00:26:42 -!- Fare [~Fare@74.125.59.116] has quit [Ping timeout: 250 seconds] 00:38:27 Fare [~Fare@63.115.78.49] has joined #sbcl 01:27:22 -!- hargettp [~hargettp@pool-71-184-185-93.bstnma.east.verizon.net] has quit [Quit: Leaving...] 01:32:27 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Quit: Leaving.] 02:15:31 hargettp [~hargettp@pool-71-184-185-93.bstnma.east.verizon.net] has joined #sbcl 02:31:05 -!- hargettp [~hargettp@pool-71-184-185-93.bstnma.east.verizon.net] has quit [Quit: Leaving...] 03:15:11 -!- Fare [~Fare@63.115.78.49] has quit [Ping timeout: 276 seconds] 03:29:57 Fare [~Fare@74.125.59.116] has joined #sbcl 05:33:49 superjudge [~superjudg@195.22.80.141] has joined #sbcl 05:34:00 tcr1 [~tcr@80-218-247-218.dclient.hispeed.ch] has joined #sbcl 06:26:50 -!- ASau [~user@95-24-202-195.broadband.corbina.ru] has quit [Ping timeout: 248 seconds] 06:29:24 ASau [~user@95-27-146-128.broadband.corbina.ru] has joined #sbcl 06:32:33 flip214 [~marek@2001:858:107:1:7a2b:cbff:fed0:c11c] has joined #sbcl 06:32:33 -!- flip214 [~marek@2001:858:107:1:7a2b:cbff:fed0:c11c] has quit [Changing host] 06:32:33 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 06:39:41 slyrus [~chatzilla@adsl-99-62-138-101.dsl.pltn13.sbcglobal.net] has joined #sbcl 07:18:46 -!- tcr1 [~tcr@80-218-247-218.dclient.hispeed.ch] has quit [Quit: Leaving.] 07:25:23 -!- cmm [~cmm@109.67.199.173] has quit [Ping timeout: 244 seconds] 07:26:37 cmm [~cmm@bzq-109-67-199-173.red.bezeqint.net] has joined #sbcl 07:28:20 -!- Fare [~Fare@74.125.59.116] has quit [Ping timeout: 250 seconds] 07:31:16 hlavaty [~user@91-65-223-81-dynip.superkabel.de] has joined #sbcl 07:31:30 nikodemus [~nikodemus@cs181063174.pp.htv.fi] has joined #sbcl 07:31:30 -!- ChanServ has set mode +o nikodemus 07:43:18 Fare [~Fare@63.115.78.49] has joined #sbcl 08:09:01 tcr1 [~tcr@217-162-207-164.dynamic.hispeed.ch] has joined #sbcl 08:27:50 what's the sbcl equivalent of memcmp? 08:29:45 attila_lendvai [~attila_le@adsl-89-132-53-222.monradsl.monornet.hu] has joined #sbcl 08:29:45 -!- attila_lendvai [~attila_le@adsl-89-132-53-222.monradsl.monornet.hu] has quit [Changing host] 08:29:45 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 08:33:30 -!- Krystof [~csr21@csrhodes.plus.com] has quit [Ping timeout: 255 seconds] 08:34:42 Do you want (equal) for arrays? 08:40:43 misterncw [~misterncw@82.71.241.25] has joined #sbcl 09:02:45 -!- nikodemus [~nikodemus@cs181063174.pp.htv.fi] has quit [Quit: This computer has gone to sleep] 09:14:50 stassats [~stassats@wikipedia/stassats] has joined #sbcl 09:21:36 tcr1: the closest thing is mismatch 09:45:30 Krystof [~csr21@howells.doc.gold.ac.uk] has joined #sbcl 09:45:30 -!- ChanServ has set mode +o Krystof 09:46:14 nikodemus [~nikodemus@cs181058025.pp.htv.fi] has joined #sbcl 09:46:14 -!- ChanServ has set mode +o nikodemus 09:49:10 I'm talking about comparing to saps 09:49:23 I mean the memory they're pointing to 09:56:55 -!- superjudge [~superjudg@195.22.80.141] has quit [Quit: superjudge] 09:57:34 superjudge [~superjudg@195.22.80.141] has joined #sbcl 09:59:28 -!- superjudge [~superjudg@195.22.80.141] has quit [Client Quit] 10:04:01 hargettp [~hargettp@pool-71-184-185-93.bstnma.east.verizon.net] has joined #sbcl 10:07:37 tcr1: memcmp? :) 10:08:19 I'm fine with that answer 10:09:52 if the foreign call overhead gets too expensive, it's not too hard to write something decent, of course. 10:10:00 but if you do that too much, i'd expect ffi overhead to be noticeable 10:15:37 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 246 seconds] 10:15:39 stassats` [~stassats@wikipedia/stassats] has joined #sbcl 10:15:49 and what about sb-kernel:ubX-bash-copy? 10:17:18 well, that's fine for copying 10:17:21 not so good for comparing 10:18:06 oh well, memcpy looks just like memcp 10:18:39 damn, memcmp 10:27:45 -!- stassats` [~stassats@wikipedia/stassats] has quit [Ping timeout: 260 seconds] 10:40:40 -!- Krystof [~csr21@howells.doc.gold.ac.uk] has quit [Ping timeout: 240 seconds] 10:46:08 ETOOCOMPLEXMACRO 10:46:23 goddammit. decoding is harder than encoding 10:54:14 http://paste.lisp.org/display/122405 # feeling clever today 10:54:41 much easier to debug than when it was in a macrolet 11:02:30 Krystof [~csr21@howells.doc.gold.ac.uk] has joined #sbcl 11:02:30 -!