| --- Log opened Wed Mar 28 00:00:03 2018 |
| 00:07 | | macdjord [macdjord@Nightstar-a1fj2k.mc.videotron.ca] has quit [[NS] Quit: Live Long and Prosper.] |
| 00:25 | <@ErikMesoy> | https://i.imgur.com/Uhyizag.png |
| 00:28 | <@abudhabi> | :D |
| 00:48 | | Derakon[AFK] is now known as Derakon |
| 00:53 | | Degi [Degi@Nightstar-ibr0nm.dyn.telefonica.de] has quit [Connection closed] |
| 00:55 | | celticminstrel [celticminst@Nightstar-m1ak9t.dsl.bell.ca] has joined #code |
| 00:55 | | mode/#code [+o celticminstrel] by ChanServ |
| 00:56 | | himi [sjjf@Nightstar-1drtbs.anu.edu.au] has joined #code |
| 00:56 | | mode/#code [+o himi] by ChanServ |
| 03:23 | | RchrdB [RchrdB@Nightstar-qe9.aug.187.81.IP] has joined #code |
| 03:25 | | RchrdB [RchrdB@Nightstar-qe9.aug.187.81.IP] has quit [Connection closed] |
| 04:05 | | Jessikat [Jessikat@Nightstar-bt5k4h.81.in-addr.arpa] has quit [[NS] Quit: Leaving] |
| 04:15 | | Jessikat [Jessikat@Nightstar-d7j.klt.132.82.IP] has joined #code |
| 04:34 | | Jessikat` [Jessikat@Nightstar-o4q.jtr.132.82.IP] has joined #code |
| 04:37 | | Jessikat [Jessikat@Nightstar-d7j.klt.132.82.IP] has quit [Ping timeout: 121 seconds] |
| 04:49 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 05:26 | | celticminstrel [celticminst@Nightstar-m1ak9t.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 05:49 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 05:52 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 06:08 | | himi [sjjf@Nightstar-1drtbs.anu.edu.au] has quit [Ping timeout: 121 seconds] |
| 06:27 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 06:27 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 06:43 | | Vornlicious [Vorn@Nightstar-ipsdsc.sub-174-211-2.myvzw.com] has joined #code |
| 06:45 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 06:45 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 07:18 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 07:18 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 07:24 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 07:32 | < McMartin> | Oh hey, nice |
| 07:32 | < McMartin> | One more fact to make the Z80 a little less terrible |
| 07:33 | < McMartin> | The non-relative branch instructions take the same amount of time irrespective of whether or not the branch is taken |
| 08:31 | | macdjord [macdjord@Nightstar-ahbhn1.cable.rogers.com] has joined #code |
| 08:31 | | mode/#code [+o macdjord] by ChanServ |
| 08:32 | < Vornlicious> | Neat |
| 09:19 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 09:19 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 10:59 | | Jessikat` [Jessikat@Nightstar-o4q.jtr.132.82.IP] has quit [[NS] Quit: Bye] |
| --- Log closed Wed Mar 28 11:12:14 2018 |
| --- Log opened Wed Mar 28 11:12:32 2018 |
| 11:12 | | TheWatcher [chris@GlobalOperator.Nightstar.Net] has joined #code |
| 11:12 | | Irssi: #code: Total of 31 nicks [23 ops, 0 halfops, 0 voices, 8 normal] |
| 11:12 | | mode/#code [+o TheWatcher] by ChanServ |
| 11:12 | | Irssi: Join to #code was synced in 12 secs |
| 11:52 | | Alek [Alek@Nightstar-7or629.il.comcast.net] has quit [Ping timeout: 121 seconds] |
| 11:55 | | Alek [Alek@Nightstar-7or629.il.comcast.net] has joined #code |
| 11:55 | | mode/#code [+o Alek] by ChanServ |
| 12:20 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code |
| 12:20 | | mode/#code [+o himi] by ChanServ |
| 12:25 | | Degi [Degi@Nightstar-ibr0nm.dyn.telefonica.de] has joined #code |
| 15:15 | | Degi [Degi@Nightstar-ibr0nm.dyn.telefonica.de] has quit [Connection closed] |
| 15:26 | | Vornlicious [Vorn@Nightstar-ipsdsc.sub-174-211-2.myvzw.com] has quit [[NS] Quit: Bye] |
| 15:26 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 15:55 | | You're now known as TheWatcher[d00m] |
| 15:59 | | Kindamoody is now known as Kindamoody|out |
| 16:12 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 16:12 | | mode/#code [+o gnolam] by ChanServ |
| 17:05 | <&jerith> | McMartin: My Rna Transcription solution uses the first Erlang throw/catch I've ever used for real. :-) |
| 17:14 | | You're now known as TheWatcher |
| 17:19 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has joined #code |
| 17:54 | | Jessikat [Jessikat@Nightstar-val89s.dab.02.net] has joined #code |
