« first day (1875 days earlier)      last day (3102 days later) » 

user55340
7:00 PM
@GlenH7 and that evil mod could point out that on their meta... Just to try to prevent a repeat.
 
@MichaelT argh why do you challenge me thusly?? I'm now dying to ask on there about the Qi energy ratio from quartz crystals to amethyst... 3 quartz to one amethyst? Do the number of sides cut matter?
 
user41796
Just to help close out an older meta thread. The question was migrated to Engineering but is too broad of a question for the Engineering site. It was closed on Engineering and the migration was rejected which sent the question back to Physics. — GlenH7 16 secs ago
 
user55340
@JimmyHoffa I'm still trying to figure out how to ask about oil engineering questions. Homeopathic oil engineering.
 
@MichaelT solution: become a moderator
 
user41796
@MichaelT So Air, Trevor, and I can race to close and delete your question?
 
7:06 PM
@enderland haha yeah right, like he wants that headache. @MichaelT's nobody's fool.
 
user55340
@enderland yes, the solution is part of the problem. How much do you need to dilute it?
 
@MichaelT well there are about 150 stack sites... so 1 part per 150?
@LightnessRacesinOrbit heh so I am considering applying for a job, and they use Docker, and it was funny when my friend asked me about if I'm familiar with their tech stack and one of the things is Docker "yeah was convincing someone to use it just this week" ;)
 
@enderland still enjoying it - really making progress now. Got my wrapper script done and a few images ready for various toolchains we use
most questions answered
 
@LightnessRacesinOrbit I wonder if anyone directly uses it or if 100% of people use a wrapper of some sort :P
 
user55340
Oh, that's it - homeopathy.se so that glen can migrate all the pseudoscience engineering questions there... Where he is a mod and can dup them to the one canonical question.
 
7:09 PM
@enderland at least my wrapper takes care of (I think) all potentially orphaned containers and dangling images
it seems to be really easy to create them
can imagine my team making a massive mess of our server with free reign around Docker
 
@LightnessRacesinOrbit yeah. lol. that was something to worry about (especially if your images and containers are big)
 
@MichaelT there can't actually be a homeopathy.se... no. I refuse. What's next, TimeCube.se? Yeah, so much for authoritative anything
 
"oh I have TWENTY GB OF DOCKER IMAGES ALIVE AND NOT BEING USED ZOMG"
 
user41796
@JimmyHoffa Not even a proposal on Area51. Yet.
 
@GlenH7 no that was when running local builds in a VM :P
 
7:11 PM
Numerology.se: All answers are right, because if I add 10 and 13 I get 1013 which can be divided by two to get 10/13 which if you squint looks like the number 7, and since 7 is a lucky number clearly all your answers are definitely right.
 
user41796
@enderland I could see where that's a problem then
 
@GlenH7 my team was building a very fancy wrapper for Docker, basically, I did a lot of testing for them when I first started since I had to learn a billion things
 
I wouldn't say mine's "very fancy"
 
I'm at risk of being banned from asking questions. All my questions were deleted and I'd like to read them to find out why to avoid asking bad questions in the future. Could anyone possibly look those up for me with a link? Thanks!
 
@LightnessRacesinOrbit your team is like 5% the size of the team where we were though, I think 250 or so devs used it for building
 
7:14 PM
@TK-421 banned from here or SO?
 
here
 
@cube oh sorry about that TimeCube jab, I didn't see you there
 
but it hides the nasty UUIDs while allowing human-readable container names to overlap across user accounts (by actually naming containers itself, and managing the human-readable descriptions within its own abstraction layer), and auto-deletes containers when they're exited from (so no stopped containers piling up!)
 
101 6
do you see any deleted questions there?
 
