2008-10-31:
[10:35] <simono> antville guys around? are you running antville on helma-trunk or the latest release (the tar.gz on downloadpage)?[10:37] * zumbrunn looks around and doesn't see any antville guys[10:37] <simono> :)[10:39] <simono> i have a problem with generic relations. when the generic object() is also the _parent it seems i have to persist the parent before adding its child[10:41] <simono> and it seems antville is working around this e.g. here http://code.google.com/p/antville/source/browse/trunk/code/Comment/Comment.properties and in other places they have a generic object as parent, but there it could be they are sure to persist the parent in time[10:50] <zumbrunn> I think there was a bug filed regarding something like that[10:51] <zumbrunn> by tobi, if I remember correctly[10:51] <zumbrunn> checking...[10:51] <zumbrunn> http://helma.org/bugs/show_bug.cgi?id=607[11:23] <simono> thx i'll check it out[11:47] <lehni> chris, you here?[11:48] <zumbrunn> yep[11:48] <lehni> have you seen this: http://www.mibbit.com/widget.html[11:48] <lehni> might be good to replace the rather clunky java irc client on your webiste[11:49] <zumbrunn> yes, I'll replace the current java one with it[11:49] <zumbrunn> I agree[11:49] <lehni> (no offense)[11:49] <lehni> ;)[11:49] <zumbrunn> I actually wanted to ask you about your opinion on that yesterday[11:49] <lehni> it works pretty well[11:49] <lehni> sometimes there is a bit of a lag[11:50] <lehni> but people using irc regularly should install a software i guess[11:50] <zumbrunn> it has come up before, and there were split opinions about which is better[11:50] <lehni> to quickly pop in its perfect[11:50] <lehni> it's up to you![11:50] <zumbrunn> slightly favoring the switch to mibbit[11:50] <lehni> what are you using on mac?[11:51] <zumbrunn> Colloquy[11:52] <lehni> will try that one[11:52] <lehni> i had a little meta object adventure day yesterday. not sure you were following ;)[11:53] <zumbrunn> yes, I was following that![11:53] <lehni> if there was a way to make these recursive lookups faster it would be such a great feature[11:54] <lehni> e.g. having the prototype provide a meta object that handles all meta functinon calls for each of its instances[11:57] <zumbrunn> it's a very cool feature, but when performance suffers even if the feature isn't used, it becomes a hard sell[11:57] <lehni> i know[11:57] <lehni> shame[11:57] <zumbrunn> partly because much of what it does can be done using JSAdapter[11:57] <lehni> the weird thing is that it's not actually the recursive lookups that make it so much slower[11:59] <lehni> it's more the ThreadLocal thing with disableMop, to avoid recursive calls[11:59] <lehni> according to my measurements[11:59] <zumbrunn> yes, I remember[11:59] <lehni> will look into JSAdapter more[12:00] <zumbrunn> so, what do you think about helma-ng? have you looked at it a bit?[12:00] <lehni> not much yet[12:00] <lehni> but what i see and hear sounds great[12:00] <zumbrunn> :-)[12:00] <lehni> i am stuck with 1.x though[12:00] <lehni> for the time being[12:00] <lehni> a few production sites and a growing cms that is based on it[12:02] <lehni> Colloquy is great, btw[12:08] <zumbrunn> time for me to go walking with the dogs[12:08] <zumbrunn> requiring snow shoes for the first time this winter[12:08] * zumbrunn will be back in a bit[12:14] <lehni> yes, early snow this year, even in london (for the first time in october since '34)[14:00] <hannesw> hi juerg -[14:00] <hannesw> just another idea/link:[14:00] <hannesw> http://groups.google.com/group/mozilla.dev.tech.js-engine.rhino/browse_frm/thread/1a9ab69548fba98/6d9a64518a6586ff?lnk=gst&q=456546#6d9a64518a6586ff[14:00] <hannesw> about implementing/extending ScriptableObject in javascript.[14:00] <hannesw> might be another solution for your problem, provided you're willing to use a special object implementation[14:03] <hannesw> so... what do you think about my latest rubyism? http://groups.google.com/group/helma-ng/browse_frm/thread/70e337f164c2a5ef[14:04] <hannesw> (off to reboot ubuntu 8.10)[14:04] <lehni> looking now[14:06] <lehni> time methods on numbers are gerat[14:06] <lehni> great i mean[14:06] <lehni> i am adding some to bootstraps too[14:08] <lehni> have you seen this:[14:08] <lehni> http://www.datejs.com/[14:13] <lehni> i have found a nice workaround for my problem[14:13] <lehni> BootStrap now can inject properties defined like this:[14:13] <lehni> { property: { _get: function() {}, _set: function(value) {} }}[14:14] <lehni> this produces getters and setters on the fly[14:14] <lehni> based on this i made a prototype called WatchableProperty which takes a property name and instals a watchable property under a different name. this property then sets native fields.[14:14] <lehni> the only downside is that they cannot have the same name, but that's fine[14:15] <lehni> so e.g. visible: new WatchableProperty('visibility', function(vale) { ... });[14:40] <lehni> i have posted some example code to the pastebin for those interested[14:40] <zumbrunn> http://helma.pastebin.com/d43b31c2c[14:40] <lehni> thanks[14:41] <lehni> i am excited about this[14:41] <lehni> feels right[14:41] <lehni> in the example, realValue would be stored in the database[14:41] <lehni> but the code would access only value
In the channel now:
Logs by date: