereafter, everything is as you have been thinking all along
Mar 13 00:43:45 <speek> but without potential of complications from cross-posting
Mar 13 00:44:16 <speek> if those in the initiatives forums decide, they can always re-advertise the initiative to more grouops
Mar 13 00:44:43 <jel> Because all citizens have a right to track them, I had figured they should be visible right there. I guess that's not a very good reason, though, for cluttering. They can always expand the Initiative somehow.
Mar 13 00:45:20 <speek> but the way in which citizens can "browse" and "search" for initiatives is a separate issue
Mar 13 00:45:45 <jel> speek: yup, that's what I said in the use-case, about advertising, instead of making the issue *available* in multiple Forums
Mar 13 00:46:30 <speek> right, but if i join another forum just to take part in one initiative, the system will start spamming me about every initiative that gets posted to that forum
Mar 13 00:46:34 <speek> maybe i didn't want that
Mar 13 00:46:43 <jel> sure, there are presentation aspects, but we want a sensible internal design, too.
Mar 13 00:47:29 <speek> the system will record which groups the initiative had been advertised to
Mar 13 00:47:37 <jel> True. Then options become subscribing to Initiatives, or having personal homepages, I guess?
Mar 13 00:47:45 <speek> that's equivalent to it belonging to multple forums
Mar 13 00:47:52 <speek> for search purposes
Mar 13 00:48:06 <jel> speek: subscribing to initiatives is, you mean?
Mar 13 00:48:39 <speek> uh, no
Mar 13 00:48:47 <jel> no, I get you now.
Mar 13 00:48:54 <speek> ok, *whew*
Mar 13 00:49:17 <speek> i think personal homepages will make a lot of sense
Mar 13 00:49:23 <speek> just like on savannah
Mar 13 00:49:32 <speek> it keeps track of just about anything you want it to
Mar 13 00:50:21 <jel> heh.. message from my hosting company: This afternoon we experienced a RAID array failure on our primary
Mar 13 00:50:21 <jel> fileserver.
Mar 13 00:50:21 <adiffer> homepages lead to another whole set of use cases. Typical ones.
Mar 13 00:50:23 <jel> =)
Mar 13 00:50:51 <speek> adiffer:yes - but wouldn't they all be part of "setting user preferences"?
Mar 13 00:51:01 <speek> stupid emoticons...
Mar 13 00:51:14 <jel> adiffer: true. Just do it the usenet way, then, with a filtered list of subscribed groups?
Mar 13 00:51:47 <adiffer> Set User Preferences would be a high level version containing lower level cases.
Mar 13 00:51:56 <jel> speek: no, you'd have "Display User's Favourite article", "Display User's local Time", etc =(
Mar 13 00:52:13 <speek> ok, here's a use case - "user goes to homepage" - and then you'll have lots of - oh, you got it
Mar 13 00:52:28 <speek> :-)
Mar 13 00:52:53 <speek> adiffer - yes, i'm just trying to see the top-level issues
Mar 13 00:53:31 <adiffer> Look at it from the non-technical user's perspective. Lots of detail. We as designers and developers can group them together as we need.
Mar 13 00:53:31 <speek> i've been doing web stuff so long, i see everything in terms of the navigation of the site
Mar 13 00:53:47 <jel> OK, so.. if we're changing to a system were you can subscribe to individual Initiatives, and Jurors subscribed to Juries, then we may as well have watched messages, and build them all upon that, right?
Mar 13 00:54:14 <jel> what about if we show them a pretty picture of a homepage, but it doesn't work? ;)
Mar 13 00:54:21 <speek> watched/joined, yeah
Mar 13 00:54:36 <speek> that'll be our prototype
Mar 13 00:55:18 <jel> no, adiffer's right about homepages for now. Too much work, too little gain. It's nothing new, no point working on it right now.
Mar 13 00:55:48 <jel> Maybe in v2, along with the distributed networking stuff, etc =)
Mar 13 00:56:05 <adiffer> Shouldn't we be able to copy homepage stuff whole sale from other GPL'd projects?
Mar 13 00:56:16 <speek> well, homepage is just another name for index page plus dynamic information from a databse
Mar 13 00:56:44 <speek> when i log in, i better be able to jump to my forums!
Mar 13 00:57:24 <speek> jel: do you have a use case that indicates a need for distributed network stuff?
Mar 13 00:57:25 <adiffer> speek: So we find something close and change a few names and queries. That's all that web programming stuff is right? 8)
Mar 13 00:57:27 <jel> Oh, yes.. there'll have to be some basic form of listing your subscribed groups when you log in. That's definite.
Mar 13 00:57:49 <speek> that's really all im talking about
Mar 13 00:58:08 <jel> OK, I was just saying that we don't want to get into extra nicities right now =)
Mar 13 00:58:12 <speek> sure, later it can become more powerful, customizable, yada yada
Mar 13 00:58:30 <speek> agree totally
Mar 13 00:59:17 <adiffer> I'm looking at use case 10, 11. The names are confusing and probably too technical.
Mar 13 00:59:22 <jel> speek: I'd basically left the whole networking thing until later. It does mention distributing events quite a lot, though.
Mar 13 00:59:51 <speek> but distributing to what/whom?
Mar 13 01:00:16 <speek> if it's the system just talking to itself, that doesn't imply distributed networking
Mar 13 01:00:32 <speek> adiffer: yeah, i agree there
Mar 13 01:00:43 <jel> adiffer: hmmm... now that you mention it, 10 is a little scary =)
Mar 13 01:01:03 <adiffer> jel: Written by a developer... 8)
Mar 13 01:01:39 <jel> It's quite a low-level component, though. I still right it in plain english? Aghhh!! But that's imprecise! ;)
Mar 13 01:01:53 <adiffer> I've got to disappear on you all. I have a conference call I'm expected to make here in a few minutes. It's for my startup business, so its exciting stuff.
Mar 13 01:01:57 <speek> plain english is great at this level, i think
Mar 13 01:02:15 <speek> i suggest we call it for the evening/day/night
Mar 13 01:02:20 <jel> No probs. We should wrap this up for now. Two hours is long enough for weekly IRC =)
Mar 13 01:02:26 <speek> ya
Mar 13 01:02:42 <adiffer> very good then. This is all logged I hope?
Mar 13 01:02:49 <jel> OK, I'll work on this stuff, especially the new Initiative subscriptions, etc..
Mar 13 01:02:54 <jel> Yup, got it logged.
Mar 13 01:03:01 <speek> great - later!
Mar 13 01:03:11 <jel> OK guys, thanks a lot. Great working with you both =)
Mar 13 01:03:12 <adiffer> OK. See you all next time then.
Mar 13 01:03:26 <jel> bye =)
Mar 13 01:03:33 <speek> bye
Mar 13 01:03:36 <-- speek (~speek@cf1.ext.eastgw.xerox.com) has left #ampu
**** ENDING LOGGING AT Wed Mar 13 01:03:42 2002