7:14 PM
enderland doesn't know
you can see your deleted quesitons in a certain time frame, I think
 
no, and yeah was happy to take that username
 
@TK-421 Real easy trick to ensure you never get Q-Banned here and learn what is and isn't on-topic: Anytime you have a question you want to ask, come in here and ask if it's on/off topic and why before posting it. Alternatively, read the HelpCenter and hope you understand it well
 
annoyingly even though I can see deleted questions, they don't show up on user profiles, they have to be linked directly
 
user41796
@TK-421 if the questions aren't showing up on your profile, you'll need to ping a site mod in order to hunt them down
 
it also hides all hub and "<none>" images from docker images so that the choice of image to use when instantiating a new build environment is ... the list of images I've created with our toolchains in
 
7:15 PM
@Ixrec moderators can but programmers mods aren't lurking here
 
We'll gladly let you know if whatever question you want to ask is a good or bad fit and why
 
Good tip, sounds fair enough. Thanks.
 
ok...what's Docker? I've heard of it but never looked at it. Some kind of sandboxing thing?
 
@TK-421 this chat is very helpful if you are interested in learning how to do that, we are pretty quick to vote to close off topic stuff but when people want an on topic thing it's rare and very well received
@JimmyHoffa yeah
 
user55340
I've written a bit on meta also.
 
7:16 PM
devenv listimages, devenv create <toolchain>, devenv list, devenv reattach <n>, devenv delete <n>, devenv exec <toolchain> <cmd> and that's about it
 
@JimmyHoffa lets you create reproducible build environments
 
@MichaelT whatever, you paid others to write that shit, we all know better.
 
user41796
@TK-421 Another trick is to look at your reputation changes That may provide the titles and make it easier for the mods to find your questions.
 
user55340
30
Q: Why was my question closed or down voted?

MichaelTHelp! My question was closed (or down voted). Why? And what can I do about it? Downvoted Duplicate Off topic What technology to take up next Recommend a software library, tool, book, research paper, blog, forum, or other resource Career or education advice (MSE) Which computer science / prog...

 
every container uses host networking and mounts host ~ to container /host/
 
user55340
7:17 PM
@TK-421 Start there, follow links.
 
@enderland relatively important. How? Controlls dependencies? Toolchain stuff? What type of build environments?
 
normal usage goes into bash for pissing about playing with source code and iteratively rebuilding; exec will do what you tell it (and will be what Jenkins uses, I guess?)
 
Found a check box for "show removed questions" where reputation changes are. That works! Thanks guys, I'll stop in before asking another question.
 
@JimmyHoffa Yeah it's a VM manager for not-VMs with a simple image-building script language, cloud storage and clever transactional caching .. where the cloud storage includes like a gazillion ready-made base images including like every distro ever
 
user41796
@TK-421 and if things aren't clear while reviewing the questions, posting a direct link to the question here in chat will allow 10k+ users to go look at it too
 
user55340
7:19 PM
 
sounds good, thanks
 
@LightnessRacesinOrbit sounds like it's usurping your active runtime and dropping your storage root into a loop storage instead
 
@JimmyHoffa uwotm8
 
user55340
If you don't have down voted (that one wasn't) questions deleted, should be ok IIRC.
 
@TK-421 yeah, if you edit something enough people in here have undelete votes that you could pretty easily get it undeleted once it's on topic too if you want
 
7:20 PM
@LightnessRacesinOrbit does it take your running environment, fake stuff up into one that has a storage rooted in some folder or tar-like-file, and then try to overlay a shell from that new location?
 
> How does a container work?
A container consists of an operating system, user-added files, and meta-data. As we’ve seen, each container is built from an image. That image tells Docker what the container holds, what process to run when the container is launched, and a variety of other configuration data. The Docker image is read-only. When Docker runs a container from an image, it adds a read-write layer on top of the image (using a union file system as we saw earlier) in which your application can then run.
 
@JimmyHoffa no
@JimmyHoffa well, sort of
it's a lot more sophisticated than a chroot jail but you're on the right track
 
@LightnessRacesinOrbit I should hope it's more sophisticated than that, if that's all it were it wouldn't deserve much notice.
What does it give you over a VM?
 
@JimmyHoffa indeed
 
Lower storage requirements?
 
7:24 PM
@JimmyHoffa Lower storage requirements, reusability, shares a kernel so it's a heck of a lot lighter to run
 
user55340
Iirc, easy deployment.
 
A Docker image is light enough to be designed to just instantiate for a single process (e.g. a Make invocation) then exit
 
@MichaelT VM deployment/cloning/starting/automation stuff isn't especially difficult these days
 
You don't need to move your whole consciousness into a separate logical system
 
@LightnessRacesinOrbit Blasphemer! That's non-monadic!!
 
7:25 PM
That being said, you can choose to launch bash if you like and do what you like in it for months
 
@LightnessRacesinOrbit note on this: if you have idiot developers people trying to get 100% code coverage with tests who add a test like "which OS am I on?" and use uname it's going to return the host system because of the shared kernel
 
@enderland that's ok - my idiot developers read from /etc/*-release :)
 
@enderland I once saw code some dumbass DBAs wrote that changed a sprocs logic based on server_name() - people who base logic on shit like that should be beaten
 
we actually found all sorts of tests that started to fail in Docker when we started switching over, and assumed that it was the docker image - but it turned out docker was a better build environment than our previous build system (lol) so really, those tests were actual fails that were undetected because of the build system being slightly different than deployment system...
 
7:28 PM
most of them weren't too serious but.. still unnerving
 
The main thing I'm hearing is it's somewhat lighter weight on storage and system resources (VM automation is very fast too; with enough resources..), at the expense of a more complicated configuration situation?
 
@JimmyHoffa one advantage that docker has is you can do local builds a lot easier too
not just on a CI system
 
@enderland with a good VM setup you can do so as well
 
@enderland once I had a test fail in the test environment only when I used the number 1e10 in it
 
7:30 PM
1e9 was fine, and all numbers were fine in dev, but somehow 1e10 in test broke stuff
 
A real big industry grade VM lab system is pretty amazing- and also hundreds of grand up to millions depending on what you want.
 
@JimmyHoffa it's the "with enough resources" part
we're a tiny company
 
@LightnessRacesinOrbit I'm not encouraging a VM over docker, just trying to contrast
 
I'll be happy if I can get basic new hardware for this server while I'm rebuilding it
 
so the main motivator isn't features so much as light weight on resources. Would you say the configuration/initial setup is particularly onerous?
 
7:32 PM
I would not
I've been at this for three days or so and I have three images the way I want them (and they're quite large as there's a lot to install for our build thingies), with three to go, a wrapper script ready to go, and a decent understanding of where everything goes and what everything does
 
neat. So, now that I know enough about docker to have some idea when I should look at it, I will never look at it for 4 years and have no recollection of it by then, which is ok because it will be obsolete, replaced by levi or canvas by then...
 
My main concern at the moment is that I'll struggle to get people to start using Docker in their own individual environments, so we'll have to keep our environment set up scripts around in parallel with the Dockerfiles (which ultimately do the same job)
but frankly as long as our main build server can build and test everything we produce, I'm happy
 
@MichaelT perl is lodging a thought into my brain which may be misremembered... I have these stream processors which are a -> [a] and I put them together continually flattening the results. But what if I really need (a,[b]) -> [a] to make a decision, should I come up with a way of joining a -> [a] and a -> [b] so I can get the input for a (a,[b]) -> [a] or an (a,[b],...) -> [a] for N number of x -> [x]s being joined...
that's sort of a part of junctions in perl, no?
 
One downside is the shared filesystem, but I think SELinux has taken care of that for me: if I set the relevant context only on /home/ (recursively), nobody can mount their container anywhere dangerous and start breaking things
at least, not critical things
and my wrapper script will only mount their homedir anyway
so (obviously) a Docker container isn't quite so fully contained; there are ways to improve that but it gets a bit more complicated, relying on SSH, Unix sockets, stuff like that to actually transfer files in and out (somewhat important for a build environment!)
 
right, the downside that a VM handles for you, but it's significantly lighter weight it sounds like as it's not stealing ram/cpu, just costing storage and considerably less of that than a VM
 
7:41 PM
much easier to have a container do SVN checkout and build, into my host homedir, and that's it (SVN checkout inside container because SVN version may differ and lead to WD version incompatibilities)
@JimmyHoffa right
 
Can docker snapshot/rollback etc? You could make the shared portion wide open and allowed for everyone so long as it's copy on write and automatically rolls back
 
@JimmyHoffa please be less vague
"the shared portion"?
 
No! No no no no!!
 
(yes, it can - each container can be "committed" to itself become an image, which may then be itself instantiated to be a new container)
 
@LightnessRacesinOrbit you mention the file system is shared, but it's sandboxed per user so that different users sharing the same file system aren't colliding so long as they stay in their /home/
 
7:43 PM
((so you'd rollback by stopping using your new 'child' container and go back to the old one))
@JimmyHoffa ah right. filesystem operations won't roll back
that's the thing - access to the host filesystem is as-is (assuming you set up your container to mount part of it somewhere). that's the one sort of fragility I see
 
@LightnessRacesinOrbit because they're not copy on write. There's no way to make docker's sandboxing force all writes to be copy deltas that it presents to that sandbox scope ?
 
@JimmyHoffa it's not sandboxing the host FS
so no
 
I would think it would do such, it seems like a logical feature they would have wanted right off
 
well, don't quote me on it - I'm still new
 
@JimmyHoffa you can host existing filesystems but nothing is by default
 
7:46 PM
2
A: Can I mount docker host directory as copy on write/overlay?

WoxxyThis is what I do: On the host: Load the directory as read only. docker run --privileged -v /path/on/host:/path/on/client-read-only:ro -it ubuntu /bin/bash On the client: On the client use OverlayFS over the read-only directory mounted from the host. mount -t overlayfs none -o lowerdir=/pat...

looks like you can do it
 
aww didn't beat the LMGTFY Jimmy. :P
 
it's just a mounting trick forcing the mount to go through some funny kernel FS module that acts as a copy-on-write fs to begin with
likely similar to how live cd's work and such, they pretend to let you write stuff to disc and will present the data back to you et al, but when you reboot they're gone because they write to a fake space
 
which is completely useless for a build environment
so I go into my container and build my project.. come out of it to retrieve my binaries, and they're not there. no point in mounting in the first place
I see the use case for this but it's orthogonal to what I've been talking about :)
 
@LightnessRacesinOrbit put them on a network drive? :)
have a folder mounted using normal mount rather than that funny FS
 
@JimmyHoffa releases end up there. random builds during development won't and shouldn't
 
7:48 PM
@LightnessRacesinOrbit it'd be a weird situation you't want to run your binaries in a different environment than you built them in though (maybe? I guess there are uses for this)
 
@enderland you wouldn't..
but when I make the binaries it's because they're going to be transferred somewhere to be used
 
@LightnessRacesinOrbit one use case for this is viewing/editing source code outside docker and building using the source tree
 
that "somewhere" may just be the releases area on the network, so fine I can just mount that
 
@LightnessRacesinOrbit so just mount everything except for /bridgeToHost with copy-on-write, builds can put stuff in that location and things will be available to the host then
 
but it may also be some piece of kit dangling off our LAN somewhere this afternoon, and my devs need the flexibility to copy the binary somewhere convenient
 
7:50 PM
also you probably won't want to have to go do some obscure thing if you want to look at some build artifact for some reason, so you might as well mount the source/build tree (for local builds)
 
@JimmyHoffa I just don't mount anything but /bridgeToHost at all - no need
@enderland yeah
 
@LightnessRacesinOrbit was just speaking to a solution to your "everyone's sharing the host FS" issue, with copy-on-write mounting you'd have isolated sandboxes with one shared host folder. It would be monadic then!! :D
 
I really want invoking Docker to be more of a glorified GCC invocation (simplifying greatly) than a separate PC to arse around in .. even though the option to go in through executing bash will be there
@JimmyHoffa well that would be cool and all but ultimately I'd still need to trust everybody to actually, y'know, do that. I guess the point I was trying to make is that it's trivial to ... not do that. which of course is not the case with a VM, in the main
 
@LightnessRacesinOrbit with devs sharing the host though and executing entire tool chains and scripts...how long before they're running scripts that are writing shit all over who knows where..
@LightnessRacesinOrbit oh, you aren't controlling their entry point?
 
