[03:17] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: Must specify both project name and product name to identify a distinct product (patch-774)
[03:24] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: changed bug listing to list package/product assignments, with corresponding status, severity, priority, etc. (patch-775)
[03:36] <stub> BradB: If you are still around, give me a ping when everything you want has been merged - I'm doing a dogfood update today.
[04:02] <BradB> oh, pqm stopped hanging...
[04:03] <BradB> s,hanging,looping,
[04:03] <BradB> i guess i have to resubmit my last merge request though
[04:04] <BradB> oh, shit, i didn't want to resubmit
[04:04] <BradB> patch-775 is what i wanted in there
[04:06] <BradB> stub: by tomorrow i'm hoping to land a few of the search widgets on the bug listing. i banged my head for a while on trying to figure out how to make a search form using a schema. something like browser:schemadisplay, but Zope 3 certainly wasn't trying to help me out on this task.
[04:08] <stub> Did you work it out? I haven't tried generating or handling forms from schemas apart form using the addform/editform gumf. 
[04:09] <BradB> no. i gave up. i might putzky around when i get back from the grocery store though. a mix of actually trying to make things work slash grokking more of zope 3.
[04:11] <BradB> stub: oh, that reminds me.
[04:11] <BradB> stub: i was going to ask...i think we need to add owners to assignments.
[04:12] <BradB> the bug listing now shows assignments, but since there's no assignment/owner/displayname, all the user'll see in that column is "-" for now
[04:12] <stub> If you have found a use case, sure (or even if you havn't - nice information to have even if it is never displayed or used)
[04:12] <BradB> stub: it's always displayed. check the bug listing (which is really a listing of assignments)
[04:14] <stub> So who should become the owner of all the existing assignments? I might set them to the owner of the bug.
[04:15] <BradB> stub: i think setting them to the bug submitter (i.e. bug "owner" :) is sufficient at this early stage.
[04:21] !alindeman:*! Quick services work so I can fix some b0rked access list; downtime should be less than a minute
[05:26] <BradB> stub: hm, i guess Zope 3 is missing a browser:form directive.
[05:28] <BradB> i want a directive where i can say: for this name, on this object, render a form using this schema.
[05:28] <stub> Similar to registering vocabularies by name? Love those flat namespaces ;)
[05:28] <stub> oic - used to attach views to objects
[05:28] <stub> That would be cool
[05:29] <BradB> well, to make a form that does something that isn't adding or editing.
[05:29] <BradB> i'm not sure what browser:schemadisplay does, because i haven't actually implemented an small example to try it out, but at best, it's very oddly named.
[05:29] <stub> editform sort of works similarly to what you want already
[05:30] <stub> display indicates render-only to me, although that would be pointless because a normal boring view attached to that interface would be just as good...
[05:31] <stub> (unless it provides a standard template)
[05:31] <BradB> stub: it's possible that browser:addform, browser:editform and browser:form could be collapsed into one directive. from the perspective of someone with a human-sized brain, i really hate when five different variations are given for doing the same conceptual thing, due to implementation details bubbling up to the UI.
[05:32] <BradB> for me those are all "creating a form"
[05:32] <stub> addform and editform would be difficult to collapse, because one is a view on an object that already exists and the other is a view on something else that lets you create an object. They both would benefit from having an abstracted out 'form' though.
[05:33] <BradB> i don't really give an ass that one happens to add an object, one happens to make changes to an object, and one submits a form to be emailed
[05:35] <BradB> it was slightly to discouraging to see /Stephan Richter/ saying earlier that Zope 3 is too hard. :)
[05:39] <stub> I suspect it might be - way too much abstraction in all those interfaces. Too many tiny chunks that nobody uses make the bigger chunks you want to use more confusing.
[05:40] <stub> (IVocabulary is defined in IBaseVocabulary, IIterableVocabulary and some others - but nobody uses anything by IVocabulary)
[05:41] <BradB> stub: not enough development going on by the people that use it to deliver things that work on billable time, i fear.
[05:42] <stub> Mmm...
[05:45] <BradB> but somehow it still feels no worse than what i've developed web apps with before (e.g. webware is not much more fun when you realize, "oh crap, i need to write my own perms system")
[05:46] <BradB> and, well, WRT the complaints stephan himself has, 1. we've known for a long time that TTW schemas are on the way and 2. if stephan's not happy with something, that probably means it'll be improved soonish.
[05:48] <stub> Its more fun than Zope2 anyway ;)
[05:48] <BradB> hell yeah
[05:49] <BradB> as complex as it sometimes feels, i'm hooked
[05:50] <stub> I've lived a pretty isolated life and only really played with CGI, Zope2, Zope3 and some Java servelet stuff
[05:51] <BradB> i've done CGI, HTML::Mason, Webware, Zope 2, Zope 3, mod_perl. i think that's all of them.
[05:56] <stub> Oh - a small stint maintaining embedded perl, which was real scarey ;)
[05:56] <BradB> heh
[05:56] <BradB> hm, what's the name of an adapter factory that creates an adapter that does nothing?
[05:57] <stub> eh?
[05:57] <BradB> i want to adapter I1 to I2, I2 is a subset of the attributes of I2, so nothing is needed to actually adapt.
