« first day (1043 days earlier)      last day (3957 days later) » 

user20683
12:13 AM
for the most part I'm of the same mind
 
user20683
some of the object syntax is annoying
 
12:29 AM
yeah -- no language is perfect, I guess :)
though.... my experience of picking up other people's Python code has been a lot better than with other languages (PHP for example) in terms of readability
 
user20683
PHP is a bad example for readability.
 
user20683
Actually PHP is a bad example for everything.
 
user20683
I dislike the enforced whitespace now and then
 
at first I was the same, but it does make everything look tidy and avoids having 'End' style ending keywords
 
user20683
true
 
2:08 PM
hey guys, does anyone have a tutorial on using IWatch in linux or just watching a system bus in general? I'm trying to watch a CAN bus for changes, which, to my understanding looks like a file from within linux, so I need to be able to tell when that changes (aka has new data for me to process)
 
2:42 PM
@Ampt Here's a site for SWatch tutorials I think it's probably the same as IWatch
 
@JimmyHoffa I don't know what I expected.
 
I'm helping
 
I think we may disagree on the meaning of the word "Helping" hahaha
Thanks for the laugh anyway, I'm about to go into a sprint presentation and needed that
 
heh that's what 8-9 in the morning is for
 
Doing my nails?
 
2:46 PM
That's what you'r presenting the sprint right?
 
user41796
@Ampt "Helping" is most definitely a relative term with this crowd
 
@GlenH7 look, I missed an e on you'r up there. Isn't that just, like nails on a chalkboard? I think I'll leave it that way just for you. Get anywhere with that fusion log yesterday?
 
I feel like the Nail painting community could really benefit by estimating the time it takes to do different styles in story points and using agile processes to ensure nail painting occurs on time with low overhead
 
@Ampt Agile processes would ensure each nail didn't match, and 2 of 3 had two patterns because it was changed halfway through the nail
2 of 3? I guess the agile process made them lose a few too, probably YAGNI
 
@JimmyHoffa First patterns, now agile. Are you some kind of software hipster? "Agile process is too mainstream"
 
user41796
2:49 PM
@JimmyHoffa I'm still in dependency hell. We've got a 3rd party set of libs that rely upon F#. It appears that some of their libs require specific versions. Time to break out the dependency walker and start ID'ing things
 
@Ampt No, I was joking, and I wish agile was mainstream
@GlenH7 Ahh, have you used ILDASM in the past?
 
@JimmyHoffa I was gonna say, keep that talk up and you'll be banished to the realm of PSP/TSP
 
user41796
@JimmyHoffa ILDASM? is that a dependency walker?
 
Intermediate-Language Disassembler
 
user41796
ah, no I have not
 
2:50 PM
C:\program files\microsoft sdks\windows\v7.1a (or so)\bin\ildasm.exe
 
user41796
k
 
It let's you see all the methods and etc available in an assembly
 
user41796
Not sure I have the fortitude to walk assembly today. <heavy sigh> It's been a long week with a lot of bickering distracting me from coding
 
but the helpful part is there's a menu in it that will let you view the entire assembly's manifest, which includes the dependency reqs with the public key token of the specific version and the version that it depends on etc
It's not assembly, it's IL
Just look at an assembly's manifest through it, you'll see quickly it has lots of helpful information about the assembly including dependencies in a clear way
reflector will surely do the same thing as well if you can find a version on hand somewhere, reflector gives even more info actually spitting the IL back to C# if you never used it
redgate has recently been trying to monetize it and the free versions were hard to come by last I saw
 
user41796
Cool, thank you for all of the pointers
 
2:53 PM
Aye, hope it helps. If I didn't like sharing what I know, why would I be on SE? heh
 
user41796
And I realize it would be IL. It's just that IL deciphering is a tangent I wasn't planning on today or this week.
 
nah it's not IL
it tears the IL apart into human readable info
compilers create assembly, disassemblers create code from it remember :P
(or so has been my comprehension of disassemblers, I've never genuinely used one)
 
user41796
I just want it to work already. :-)
 
haha yeah. Good luck with that. This is why .net assemblies should always have the "SpecificVersion" flag on their refs set to false.
 
user41796
and we had battled that with this vendor before. So I'm miffed we're having to deal with it again. Unfortunately, I'm getting inconsistent behavior while trying different things. Which really just means I'm flailing away instead of sleuthing.
 
2:57 PM
@GlenH7 by chance, have you tried executing the process in a 32 bit context?
 
user41796
have to run this particular nasty in 64 bit since I need the memory space. It's ... ugly
 
I've had a variety of times where 3rd party libs didn't dictate but turned out they were compiled in 32 bit or depended on a 32 bit only lib, .NET just says can't find lib when you look for a 32 bit one
 
user41796
interesting; that may indeed be part of the challenge then
 
If at all possible just for testing purposes, you might try executing the process in 32 bit just to see if your dependency problems just disappear
 
user41796
would ildasm call that out?
 
