#grok Tuesday, 2007-09-04

*** faassen has quit IRC00:20
*** sod has quit IRC01:47
*** sod has joined #grok01:48
*** michelp has joined #grok02:36
*** fcorrea has quit IRC03:03
*** philiKON has quit IRC04:09
*** seocam has joined #grok04:13
*** sod has quit IRC04:44
*** kushal has quit IRC04:54
*** seocam has quit IRC04:58
*** kushal has joined #grok07:08
*** baijum has joined #grok07:47
*** LeoRochael has quit IRC08:15
*** cursor has joined #grok08:43
*** j-w has joined #grok09:00
*** timte has joined #grok09:25
*** juh has joined #grok09:26
*** juh has quit IRC09:39
*** juh has joined #grok09:52
*** norro has joined #grok10:34
*** norro has quit IRC11:22
*** philiKON has joined #grok11:26
*** d-bug has joined #grok11:52
*** juh has quit IRC11:53
*** d-bug has left #grok11:53
*** juh has joined #grok11:54
*** juh has quit IRC12:11
*** juh has joined #grok12:13
*** juh has joined #grok12:14
*** juh has quit IRC12:21
*** juh has joined #grok12:22
*** davidstryker has joined #grok12:40
*** davidstryker has left #grok12:40
*** juh has quit IRC12:52
*** LeoRochael has joined #grok13:07
*** brandon_rhodes has quit IRC13:19
*** kushal has quit IRC13:20
*** kushal has joined #grok13:20
*** juh has joined #grok13:21
*** kushal has quit IRC13:29
*** kushal has joined #grok13:29
*** kushal has quit IRC13:30
*** kushal has joined #grok13:30
*** kushal has quit IRC13:35
*** kushal has joined #grok13:35
*** rocky has quit IRC13:39
*** kushal has quit IRC13:40
*** kushal has joined #grok13:41
*** rocky has joined #grok14:15
*** ignas has joined #grok14:33
*** rocky has quit IRC14:35
*** rocky has joined #grok14:41
*** ruda_porto has joined #grok14:47
*** fcorrea has joined #grok15:12
*** kushal has quit IRC15:18
*** cursor has quit IRC15:24
*** regebro has joined #grok15:45
*** gumpa has joined #grok15:54
*** baijum has quit IRC15:57
*** brandon_rhodes has joined #grok15:57
j-wthe order in which event subscribers are "called" is not deterministic, is it?16:17
philiKONright16:18
philiKONwell, heuristically there's an order16:18
philiKONit's an implementation detail16:18
j-wbut how can the indexing work correctly then?16:18
j-wI mean, for example, say I have an index that uses an attribute on the dublincore annotation16:19
j-wis the annotation (modified e.g.) update *before* the object is indexed?16:19
j-w(hmm, my explanation is bit too brief maybe)16:20
philiKONwell, when you change annotations you're supposed to send an event too ;)16:21
philiKONs/;)/:)16:21
j-win that case the object is indexed twice basically16:21
j-wright?16:21
philiKONobjects aren't indexes. attributes are16:21
philiKONunfortunately, the  indexes don't yet interpret the information sent along with the event16:22
philiKONso, yes, everything would be indexed twice then16:22
j-wok16:22
philiKONa way to work around that is to remember to-be-reindexed  objects in a queue16:22
philiKONand then separately index them asynchronously16:22
* j-w looks at the zope.dublincore event subscribers16:23
j-wwell, there's not reallt a problem other than my attempt to understand the mechanism :)16:23
j-wthe thing is: zope.dublincore's event subscribers do not send out events (I think)16:24
philiKONheh ok16:24
j-win other words, the indexes based on dublincore attributes might not be "up to date"16:24
philiKONi guess they don't because a) when they were written we didn't have that additional info to send along with the event so that b) they would call each other in recursion16:24
j-wok16:24
philiKONwhat they should really do is send an event16:24
philiKONbut sned along the information16:24
philiKONand when they themselves are called, the should bail out if they detect the 'modified' dublincore attr was changed16:25
philiKONs/the should/they should/16:25
j-w"that additional info" is sent along with the event how?16:26
* j-w wonders if there's anything written about this somewhere16:27
philiKONj-w: page 31016:27
philiKONj-w: Ex 16.2.116:27
j-wLOL16:27
philiKONlines 29-3016:27
* j-w reads16:27
philiKONjust beware of the erratum there16:28
j-w:-)16:28
philiKONhttp://worldcookery.com/Errata16:28
j-waaaah! right!16:28
j-wI overlooked this "detail" in the example16:29
j-wthanks a LOT!16:29
philiKONnp16:29
j-wthis reminds me: would adding the URL to the irc logs in the channel's topic be useful?16:29
philiKONyeah, sure16:30
*** j-w changes topic to "http://grok.zope.org | http://philikon.de/files/grok-bbq-sprint.pdf | irc logs: http://jw.n--tree.net/irclogs/%23grok"16:31
*** ruda_porto has quit IRC16:38
*** nathany has joined #grok16:42
j-wphiliKON: you actually made a bug report for this a while ago...16:42
j-wfor the fact that zope-3 does not yet fully use the extra information to pass with an event.16:43
fcorreaphiliKON: in your 6th errata in http://worldcookery.com/Errata about PersistentList is wrong. Instead of mentioning the way it was in the book, you actually fixed that in here : PersistentList is mentioned twice (PersistentList and PersistentDict) while the text meant to say "PersistentList and PersistentDict."17:00
fcorreaI think you meant: PersistentList is mentioned twice (PersistentList and PersistentList)17:01
*** ignas has quit IRC17:12
*** sod has joined #grok17:15
*** timte has quit IRC17:16
*** j-w has quit IRC17:27
*** kushal has joined #grok17:33
*** juh has quit IRC17:42
*** juh has joined #grok18:06
*** juh has quit IRC18:10
*** ruda_porto has joined #grok18:55
*** cursor has joined #grok18:56
*** kushal has quit IRC18:57
*** norro has joined #grok20:22
*** cursor has quit IRC20:36
*** seocam has joined #grok20:41
*** rocky has quit IRC20:46
*** seocam has left #grok21:23
*** tiola has quit IRC22:03
*** fcorrea has left #grok22:34
*** gumpa has quit IRC23:03
*** ruda_porto has quit IRC23:12
*** regebro has left #grok23:14
*** brandon_rhodes has quit IRC23:37

Generated by irclog2html.py 2.5 by Marius Gedminas - find it at mg.pov.lt!