- ChanServ has set mode +o Krystof 11:06:02 -!- hargettp [~hargettp@pool-71-184-185-93.bstnma.east.verizon.net] has quit [Quit: Leaving...] 11:07:06 hargettp [~hargettp@pool-71-184-185-93.bstnma.east.verizon.net] has joined #sbcl 11:42:20 -!- Fare [~Fare@63.115.78.49] has quit [Ping timeout: 240 seconds] 11:50:21 -!- hargettp [~hargettp@pool-71-184-185-93.bstnma.east.verizon.net] has quit [Quit: Leaving...] 12:03:25 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has left #sbcl 12:11:22 -!- Krystof [~csr21@howells.doc.gold.ac.uk] has quit [Ping timeout: 248 seconds] 12:16:24 http://paste.lisp.org/display/122405#1 -- got decoding to work pretty decently, i'd say 12:23:14 *bsbutler* cheers for nikodemus 12:24:30 -!- bsbutler is now known as redline6561 12:24:32 -!- redline6561 [~user@c-66-56-55-169.hsd1.ga.comcast.net] has left #sbcl 12:29:07 it's funny that babel conses less while being 2x slower 12:37:00 I'm curious what those times look like on other Lisps. (Albeit not curious enough to try myself at this point.) 12:37:26 There must be a reason any code I've had that uses cffi was mindbogglingly slow on, say, CCL, compared to SBCL (aside from CCL's compiler not playing in the same league, which shouldn't matter _that_ much for FFI code). 12:38:02 what kind of code 12:38:02 ? 12:38:25 if string translations were done, then babel may as well be at fault 12:38:34 you should see how slow it's on Clisp :D 12:39:01 yeah, clisp manages to underperform even compared to ccl :-) 12:40:22 I'm mostly thinking in terms of toy benchmarks I wrote when debugging CommonQt. The full CommonQt calling sequence is still annoyingly Lisp-code-heavy, but I extracted little snippets that focus on the FFI parts. 12:41:06 It was a huge improvement on CCL (and minor improvement on SBCL) to cache away anything obviously related to C strings. But even afterwards stuff was sluggish, and I haven't -- yet -- investigated further. 12:52:26 SBCL typically gets more love in both CFFI and Babel development, that might explain some of it. 12:52:50 stassats [~stassats@wikipedia/stassats] has joined #sbcl 12:54:30 lichtblau: was it OSX that required all GUI callbacks to run in the initial/main thread ? 12:58:21 superjudge [~superjudg@host-90-238-94-205.mobileonline.telia.com] has joined #sbcl 12:59:51 yes, I have no MacOS experience, but that's the stated reason other libraries have this requirement. 13:00:49 s/callback/any GUI operation whatsoever/ as I understand it 13:06:24 the reason babel is slower while consing less is that computes the exact length of the decoded string first 13:07:10 whereas my stuff guesstimates, and if it wasn't enough allocates another buffer, etc till everything has been decoded and then just copies them to the final result string 13:07:34 one of the things that still needs love here is the guesstimation, though 13:08:00 i also play a dirty trick with %shrink-vector 13:08:57 i'm thinking that for relatively short strings, the guesstimate can be "large enough for sure", whereas longer strings should be random sampled to see what they look like 13:08:59 yeah, exactly what I was thinking of 13:10:04 octets-to-string yields a number of chars <= to that of octets, so allocating a string of equal length then using %shrink-vector seems ok 13:10:55 but what happens with the no longer used memory at the end of the string ? 13:11:16 does it get freed only when the GC moves the string ? 13:11:18 it's reclaimed when the GC copies the string 13:11:29 ok 13:12:22 but yes, there should be a sanity check not to return a string with large amount of junk at the end, in case it ends up being pinned for a long time 13:13:28 could we add such a function to alexandria ? 13:13:35 Krystof [~csr21@howells.doc.gold.ac.uk] has joined #sbcl 13:13:35 -!- ChanServ has set mode +o Krystof 13:13:37 nikodemus: macroexpand-1 ;) 13:14:00 instead of macroexpand? probably :) 13:14:33 -!- superjudge [~superjudg@host-90-238-94-205.mobileonline.telia.com] has quit [Quit: superjudge] 13:14:45 it's also reclaimed when the GC doesn't copy the string, I think (in page increments) 13:16:09 you're right 13:16:15 um 13:16:28 wait, i don't think we allocate anything to pinned pages? 13:17:08 and i think those pages remain in the same allocation region despite %shrink-vector 13:17:30 and pinning IIRC happens by allocation regions. or maybe it doesn't anymore. not sure 13:20:45 I don't know, but large strings won't be copied (they'll be promoted in-place but left in large object pages) 13:21:38 Not sure what'll happen for pinned large objects. 13:36:30 same thing, iirc 13:43:19 -!- flip214 [~marek@unaffiliated/flip214] has quit [Remote host closed the connection] 13:51:41 -!- Krystof [~csr21@howells.doc.gold.ac.uk] has quit [Ping timeout: 240 seconds] 14:06:30 homie [~levent.gu@xdsl-78-35-146-172.netcologne.de] has joined #sbcl 14:18:45 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 260 seconds] 14:18:50 redline6561-work [~redline65@66.6.146.58] has joined #sbcl 14:21:54 -!- misterncw [~misterncw@82.71.241.25] has quit [Remote host closed the connection] 14:35:15 loke [~elias@bb116-14-45-139.singnet.com.sg] has joined #sbcl 15:01:01 misterncw [~misterncw@82.71.241.25] has joined #sbcl 15:03:23 -!- misterncw [~misterncw@82.71.241.25] has quit [Remote host closed the connection] 15:09:20 Krystof [~csr21@158.223.161.121] has joined #sbcl 15:09:20 -!- ChanServ has set mode +o Krystof 15:14:30 -!- hlavaty [~user@91-65-223-81-dynip.superkabel.de] has quit [Ping timeout: 246 seconds] 15:16:52 http://paste.lisp.org/display/122405#2 -- wow, our current utf-16 is _slow_ 15:18:16 of course, given how the current sbcl conversions work, it might be that it's decent on streams and just do-slow on string<->octets 15:18:27 dog-slow, even 15:22:23 -!- Krystof [~csr21@158.223.161.121] has quit [Read error: Operation timed out] 15:35:40 Fare [~Fare@63.115.78.49] has joined #sbcl 15:48:07 pdlogan [~patrick@174-25-37-137.ptld.qwest.net] has joined #sbcl 15:58:45 Krystof [~csr21@howells.doc.gold.ac.uk] has joined #sbcl 15:58:45 -!- ChanServ has set mode +o Krystof 16:37:11 -!- tcr1 [~tcr@217-162-207-164.dynamic.hispeed.ch] has quit [Quit: Leaving.] 16:56:27 http://paste.lisp.org/display/122395#3 -- first cut at BOM support 16:57:00 -!- Krystof [~csr21@howells.doc.gold.ac.uk] has quit [Ping timeout: 240 seconds] 18:11:13 -!- nikodemus [~nikodemus@cs181058025.pp.htv.fi] has quit [Ping timeout: 258 seconds] 18:36:13 Krystof [~csr21@csrhodes.plus.com] has joined #sbcl 18:36:13 -!- ChanServ has set mode +o Krystof 19:33:38 -!- redline6561-work [~redline65@66.6.146.58] has quit [Ping timeout: 258 seconds] 19:37:19 redline6561-work [~redline65@cei-gla-cpk2.coxinc.com] has joined #sbcl 19:39:37 attila_lendvai [~attila_le@apn-94-44-172-67.vodafone.hu] has joined #sbcl 19:39:37 -!- attila_lendvai [~attila_le@apn-94-44-172-67.vodafone.hu] has quit [Changing host] 19:39:37 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 19:42:27 -!- Krystof [~csr21@csrhodes.plus.com] has quit [Ping timeout: 252 seconds] 19:42:31 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Read error: Connection reset by peer] 19:42:35 tcr1 [~tcr@80-218-247-218.dclient.hispeed.ch] has joined #sbcl 19:43:04 attila_lendvai [~attila_le@apn-94-44-172-67.vodafone.hu] has joined #sbcl 19:43:04 -!- attila_lendvai [~attila_le@apn-94-44-172-67.vodafone.hu] has quit [Changing host] 19:43:04 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 19:56:29 Krystof [~csr21@csrhodes.plus.com] has joined #sbcl 19:56:29 -!- ChanServ has set mode +o Krystof 20:09:05 -!- redline6561-work [~redline65@cei-gla-cpk2.coxinc.com] has quit [Remote host closed the connection] 20:14:41 redline6561-work [~redline65@cei-gla-cpk2.coxinc.com] has joined #sbcl 20:16:36 -!- homie [~levent.gu@xdsl-78-35-146-172.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 22:56:43 nikodemus [~nikodemus@cs181063174.pp.htv.fi] has joined #sbcl 22:56:43 -!- ChanServ has set mode +o nikodemus 23:03:40 -!- redline6561-work [~redline65@cei-gla-cpk2.coxinc.com] has quit [Remote host closed the connection] 23:25:41 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Read error: Connection reset by peer] 23:27:37 -!- tcr1 [~tcr@80-218-247-218.dclient.hispeed.ch] has quit [Quit: Leaving.] 23:31:27 -!- pdlogan [~patrick@174-25-37-137.ptld.qwest.net] has left #sbcl 23:45:56 -!- nikodemus [~nikodemus@cs181063174.pp.htv.fi] has quit [Read error: Operation timed out]