@JimmyHoffa hence the wrapper :)
 
7:54 PM
@JimmyHoffa I'm providing a pleasing abstraction around it but ultimately if they want to use Docker at the Docker level there's not much I'm going to try to do about that
@JimmyHoffa but again my point was as a description of Docker itself - and speaking of Docker itself, the fact is that this is the case :)
 
@LightnessRacesinOrbit ah gotcha. Levi's will solve these problems I'm sure..
 
@LightnessRacesinOrbit yup, I did that a few times when testing, "so I need to do something our wrapper doesn't do yet..."
 
@LightnessRacesinOrbit cool, makes sense. Haskell has a variety of such sandboxing things available - especially with Stack now which is bloody awesome.
and of course, it's monadic.
 
@JimmyHoffa well, again, only /home/ will be mountable at all thanks to SELinux config (which only I control) so there are limits to what they can bugger up, even if they abandon my script, go to the Docker level and really try
and that's at the cost of their job, so...
 
@LightnessRacesinOrbit no! Don't job them! (why does "brain" someone refer to debraining them?)
Jan 5 '13 at 4:47, by Jimmy Hoffa
I'm helping :D
 
7:56 PM
is it possible that you're a little intoxicated, @Jimmy?
 
@LightnessRacesinOrbit SCOTCH
 
user55340
Little?
 
nah he's always like this
 
> C# developer
never mind
I keep forgetting I left the land of sanity behind me in my SO cliques
 
user55340
And he claims understanding of Haskell and monads.
 
7:57 PM
I'm automatically suspicious of anyone who keeps prattling on about monads and monoids and Haskell and such things.
 
user55340
It's for the best.
 
>>= mempty :| scotch
 
user41796
@LightnessRacesinOrbit sanity is only for those who don't program on a regular basis
 
speaking of...
 
user41796
8:00 PM
Language choice merely affects the rate at which you approach insanity
 
protip, when wondering how your function works instead of trying to conceptualize your logic in your head, just look at the test you already have written that covers that exact scenario to see
 
I use Javascript and I'm still perfectly sane
 
user41796
@Ixrec And you'd be a liar.... :-D
 
user41796
You can have one or the other, but not both. They are mutually exclusive
 
@GlenH7 does javascript break mutually exclusive principles?
;)
 
8:02 PM
C++ is the one that drives me insane, because BUILD SYSTEMS ARE HORRIBLE
both of the actual languages are fine
 
@Ixrec our build system had cmake scripts calling shell scripts calling java executables WHATS NOT TO LOVE
 
@enderland our company's build systems might be slightly worse than yours
 
@enderland only if one of the principles is ''
 
do you have a file full of empty functions whose sole purpose is to "dummy out" symbols at link-time which your code's dependencies expect to see despite never actually calling?
 
interestingly we learned about this only because we tried a docker build without a JRE installed in our docker
 
8:04 PM
we have a Perl script whose sole purpose is to generate those dummy files
 
hey we had something like that too, actually. :|
 
eek
well...were half of your dummies Fortran symbols?
 
we had some perl black magic going on generating some sort of temp files
no
 
@Ixrec yes, if C++ didn't have a toolchain clearly built by someone following a recipe to summon Cthulhu, it would probably have a lot lower cost of entry
 
we had an amusing bug ticket the other day because someone managed to dummy out main() and broke everyone else's code
 
user41796
8:07 PM
main() is for suckers.
 
@Ixrec sounds like they had a deadline looming and decided to pull the fire alarm instead of getting marched into the principles office
 
hopefully
 
@Ixrec "both"?
 
@LightnessRacesinOrbit Javascript and C++
 
@JimmyHoffa "a"?
 
8:08 PM
my brain is mush right now
 
I love both C++ and Javascript, they do their respective ends of the spectrum quite well (when subsetted properly), the only real difference between them for me is that C++ always requires fighting with a build system of some kind
 
Every time I talk to someone who's worked with C++ about learning C++, they say "Don't learn C++." Everyone in The Lounge says that to me. Not sure if they are protecting my sanity or pulling my leg.
 