| 17:59 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 19:10 | <&[R]> | https://www.eff.org/deeplinks/2018/03/responsibility-deflected-cloud-act-passes <-- Patriot Act 2.0 |
| 19:13 | < McMartin> | This one is weird to me |
| 19:13 | < McMartin> | THe big case that was sparking this? Both sides said they'd rather have this pass than fight the case |
| 19:14 | <&[R]> | Which case? |
| 19:15 | < McMartin> | Microsoft vs USA, regarding, essentially, to what degree MS can use Ireland as a data haven |
| 19:17 | <&jerith> | I should note that this makes my professional life somewhat difficult, because it effectively stops us using any services owned or operated by US companies. |
| 19:18 | <&jerith> | Perhaps not for company email or whatever, but certainly for our user data. |
| 19:19 | < McMartin> | I expect to see a fair amount of that in the near future, yes. |
| 19:19 | <&jerith> | Considering that some of our projects involve national-scale health systems and some of our projects (potentially overlapping with the health stuff) operate in countries with oppressive governments. |
| 19:20 | < McMartin> | Yeah. The EFF and ACLU are both correct that this is an extremely ill-considered law. |
| 19:20 | < McMartin> | (So is FOSTA, for the record, but the people suffering immediately for that one are mostly people that everyone loves to hate, but it ended up passing over the objections of the constituency it was notionally serving.) |
| 19:21 | < McMartin> | Things will get even more exciting if Oracle manages to prevail on its renewal of its claims against Google re: the Java APIs |
| 19:22 | <&jerith> | Add the no-US-funding-for-anything-that-mentions-pregnancy-termination law and the current administration is doing significant harm to foreign organisations attempting to improve the lives of foreign people. |
| 19:22 | < McMartin> | Because even if that prevails on a narrow reading, I see no way that doesn't also apply to tomcat, and 90% of the dynamic web in the US will be torn down as an unacceptable legal risk. |
| 19:23 | < McMartin> | The current government in the US is also aggressively hostile to US organizations attempting to improve the lives of US people. |
| 19:23 | <&jerith> | Yeah. |
| 19:23 | <&jerith> | My point was more that they are systematically interfering in other people's jursdictions. |
| 19:24 | < McMartin> | I mean, so was the initial funding |
| 19:25 | < McMartin> | The question that was legit at issue with the MS case boiled down to "to what degree can US companies evade oversight by the US by taking advantage of other people's jurisdictions" |
| 19:25 | < McMartin> | And while it's a not unpopular view on the Internet that of course the US government should not be able to get information about what US companies do involving US citizens, even with court orders, that isn't likely to fly overall |
| 19:26 | <&jerith> | There are presumably legitimate channels for acquiring the relevant information from the other jurisdictions. |
| 19:27 | <&jerith> | Without violating other people's legitimate privacy legislation. |
| 19:27 | < McMartin> | The dodginess of this wrt Ireland was apparently specifically one of the issues. |
| 19:27 | | Degi [Degi@Nightstar-ibr0nm.dyn.telefonica.de] has joined #code |
| 19:27 | < McMartin> | Like, MS was evading subpoenas successfully by saying "we actually don't know how much of this order we can comply with without violating Irish law" |
| 19:28 | <&jerith> | The appropriate response to that is not a US law requiring MS to violate Irish law, though. |
| 19:30 | < McMartin> | I would not conclude that this is what it actually said without checking the bill itself. |
| 19:31 | <&jerith> | Agreed. |
| 19:31 | <&jerith> | But I've seen that intepretation multiple times. |
| 19:31 | < McMartin> | If it is, MS will spark an international incident basically immediately and it's not like the EU is going to be shy about picking the fight. |
| 19:32 | <&jerith> | And it's about the only way to bypass the "this violates Irish law" thing. |
| 19:33 | < McMartin> | That does raise the reciprocal question |
| 19:33 | < McMartin> | To what degree are you relying on US protections for US companies to protect citizens from their own repressive governments? |
