--- Log opened Sun Apr 12 00:00:17 2009
09:14 < dancor> @users
09:15 < dancor> weird, there is a lambdabot commit from Thu Jan 22 18:20:30 2009 "make happs a default channel"
09:40 < gwern> doesn't mean lispy or whomever is hosting lb has pulled those changes
11:33 < mae> first i heard of tupil
11:33 < mae> where is the front end? :)
11:35 < mae> all: sorry i haven't been around, its been crazy here, i tried to switch isps unsuccessfully, and my wife is getting more demanding........ (baby is coming)
11:35 < mightybyte> Congratulations
11:35 < mae> anyhoo, if there are any major issues that need to be talked about, lets have it on the mailing list so i can respond asynchronously
11:35 < mae> thx! :)
11:35 < mightybyte> It did seem that #happs activity had decreased lately.
11:38 < mae> yeah it didn't help that my work had an emergency yesterday
11:38 < mae> an accident on one of att's major lines
11:38 < mae> took 24 hours to fix
11:38 < mightybyte> Ouch
11:38 < mae> so we had to switch all the stuff to our backup isp
11:39 < mae> so ppl could use email etc
11:40 < mightybyte> Not fun
11:43 < mae> yeah..
11:43 < mae> so looks like a lot of activity on the ml
11:43 < mae> and patches too
11:43 < mae> exciting to see
11:43 < mightybyte> Yeah
11:43 < mae> i think i will switch to 2 month release cycles
11:43 < mae> 1 month is too short
11:44 < mae> after 0.3
11:44 < mightybyte> Especially for sharding 0.4
11:44 < mae> i am still unsure of the sharding functionality making it by then
11:44 < mae> we have a fork in the road
11:44 < mae> one is to pursue the implementation with spread
11:45 < mae> another is solidsnacks algorithm
11:45 < mightybyte> Ahhh
11:45 < mae> i really need to blog about that
11:45 < mae> he wants people to find the holes in his algorithm and prove its correctness before he prototypes it
11:46 < mae> ahh in any case, i'll try to get back on here next week and be more active
11:46 < mightybyte> Who is "he"?
11:46 < mae> solidsnack
11:46 < mae> um
11:46 < mightybyte> Oh, right.  At first solidsnack sounded like the name of the algorithm.
11:46 < mae> tupil.com
11:47 < mae> i think his name is chris eidhof
11:47 < mae> so basically, up to this point i've been using the 'itch scratching method' to work on the project
11:47 < mae> when i wrote that roadmap initially it was kind of pie in the sky
11:47 < mightybyte> Hmmm, I thought Chris used the name "chr1s" on IRC.
11:47 < mightybyte> Ok
11:47 < mae> i am not having an immediate need for sharding, as i have not hit that scalability wall yet
11:48 < mae> so lately my focus has been on making the development much more painless
11:48 < mae> are you in need of sharding?
11:48 < mae> mightybyte: maybe i'm mixed up :)
11:48 < mightybyte> Yeah, I think that's the right thing to do, but sharding will take awhile to get workout out, so I think it should be started early.
11:48 < mightybyte> No, I don't need sharding right now.
11:49 < mae> mightybyte: right, so again, i don't think until someone "really needs it" will it be able to be developed effectively
11:49 < mae> because A: there won't be the effort put in, B: it will be an estimated, calculated use case, and not a "real world" scenario
11:49 < mightybyte> But by that time, whoever needs it will be in a tough spot.
11:50 < mightybyte> But you're right.  Large-scale is difficult to develop for when you don't actually have the traffic.
11:50 < mae> thats true :)
11:50 < mae> we should discuss this on the ml, I gotta run right now.
11:51 < mightybyte> Ok
11:51 < mae> if you think its important, please post your thoughts
11:51 < mae> and lets get a thread going
11:51 < mightybyte> At some point, I think we should make it a mandatory part of a release.
11:51 < mightybyte> Doesn't have to be 0.4 though.
11:51 < mae> interesting idea.
11:52 < mae> but again, i don't think we can force people to work on something that isn't practical for them :)
11:52 < mightybyte> Of course.
11:52 < mae> and I really can't do the sharding stuff alone, I don't have the bandwidth
11:52 < mightybyte> Me either right now.
11:53 < mae> maybe if i take some paternity leave
11:53 < mae> pair coding
11:53 < mae> feed the baby
11:53 < mae> code
11:53 < mae> feed the baby
11:53 < mae> code
11:53 < mightybyte> lol
11:53 < mae> he he he.
11:53 < Zao> feed the baby code
11:53 < mae> that too
11:53 < mightybyte> Write code to feed the baby.
11:53 < mae> ok gotta run, like i said, bring it up on the ML :)
11:53 < mightybyte> Catch you later.
11:55 < t345> does happstack support native compilation like Ocsigen?
13:09 < t345> ok, looks like there's no answer to that but trying out :D
15:32 < cocon> how can I generate good random numbers for session IDs?
15:32 < cocon> does happstack come with session management / authentication?
16:12 < mightybyte> cocon: http://softwaresimply.blogspot.com (look at the 3 most recent posts)
16:17 < mightybyte> cocon': (just in case you missed it) Check out the 3 most recent posts at http://softwaresimply.blogspot.com
16:26 < mmmdonuts> gwern, about that compressedResponseFilter patch for happstack, was that already in the repo you used, or did you have to apply it yourself?
16:26 < mmmdonuts> I ask because when I retrieved the happstack repo from patch-tag, it doesn't seem to have that patch.
16:26 < gwern> mmmdonuts: it's there if you search
16:26 < gwern> oddly, it seems to be out of date order
16:26 < mmmdonuts> Huh - ok, thanks.
16:27 < gwern> I think what happened is that this patch got recorded but not applied for a long time:
16:28 < gwern> Thu Mar 26 00:54:17 EDT 2009  Gregory Collins <greg@gregorycollins.net> * fix-timeout-leak-trunk.diff
16:28 < mmmdonuts> Wow, I found it, thanks.  Completely missed it the first time.
16:28 < gwern> so you have to go down like 2 or 3 pages
--- Log closed Mon Apr 13 00:00:20 2009