Perl 6 - the future is here, just unevenly distributed

IRC log for #pdl, 2015-01-29

| Channels | #pdl index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:19 sivoais mohawk: that's gonna have to wait until the weekend unfortunately
01:20 sivoais trying to get some writing done
03:29 mohawk sivoais, fair enough
03:29 mohawk good luck with the writing
03:29 mohawk does anyone know why med2d might be misbehaving? i've got a nice new layout of PDL and that's the one test that's failing (in t/bad.t and t/image2d.t)
03:31 mohawk hang on, i've got it
03:31 mohawk it's because i moved image2d into the "kitchen sink" area and it's loading the system-installed one instead
03:43 mohawk that's quite a relief
03:54 drrho joined #pdl
12:15 opkick [pdl] wchristian pushed 20 new commits to master: http://git.io/FV7C
12:15 opkick pdl/master c432bdb Ed J: Put Changes file actually in the distdir and therefore the distro
12:15 opkick pdl/master 72bc2c4 Ed J: Sort MANIFEST* and last chunk of .gitignore
12:15 opkick pdl/master 6bbef1a Ed J: Remove unnecessary Basic/SourceFilter cruft
15:49 mohawk aww yeah
15:49 mohawk chm++
17:30 mohawk i think i may be suffering from PDL Traumatic Stress Disorder
19:15 opkick [pdl] wchristian tagged v2.007_09 at d5801be: http://git.io/FKs1
19:15 opkick [pdl] wchristian pushed 2 new commits to master: http://git.io/FKsX
19:15 opkick pdl/master a2c1f04 Chris Marshall: Update Known_problems and Release_Notes for CPAN devel release
19:15 opkick pdl/master 609efb0 Chris Marshall: Update VERSION to 2.007_10 for more development
19:50 sivoais vicash: what do you think of <https://tradewave.net/>. Just got linked to it by a friend
19:50 sivoais "Build your own automated
19:50 sivoais Bitcoin trading strategy."
19:51 * vicash looks
19:53 vicash sivoais: it is a good idea.. the fact is that English based written logic strategies have been around for years but they all evolved into their own complex DSLs that ended up being super ugly
19:55 vicash but the point that such languages can actually make a saleable product is good
19:57 vicash as far as i see there is no money in making such products per se. the money is making sure that your platform can get commissions per trade. the amount of money you can make by selling a piece of software like this is too less for the hours of effort you put in building it and making it viable to every possible type of trader.
19:57 vicash the real way to make money is have a platform that ppl buy and trade on, so then you make money on those commissions etc via market making
20:05 vicash i see that they have a pricing plan but i wil be surprised that they break even with it.. having worked for companies that build such platforms, these are just teasers to on-board users after which they end up building market making systems to profit on each trade that the user does
20:21 sivoais I see. I wonder what they do with the trading strategies that people write for the platform. They could run any strategy they want to build a meta-strategy.
20:31 vicash well the fact is that these are not targeted towards institutional traders. they are targeted to mostly amateur folks and some power-user type traders
20:32 vicash so they dont really care what kind of strategies you run, the prices you see are already way delayed
20:32 vicash like for example, Interactive Brokers sends you data every 100ms. they take snapshot of existing data every 100ms and stream that to you..
20:32 vicash but you know that stock market prices are much faster than that
20:33 vicash obviously using such platforms means not doing trading at the millisecond level. as retail investors you make money by trading maybe over a few days or if you do day trading over 15 minutes to a couple of hours.. there the data every 100ms or 500ms will not matter and price moves can be higher than the market making levels
20:35 vicash at least for Interactive Brokers which I use personally for my own automated trading systems, the killer money maker for them is commission per trade
20:35 vicash regardless of whether i make profit or loss, they make their commission
22:34 mohawk the swine
22:37 mohawk so with a perl without DynaLoader (using it as i had it perlbrewed to check PDL when no PDL already installed), PDL::Core refuses to load
22:37 mohawk i think the ppgen stuff is a bit broken
22:40 mohawk or, it might just be Core.pm.PL
22:42 mohawk ooooor, it might be EUMM broken
22:42 * mohawk sobs
22:55 jberger no DynaLoader?
22:56 mohawk yes, configured with -Uusedl
22:56 mohawk for use in embedding, probably
23:04 jberger ah, I see
23:04 jberger PDL in an embedded environment, sound like fun
23:07 mohawk ha ha
23:07 mohawk why not
23:07 mohawk run PDL on one's wristwatch
23:07 mohawk however, that's just a scenario where such a beast might make sense
23:08 mohawk i have such a beast, and EUMM is supposed to work with it
23:21 mohawk so latest and greatest EUMM worked, so that's awesome
23:48 mohawk ah, it wasn't the latest version
23:49 mohawk it's an ordering thing
23:49 mohawk EUMM was relying on static having being built, by having it sequenced in the deps
23:49 mohawk whereas it needs to be made an actual dependency
23:51 mohawk ...and that works

| Channels | #pdl index | Today | | Search | Google Search | Plain-Text | summary