| 19:34 | <&jerith> | We're relying on things like hosting certain services outside the jurisdictions of those governments. |
| 19:35 | | RchrdB [RchrdB@Nightstar-qe9.aug.187.81.IP] has joined #code |
| 19:35 | <&jerith> | But, and I can actually talk about this now that it's public and the NDA no longer applies, we are starting to provide access to various services over WhatsApp. |
| 19:36 | < Emmy> | neat |
| 19:36 | <&jerith> | Aside from being vastly cheaper and nicer to use than SMS/USSD, it largely mitigates the ability of telcos to monitor the messages. |
| 19:37 | <&[R]> | Except now Trump fucked that up |
| 19:37 | < McMartin> | Not sure how up-to-date this is compared to the other things, but there's one obvious difference here between the MS case and what you face that redounds to their benefit and not yours |
| 19:38 | < McMartin> | https://www.geekwire.com/2018/congress-considers-tech-backed-cloud-act-privacy-human-rights-groups-raise-concerns/ |
| 19:38 | < McMartin> | Specifically: it sounds an awful lot like this bill is a no-op absent additional treaties being signed with the relevant nations or organizations. |
| 19:38 | <&jerith> | Except now WhatsApp may be required to just hand over all the data (even if it's just metadata with end-to-end crypto on the actual messages) to anyone the US government decides should have it. |
| 19:38 | < McMartin> | If part of the issue is that the legal procedures for doing discovery across the Atlantic really are getting overwhelmed in the US and EU, the US and EU will be able to work something out that MS will be OK with. |
| 19:39 | < McMartin> | But repressive governments will happily work with Trump to fuck over people Trump doesn't care about. |
| 19:40 | < McMartin> | Any nation that really wants to do its damage still can, of course, by simply forbidding data about their citizens to leave their borders in the first place. |
| 19:40 | < McMartin> | Then all you have to do is check what's accessed and make *that itself* the crime. |
| 19:41 | < McMartin> | That is at the moment an unusually high opportunity cost |
| 19:41 | < McMartin> | I'll be shocked if the EU nations don't do it in the next five years though, on "nobody else is trustworthy" grounds. |
| 19:42 | <&jerith> | The South African Protection Of Personal Information Act (POPI) specifies a bunch of requirements for the storage and processing of personal information. |
| 19:43 | <&jerith> | Kind of like PCI/DSS or whatever for credit card data. |
| 19:44 | <@iospace> | https://arstechnica.com/information-technology/2018/01/now-even-youtube-serves-ads-with-cpu-draining-cryptocurrency-miners/ |
| 19:45 | <&jerith> | I'm not familiar with all the details, but our security team has told me we're allowed to keep certain things in certain providers in certain jurisdictions (like AWS in Ireland) under certain conditions without violating POPI. |
| 19:45 | < McMartin> | I think if the EU plays its cards right they can preserve this |
| 19:45 | < McMartin> | What would be necessary would be data segregation of US-sourced data from other ones within the server |
| 19:46 | < McMartin> | Er, rather, data about US citizens |
| 19:46 | <&[R]> | Which is a massive pain in the ass |
| 19:46 | < McMartin> | Sure is, but people are assholes, so ass pain is a fact of life |
| 19:46 | <&jerith> | What this new law *may* do is suddenly change those rules such that it is no longer legal under South African law to use those services. |
| 19:47 | | * McMartin nods |
| 19:47 | < McMartin> | Worth taking precautions, but I wouldn't treat it as a fait accompli |
| 19:47 | < McMartin> | After all, the EU will not do anything to make it suddenly illegal for *EU* companies to use those services. |
| 19:48 | <&jerith> | Sure, but it makes an already difficult issue more difficult. |
| 19:48 | < McMartin> | If the US becomes a sufficiently recalcitrant customer that they nevertheless want to keep doing business with, it's the US that will need to be split out. |
| 19:48 | < McMartin> | That's true, but that can also be phrased "there's a lot of nasty issues in this space that we've been actively neglecting for a long time". |
| 19:49 | <&jerith> | "[...] without a new procedure for handling these requests, foreign countries could simply declare that all data generated by its citizens be stored within its borders, making it applicable to its laws." |