@enderland you been doing lots of coding lately? Gotten your wish? You seem to be in a constant state of transition, can't keep track of whether you're doing dev work, school work, or getting stuck with old legacy operational junk
 
@RobertHarvey probably both tbh, but as long as you're not tied to an evil legacy codebase C++ can be quite satisfying
 
@RobertHarvey haha "Lounge" and "sanity"
2
 
8:13 PM
the main thing to remember is that it is a language where you must read the entire manual before writing any code
or you will hurt yourself
 
@RobertHarvey also, enough with the "everybody"; these guys can't call you on it, but I know for a fact that's not true, so pull the other one!
 
@Ixrec but seriously, the manual is like 10k pages!
 
and by "manual" I mean a book like The C++ Programming Language and then Effective C++, one book won't cut it
 
@JimmyHoffa no, I'm tryign to wrap my mind around a complicated query schema right now. I'm not really doing much coding, almost all ETL work
 
user41796
< grabs popcorn />
 
8:14 PM
I've been FT now for 3 years or so, most of which I was doing coding...
 
and of course for modern C++ add one or two more books for all the new features
 
@Ixrec All 1000 pages of it? That's how long Bjarne's book is.
 
user55340
In college the lab I hung out in used Bjarne as a synonym of the headache you get from dealing with too much complexity.
 
@RobertHarvey you can probably skim some parts
 
I spend most of my life in C++. I'm really quite accustomed to it. It is fine for what it is, at the layer on which it sits. (Even if C++14 is a bag of shite and the future will be even worse.) But as a consequence of focusing on that gritty world, I have very little experience with higher-level stuff, Python, Ruby on Rails, Github, groovy new-fangled tools people talk about in opium dens like this one. My little dalliance with Docker this week has made me a little hungry for more of the poppy.
 
8:16 PM
I'd avoid skipping anything, except maybe advanced template metaprogramming (basic is enough to avoid hurting yourself)
 
psr
@Ixrec avoid hurting yourself?
 
@LightnessRacesinOrbit "Even if C++14 is a bag of shite" and "I have very little experience with higher-level stuff" are actually consistent viewpoints.
 
@psr to learn where the line is between useful templates and cthulu templates =)
 
user55340
@LightnessRacesinOrbit just stay away from php if you have common sense.
 
@RobertHarvey I use "higher-level stuff" not in the very narrow sense that the nonsense added to C++11 and C++14 might be described
 
8:17 PM
@LightnessRacesinOrbit outside of C++ circles there tends to be pretty common agreement that if it's really easy to screw something up, it should be fixed or avoided altogether, C++ folks lean towards the "just learn how to do it safely" instead approach
 
@LightnessRacesinOrbit dare I ask what's wrong with C++14? (I really like what's in there; it may have a lot of corner cases and inconsistencies but no more so than C++98)
 
C++ was always a language built on hacks, but it's really bad at that now. Really, really bad. The glamour of how you can compose metaprograms and impress your retired CS professor is not really relevant to that point.
 
user55340
Some php O_o : ideone.com/ppfv3f
 
@Ixrec anything useful is utterly illegible and cosmically esoteric
the language's rules for doing just the most basic things are absurdly complex
 
Wasn't that always true?
 
^
 
user55340
And Ruby and JavaScript.
 
@GlenH7 you sure saw it coming, I'm with you on this one :D
 
user55340
 
adding more things makes it worse, especially with all the TMP gubbins people are focusing on so much
I mean jesus christ look at the mess that is std::enable_if
Powered by SFINAE, a disgusting hack by any definition!
it's really extraordinary
 
8:20 PM
@LightnessRacesinOrbit Haskell doesn't have SFINAE! :D
 
alternatively, we could point out the fact that enable_if is less useful in C++14 than in C++98 because more of its use cases are covered by less arcane features now
 
@Ixrec such as
 
constexpr stuff and lambdas with auto parameters are the main things that come to mind
 
(I'm not disputing — I honestly haven't paid much attention to it)
hm I wouldn't say any of those take care of enable_if's use cases ... though I guess I can see how they might make those use cases obsolete
if you accept the distinction
 
I do =)
 
8:22 PM
constexpr seems like a bit of a mess too, though I won't profess to be an expert
 
pretty much every feature is a mess if you think about it too hard, though constexpr is actually quite devoid of gotchas
 
lambdas seem to be about the only new feature from 2011 and since, that isn't a total travesty
 
the worst you can do is fail to understand why the compiler thinks it can't be constexpr, and give up and let it go back to being runtime-only
 
right
don't get me started on std::move
inb4 no the single argument one
 
@LightnessRacesinOrbit why? did they forget std::stop and you'll just keep going until exhaustion?
 