[05:58] <BradB> s/attributes of I2/attributes of I1/
[05:58] <stub> There is an implicit adapter, in that if you IFoo(obj) and obj already provides IFoo then you just get obj back
[05:58] <BradB> yeah, but this is with b:schemadisplay
[05:58] <BradB> it's telling me it can't adapt my I1 to an I2
[05:59] <BradB> so i have to say that I1 provides I2, I guess
[05:59] <BradB> instead of registering a "null" adapter to achieve that
[05:59] <stub> yes. Or have the l1 interface inherit from the l2 one
[06:01] <stub> oh fuck off
[06:02] <BradB> heh, that was me. i love that bug.
[06:02] <stub> pqm is not a labomized monkey, it is an autistic one that like to sit in the corner banging its head against the wall ;-(
[06:02] <BradB> heh
[06:02] <stub> lifeless: pqm ping
[06:03] <stub> elmo: pqm ping
[06:03] <BradB> stub: i thought you killed the pqm infinite loop earlier? because earlier it was looping on your merge.
[06:03] <stub> not me. Just happened that the patch I submitted before I went to bed got processed just after I had my morning coffee.
[06:04] <stub> I don't have that sort of access to chinstrap
[06:04] <BradB> so yeah, b:schemadisplay simply displays those fields.
[06:05] <BradB> wee
[06:05] <stub> I think you can build what you want using editform to render and validate a form, which can later be wired up into a .zcml directive.
[06:07] <BradB> yeah, i was thinking of doing it that way earlier. a bit of a dirty workaround until browser:form is implemented, but it'll do, i suppose.
[06:07] <stub> form = SchemaForm(IFoo, default_ifoo)
[06:07] <stub> form.render(request)
[06:08] <stub> result = form.handle(request) # returns an object() providing IFoo with attributes all set
[06:08] <BradB> sweeeet.
[06:08] <stub> The zcml directive would just register a SchemaForm with a givenname and template
[06:09] <stub> Oh... need a form.validate in there, and request can be passed into the constructor
[06:10] <BradB> is this science fiction, or you giving an example of what to put in the view?
[06:11] <stub> It is science fiction, but I think it can be worked up fairly easily using the existing editform
[06:11] <BradB> yeah, it would seem that way
[06:13] <stub> But not this weekend ;)
[06:13] <stub> Oops... need to head off before the shops close...
[06:24] <stub> There is a patch in the queue to add owners to *BugAssignment and the Malone & test case fallout btw.
[11:44] <dilys> Merge to 	rocketfuel@canonical.com/launchpad--production--1.3: cherry pick fixed ArchiveLocation instantiation (patch-1)
[11:52] <SteveA> don't forget to put on your lens
[12:51] <sabdfl> morning, ish
[12:57] <SteveA> hi
[12:58] <SteveA> a couple of python wizards are arriving at my place today, to stay in the spare room for a week or two.
[12:58] <SteveA> There's a pypy (python-in-python) sprint in vilnius.
[01:18] <sabdfl> cool
[01:49] <SteveA> If my ubuntu CDs arrive before they leave, I'll spread a few around, and give some to Laura Crighton to take back to Strakt.
[01:49] <SteveA> if not, POV have a cd burner.
[03:52] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: *BugAssignment.owner (patch-776)
[04:02] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: polish malone (patch-777)
[05:07] <BradB> morning
[05:07] <BradB> SteveA: ping
[10:24] <SteveA> hi Brad
[10:24] <BradB> hi
[10:26] <BradB> on phone :)
[10:34] <BradB> back!
[10:34] <BradB> maybe i should take this to #zope3-dev
[11:26] <SteveA> BradB: I reckon I can look at doing this form thing early next week.  Do you think it would speed up development?
[11:27] <BradB> SteveA: yes. e.g. for the bug search form, it takes too much effort to hand-code HTML widgets + validation.
[11:27] <SteveA> yeah
[11:27] <SteveA> I did do something a bit like this for my first zope3 project
[11:28] <SteveA> but it was what I ended up near, after hacking / customizing / forking edit-form
[11:28] <SteveA> and it was a mess
[11:28] <BradB> eek
[11:28] <SteveA> so, I think your ideas of doing it afresh, and simply, are better
[11:28] <BradB> in the abscense of this, i'll probably have to hand-code HTML, but it'll just cost more money to maintain until a make-my-life-easier browser:form directive exists
[11:29] <SteveA> it would be useful to me if you hand-code this one in HTML.  then, it becomes the first candidate for replacement
[11:29] <SteveA> but it provides a real first example
[11:29] <BradB> yes, i will, because i want something working this weekend no matter what.
[11:29] <BradB> (e.g. even if hand-coding is necessary, which for now it is)
[11:29] <SteveA> ok.  let's talk about it more next week, when you've done the form, and I've emptied some of my todo list
[11:30] <BradB> sure
[11:30] <BradB> thanks
[11:30] <SteveA> thanks for bringing this up!
[11:30] <BradB> no prob :)
[11:30] <SteveA> I think this might lead to more sane add/edit forms eventually too
[11:30] <BradB> i hope so.
[11:30] <SteveA> if you get an example with more than one form on a page, let me know
[11:31] <BradB> ok
[11:31] <SteveA> I'd rather work from a real-life requirement than a kind of "it would be neat", of course
[11:31] <BradB> yes