| 19:50 | <&jerith> | This is already the case with certain kinds of data. |
| 19:50 | < McMartin> | And I can easily see the EU deciding that That's Fine. |
| 19:50 | < McMartin> | If the US wants to be nosy, lock them out. |
| 19:50 | < McMartin> | That's what puts *you* in the bind but not them, as it were. |
| 19:51 | < McMartin> | (And MS doesn't care because they can just set up the subsidiaries and be done with it) |
| 19:51 | <&jerith> | Also, the damage to us has mostly already been done. |
| 19:52 | <&jerith> | Because the fact that such a law could pass so quietly and with so little challenge means that we can't trust that it won't happen again, even if it goes away this time. |
| 19:53 | < McMartin> | Yeah, but that's always been true |
| 19:53 | <&jerith> | It hasn't been this publicly true in a while, though. |
| 19:53 | < McMartin> | As an American, the idea that Americans would be permanently reliable resources for other nations is kind of jarring to me, I think in part because the current forces at play in our government have always been extremely present in domestic politics |
| 19:53 | < McMartin> | Yeah |
| 19:54 | <&jerith> | And in most cases it's been stuff that doesn't directly effect other nations so much. |
| 19:54 | <&jerith> | *affect |
| 19:55 | <&jerith> | Although if someone secedes over this I guess it *could* effect other nations... |
| 19:55 | < McMartin> | That's a highly unlikely outcome. |
| 19:57 | < RchrdB> | wait, is this CLOUD act thing a complement to the PATRIOT act? |
| 19:57 | <&jerith> | A marginally less unlikely outcome is civil war somewhere (possibly the Balkans) as a result of someone invoking the provisions of the CLOUD act. |
| 19:57 | <&[R]> | RchrdB: it's a mass surveilance thingy |
| 19:57 | < McMartin> | RchrdB: Not at all unless you mean "they both end up incorporated into the US Code" |
| 19:57 | <&[R]> | IE: same class of damage |
| 19:58 | < McMartin> | It is about relaxing the requirements for Mutual Legal Assistance processes. |
| 19:58 | < McMartin> | The concern is that this can be used to do end runs around nations' notional protections for their own citizens |
| 19:59 | < RchrdB> | oh! |
| 20:00 | < RchrdB> | It creates a system where the US can offer a MLAT to some other state, which in return will be enticed to give similar access to the US? |
| 20:00 | < McMartin> | It weakens the requirements on the US side. |
| 20:00 | < McMartin> | We already can and indeed have offered and accepted many a MLAT over the years. |
| 20:01 | < RchrdB> | and that MLAT can involve an assumption that the non-US country's judicial branch is setting as stringent a standard for access requests as the US's would be |
| 20:01 | < McMartin> | To the extent that there is a problem here, the US system for processing MLA requests is presently preposterously overworked and takes years to get shit through |
| 20:01 | < RchrdB> | so it expands the permissiveness of the terms that a MLAT could have in it |
| 20:02 | < McMartin> | Right, which in turn could, as jerith notes, mean that US companies are no longer legal partners for third nations' operations. |
| 20:03 | < McMartin> | While I said "the problem" above, that's from the outside. |
| 20:04 | < McMartin> | It seems like the inciting incident here is a protracted clusterfuck involving the MS Azure Cloud, the US, and Ireland |
| 20:05 | < McMartin> | So a big factor in how bad this is or isn't, IMO, is likely to be how aggressive Ireland (and by extension the EU) is in response. |
| 20:05 | < McMartin> | As a rule the EU seems to be much more aggressive about this stuff than the US is, at least with respect to their own citizens. |
| 20:05 | <&jerith> | Another legitimate concern (although one that doesn't affect me as much) is that the reduction in due process could potentially expose a whole lot of information pertaining to US citizens to foreign governments. |
| 20:06 | <&[R]> | Shhh shhh |
| 20:06 | < McMartin> | Yeah, almost any issue with it is going to end up reciprocal |
| 20:06 | <&[R]> | Putin is totally not a part of this |
| 20:06 | < McMartin> | But that's also a case where the EU's response will be... interesting. |