8:25 PM
before we move on to std::move, one of the noteworthy differences between C++11 and C++14 is that C++14 allows a lot more stuff to go on inside of constexpr functions; previously you were limited to a single return statement
 
@JimmyHoffa well, for starters, it doesn't move anything
@JimmyHoffa it doesn't actually come remotely close to moving anything
@JimmyHoffa which seems strange for something called std::move
 
yes, you have to just know that std::move does not move
it was not a great choice of name
 
it's named after a use case, which is bafflingly stupid and entirely contrary to how the rest of the language is constructed
 
@LightnessRacesinOrbit semantic naming never was a strong point for low level languages
 
user55340
std::head:::case
 
8:26 PM
it's hardly the first time C++ has chosen a bad name for something standard (random: the one problem with constexpr is that's also a dumb name)
 
and yes I'm getting sick of the build process too
lol I write one python script and listen to me
 
to be fair every language has some dumb names in the standard
looking at you Haskell monad "return"
 
ooooh you said "standard" and "Haskell"
 
user55340
Retrum? Red rum?
 
7
Q: What does it mean to be "standardised", and why is Haskell it?

Lightness Races in OrbitWikipedia claims that Haskell is "standardised", but the Haskell standard is not ratified by an internationally recognised standards body such as ISO; not even by a national body such as BSI or ANSI. So what really are the criteria for a programming language to be labelled "standardised"? Does i...

AHA
 
8:28 PM
> God appears, and god is light
> To those poor souls who dwell in night
> But does a human form display,
> To those who dwell in realms of day
> -Blake
 
haha and someone suggested migration to Programmers #irony
 
as for move semantics, I believe they did quite well with that because although it is incredibly tricky and complicated, resource management always is, and fundamentally that whole topic is about ways to tell the compiler how to do a certain class of optimizations that are usually only necessary for classes that manage resources; if you don't manage resources you can virtually ignore move semantics and silently reap the benefits
 
@Ixrec I'm not disagreeing, I know why they used the name (it gives a pretty imperative appearance to do notation stuff), but yes I don't care for it
 
@Ixrec that was rather convenient yes
 
psr
@Ixrec I'm sure they just wanted the do notation to be readable and approachable to non-Haskell programming dolts. Normally they would just call it a Nytharlomorphism or something.
@JimmyHoffa Beat me to it.
 
8:32 PM
@psr I typically just make an alias for nytharlomorphism to return in all my Haskell so I can use the correct name.
 
user55340
@LightnessRacesinOrbit btw, our sanity quotient may be a bit... Questioned. Very high blue name ratio, Haskell, JavaScript and even mumps coders.
 
@MichaelT and the Javascript coder prefers it to all other languages =)
 
psr
@JimmyHoffa @JimmyHoffa - It just makes sense because only Nyarlathotep is pure.
 
@MichaelT I'm so sorry
OMG FEEDS SHUT UP
 
user55340
(JavaScripting here at the moment. Jquery Ajax calls to a Java backend)
 
user55340
8:34 PM
Use mobile. No feeds.
 
psr
@MichaelT I know, when they were talking about dumb names in the standard MUMPS has one by the time you read the title.
 
user41796
@MichaelT +1 for more jQuery
 
the thing in C++11/14 that I feel they did screw up was universal/braced initialization; it could have been objectively superior to parentheses and equals initialization, but nooooo they had to make auto x = {...} deduce to std::initializer_list for no reason whatsoever and now everyone's afraid of both styles
 
user55340
At some point I might d3 and add red circles.
 
> C++ developers finally learned to be afraid of their own language, news at 11!
;D
 
8:43 PM
@JimmyHoffa well, the good ones are already afraid of raw owning pointers
 
@MichaelT scotch are you scotch questioning my SCOTCH sanity scotch !?!
 
user55340
There is nothing to question.
 
@MichaelT alright scotch then! Carry on..
 
@Ixrec oh yes
@Jimmy wtf
 
user41796
@LightnessRacesinOrbit sed s/wtf/scotch/
 
user41796
8:48 PM
probably one of the best auto-replacements you could put in place and just leave there for eternity
 
@GlenH7 it would result in a lot of wtfs scotches per minute
 
user41796
1 min ago, by GlenH7
probably one of the best auto-replacements you could put in place and just leave there for eternity
 
We just had a phone interview with someone to get my job
 
user41796
@durron597 And they're awesome, right?
 
user55340
Those are always fun.
 
8:50 PM
my base salary is right in the middle of the posted range on the SO careers page
the guy is like "this is for a senior position, right? that salary is kind of low for senior developers"
let's just say it was difficult to not say cough no shit cough
 
user55340
Heh. Hint to boss then.
 
user55340
Ok, can't type this fast enough on mobile...
 
@LightnessRacesinOrbit scotch on you, you've heard me proclaim a file system monadic, spew various general nonsense and quote blake, yet you would still say scotch? I can't be blamed for you only now realizing scotch.
 
user55340
So, when I moved to Silly Valley back in '96, overheard some newer managers talking with their mentor about the developers they had and complaining that a few of them were getting paid more than they (the jr managers) were. The mentor said "good - try to make sure you keep them."
 
@durron597 "yeah, if I were you I'd apply to other companies, like Amazon it's a low cost of living here"
 
user114359
8:53 PM
A few jobs ago when I quit my boss didn't even bother to ask why. He already knew I was way underpaid and he already got blocked by his management when I asked him to fix it.
 
user55340
Was amusing looking at the shock on the jr. managers faces.
 
Honestly, I took a lower pay than I wanted because I was out of the industry for so long. And, to be honest, looking back, I wasn't worth what I thought I was, again for the same reason.
 
user55340
i think it's reasonable to say that most non tech places undervalue technology professionals.
 
(how many more times can I fit the word honest into one message?)
 
user41796
@durron597 honestly, we don't know. Honest-to-goodness.
 
8:56 PM
But now that my skills are much more current, I'm still getting paid almost the same amount. No wonder I'm leaving!
 
enderland is trying to stalk @durron597's company on careers
 
@GlenH7 That's an honest answer, honestly.
 
@durron597 yeah, one of the oddities in our industry is that pay increases aren't anywhere near at the rate commensurate people's value increase. As a decent engineer starts a company and begins learning the domain, after a couple years the value they're providing just from what they've learned of the domain is wayyyy more than earlier on.
 
user55340
This goes back to the most people want crap? No, they don't. They want what they think they can afford and what value they afford to technology professionals.
 
user114359
@durron597 You know what is really demoralizing? Being paid less than when you started in inflation-adjusted dollars but not being able to do anything about it because of the Great Recession.
 
8:57 PM
@JimmyHoffa I've wondered about this, if I leave my current company what I could get in terms of salary/etc...
 
@Snowman I'm kind of expecting to change jobs every 3-4 years for the rest of my life at this point. Unless Amazon figures it out.
 
It's funny that job hopping is a stigma, and yet doing it every 2 years will yield an enormous pay increase over time, compared to the piddling normal 1-5% yearly (5% is pretty damn uncommonly good) you get around industry
 
glassdoor suggests a slight average pay increase from what I'm making now, actually
 
user55340
@enderland set a value for yourself and interview to find your marketability. If it's above the number, take the job.
 
I tell this story all the time, my a family friend worked as an intern at... FNC I think, having a car pick them up at 6 AM every day for zero money for two years
 
user114359
8:59 PM
@durron597 I have had four jobs in two years
 
user114359
But I think I finally got it right, this new job is probably the one I will stick with for 4+ years.
 

« first day (1875 days earlier)      last day (3102 days later) »