| 20:07 | < RchrdB> | in unrelated news, someone dun goofed with page tables https://twitter.com/0xabad1dea/status/978999381084033024 |
| 20:07 | < McMartin> | [R]: Let's be real here, the AIVD is just as likely to be Entirely Uninvolved~ |
| 20:07 | < RchrdB> | "In short - the User/Supervisor permission bit was set to User in the PML4 self-referencing entry. This made the page tables available to user mode code in every process. The page tables should normally only be accessible by the kernel itself." |
| 20:08 | <&[R]> | 'Ah the old sage wisdom of “you can’t break into my house if I let you in”' |
| 20:08 | < RchrdB> | indeed |
| 20:08 | < RchrdB> | The patches MS put out in Jan and Feb this year for mitigating Meltdown contained this, apparently the patches they put out in March fixed it |
| 20:09 | <&[R]> | Wonder why they even had that "feature" |
| 20:09 | < McMartin> | The initial Meltdown/Spectre patches outside of firmware involve playing silly buggers with the cache and TLB, AIUI |
| 20:09 | < McMartin> | Basically disabling them with software that's intentionally bullshit to make information that leaks out of it useless |
| 20:10 | < McMartin> | Looks like some of it was *too* bullshit >_> |
| 20:10 | < Mahal> | Yall have seen Total Meltdown I gather? |
| 20:10 | < McMartin> | ... not by that name? |
| 20:10 | < Mahal> | Win7 and 2k8. |
| 20:10 | < RchrdB> | is this a film or an article? |
| 20:10 | < Mahal> | The ms patch that 'fixed' meltdown and made the page table fully and easily accessible |
| 20:10 | < McMartin> | Oh |
| 20:10 | < McMartin> | That's literally what we're talking about right now. |
| 20:10 | < Mahal> | Yes I was asking for clarification :) |
| 20:11 | < McMartin> | I just heard of it three minutes ago :) |
| 20:11 | < McMartin> | I did not know tha tname. |
| 20:11 | <&jerith> | Irrelevant to me personally, thank Eris.~ |
| 20:11 | < RchrdB> | ah! so, you heard of it before I did :) |
| 20:13 | <&jerith> | On the other hand, I have spent the past two days largely sitting around waiting for incompetent functionaries in Uganda to give me a working network so I can turn the systems they moved across the city back on. |
| 20:14 | <&jerith> | Yesterday they didn't have IPs for us, despite continuous requests since November for said IPs. |
| 20:14 | < Mahal> | Sigh |
| 20:15 | <&[R]> | Been there, I feel your pain man |
| 20:15 | <&[R]> | Or woman |
| 20:16 | <&jerith> | Today they had IPs for us, but their firewall blocked all outbound traffic, then briefly blocked no outobund traffic, then blocked all outbound traffic that was not DNS or HTTP on port 80, then blocked all outobund traffic not on a list of ports we provided, except NTP was broken the time except for the brief "no firewall" interlude. |
| 20:16 | <&jerith> | *the whole time |
| 20:17 | <&jerith> | Turns out quite a lot of the distributed systems in our hosting cluster get upset when all the bits think the time is different. |
| 20:19 | <&jerith> | We quoted two people for three days for this migration, spent at least twice that justifying our quote, and were then told to do it in two days. |
| 20:20 | <&jerith> | We took three days anyway (including some after-hours work on Monday that pissed me off immensely) and ended at 5pm (our time) today with the important bibts of the system still turned off. |
| 20:20 | <&jerith> | *bits |
| 20:21 | < Emmy> | ah. i can imagine the frustration |
| 20:21 | < Emmy> | i had a somewhat similar episode monday with repairing a database |
| 20:22 | <&jerith> | Total accomplishments: backups and clean shutdown before the move (on Monday night), and partial setup in the new DC. |
| 20:22 | < Emmy> | because the last backup was 2.5 weeks ago. |
| 20:22 | < Emmy> | ...yeah they gonna pay. |
| 20:22 | <&jerith> | The partial setup would have taken two hours rather than two days if they'd actually given us the IPs and a non-broken network. |
| 20:23 | <&jerith> | We are now, as per their directive, no longer allowed to work on this until more budget is approved. |
| 20:29 | | Kindamoody|out is now known as Kindamoody |
| 20:30 | <&jerith> | When I said "waiting around" up there, I meant "spending 15 to 30 minutes tinkering with property tests in Elixir, between 5 minute sets of checking whether the Uganda stuff has changed". |
| 20:33 | <&jerith> | Elixir's StreamData stuff (which is a candidate for stdlib inclusion) is pretty cool, though. |
| 22:45 | < McMartin> | jerith: You got me to run the numbers on body vs tail recursion for cost analyses |
| 22:45 | < McMartin> | Putting the "Decrement Part of Register" in "Contents of Decrement part of Register" |
| 22:46 | <&jerith> | McMartin: I can't remember where the blog post I saw regarding that was. |
| 22:46 | < Mahal> | jerith: that sounds like a comprehensive clusterfuck. |
| 22:47 | < McMartin> | I was sleazy and treated all operations as constant time and space |
| 22:47 | < McMartin> | Which is still enough for a good rule of thumb |
| 22:48 | < McMartin> | (Basically: if your number of iterations is linear in the number of items in your input or your output, body recursion will probably be faster by some constant factor and may not consume additional max space and will almost certainly produce less garbage.) |
| 22:49 | <&jerith> | Ha! Here it is: https://pragtob.wordpress.com/2016/06/16/tail-call-optimization-in-elixir-erlang-not-as-efficient-and-important-as-you-probably-think/ |
| 22:49 | < McMartin> | Unless making a tail call is significantly cheaper than making a normal function call, my analysis was more shallow and applied even to Scheme and ML. |
| 22:50 | <&jerith> | Mahal: Yes, and this is only the latest in the long series of clusterfucks in that project. Aside from this server migration, we're also in the process of handing over the whole shebang to someone else and getting the hell out of it. |
| 22:53 | < McMartin> | jerith: This is a great article and you should totally link it in that conversation. |
| 22:54 | <&jerith> | I'll let you do that. I originally got it from a conversation on someone else's Elixir solution. :-) |
| 22:55 | <&jerith> | The important thing is that there's no call stack depth limit, so you can recurse as much as you have RAM for. |
| 22:55 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has quit [Ping timeout: 121 seconds] |
| 22:55 | < McMartin> | Sounds good. |
| 22:56 | <&jerith> | Just make sure you tail-recurse in any loops you don't intend to return from.~ |
| 22:56 | < McMartin> | Also, this gist makes me happy and I should figure out how to make the erlang shell do it so I can look at Elixir's internals. |
| 22:56 | < McMartin> | A loop that iterates infinity times is far larger than its largest static input.~ |
| 22:56 | < McMartin> | https://gist.github.com/sasa1977/73274c2be733b5321ace |
| 22:59 | <&jerith> | Where'd you find that? |
| 22:59 | < McMartin> | A couple links deep from yours. |
| 23:01 | <&jerith> | Aside from the Code.compile_string() and IO.inspect() calls, that's basically Erlang code. |
| 23:03 | <&jerith> | I can never remember the format string syntax exactly, but IO.inspect(foo) is more or less equivalent to io:format("~p~n", [Foo]). |
| 23:03 | < McMartin> | You may recall I was curious about how tightly coupled BEAM was to Erlang's powers and pitfalls, and how Elixir went about improving those. |
| 23:03 | < McMartin> | This looks like The HOWTO I Needed |
| 23:04 | <&jerith> | As of OTP 20, BEAM's grown support for an extra layer of abstraction in its AST storage. |
| 23:04 | < McMartin> | Fedora's OTP implementation seems to be a bit stale. |
| 23:05 | <&jerith> | This makes it possible to get Elixir AST in there instead of Erlang's while still allowing all the tools that expect Erlang AST to work. |
| 23:06 | <&jerith> | Before that, Elixir-generated .beam files were identical to Erlang-generated ones. |
| 23:06 | <&jerith> | Well, there are some extra library modules for Elixir's stdlibb. |
| 23:08 | <&jerith> | The really interesting stuff in the language (macros, syntax stuff, etc.) is all essentially preprocessor work.~ |
| 23:08 | < McMartin> | One of my be questions is indeed the degree to which Elixir is "erlfront" |
| 23:10 | < McMartin> | Does the : prefix mean "atom"? |
| 23:10 | <&jerith> | Yes. |
| 23:11 | <&jerith> | Elixir uses TitleCase for modules, snake_case for identifiers, :colon_prefix for atoms. |
| 23:12 | <&jerith> | "" for Elixir strings (UTF8-encoded binaries), '' for Erlang strings (which it challs charlists). |
| --- Log closed Thu Mar 29 00:00:27 2018 |