2:59 PM
Yeah
I think it's at the bottom of the manifest it says x86 or x64 or any cpu
but you have to check all the refs to see that
could be 1 dll 3 nodes away that's 32 bit, causing the whole chain to fail when all nodes leading up to that one are any cpu
 
user41796
ok, looking at manifest now
 
I don't recall exactly how it's delineated, but I've used manifests to tell before and I recall it was in the last section at the end
 
user41796
er, manifest of assembly that the fusion log viewer called out as bedeviled
 
maybe ILONLY means any cpu
.corflags 0x00000003 // ILONLY 32BITREQUIRED
just pulled that from an assembly compiled requiring 32bit, so if it says ILONLY it's any cpu, (which is corflags of 1)
 
user41796
.imagebase 0x00400000
.file alignment 0x00000200
.stackreserve 0x00100000
.subsystem 0x0003 // WINDOWS_CUI
.corflags 0x00000001 // ILONLY
// Image base: 0x050F0000
 
user41796
3:04 PM
might have a smoking gun there then
 
that says any cpu, should be fine
you have to check it's dependencies though is the problem
if one of it's dependencies is 32 bit it may be causing the binding failure
all the .assembly extern listings at the top are it's dependencies
 
user41796
.assembly extern FSharp.Core
{
.publickeytoken = (B0 3F 5F 7F 11 D5 0A 3A ) // .?_....:
.ver 4:0:0:0
}
Appears to be spec'ing the 4.0 version?
 
Yep
Are you not running in 4.0?
 
user41796
backstory - I have a beta version of the libs since I needed a particular bug fix. We had either FSharp 2.0 or other version (not 4.0) in that same folder and ref'd as well. Dropped new libs in, other projects built fine but I haven't tried to execute them.
 
user41796
My particular project dies at runtime, not compile / linkage.
 
3:08 PM
Are you running in .NET 4?
 
user41796
sorry, yes, we're using .net 4.0
 
user41796
oh, and note that we have the fsharp assembly included in that 3rd party folder to "avoid" issues with folks having other F# versions installed. AKA one stop referencing.
 
Well not certain if FSharp perhaps has versions independent of the runtime like you mention you have a non-4.0 version in that folder, but that may be your problem right there
 
user41796
it does; I have 2.0, 1.9, 4.3, and 4.0 versions of fsharp.core living on my computer
 
all in the gac?
 
user41796
3:12 PM
no, not all in gac. Our dev machine configuration process is not as robust as I would like. Which is also why we use one stop referencing in this case. I'd have to x2 check the GACs version
 
Using the GAC stinks, and to be certain I'm not sure you can even load two versions of the FSharp.core into one appdomain (I really doubt you can)
GAC would allow finding both versions correctly if you wanted but again, I doubt even that would solve your problem. I think you just have an old fashioned version conflict. You need to get everything onto F# 4.0
Because you can't change the dependency of that 3rdParty bit
2
Q: Can't load FSharp.Core version 4.0.0 in infer.net fun // hard link dependency in a dll ?

nicolasI am trying Infer.Net (An F# Library for Probabilistic Programming) And running the examples in VS11 Beta lead to the error : Could not load file or assembly 'FSharp.Core, Version=4.0.0.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a' or one of its dependencies. The system cannot find...

look at the second answer to that question
something like that might work
 
user41796
Oy. Explicitly ref'ing the fsharp v4.0 assembly got me past the exception. Now to unwind this pile and present the results back to boss-man.
 
redirecting other versions
 
user41796
good link there; didn't know that was possible. We have a good relationship with the vendor, so I can point out the apparent dependency and ask them to fix it. Their turnaround can be pretty good on things like that
 
If it's an F# shop I would imagine they're doing things fairly right
 
user55340
3:16 PM
ORA 27102... out of memory. sad face.
 
user41796
They do some pretty interesting stuff. Heavy, heavy use of reflection for gluing stuff together.
 
@MichaelT Quick, plug some usb sticks in!
@GlenH7 Meta-Programmers, not uncommon among the FP crowd
 
user41796
@JimmyHoffa The founder of that firm is heavily biased that way. :-) We actually had to demonstrate to him some use cases where immutability was really bad for us.
 
Ah well, the alternative bias is way worse
 
user41796
True; and he was very reasonable about it once we walked him through the domain. The lights came on in his head and it was clearly an "ah-ha" moment. Good collaboration
 
user41796
3:20 PM
@MichaelT So you're running out to your local BBuy this AM? :-D
 
user55340
Nope... its a "hmm... oracle has funky memory requirements" and dusting off my sysadmin hat.
 
user41796
 
user41796
I have a new meme
 
user55340
> By default, Oracle 10 will allocate 40% of the total system RAM to the SGA and PGA.
 
@MichaelT ??? shit, everybody get down, he's about to get the perl!
 
user55340
3:23 PM
perl? Nope... its a "what does projadd do?"
 
@MichaelT SGA/PGA? And 40% sounds like a horribly low default for a DB
 
user41796
I was kind of surprised it was that low too. My biased assumption is that Oracle assumes it's the only thing on the box and all your memory belongs to them.
 
well needs to leave for OS, I would have expected like 80%
 
user41796
If they could pull off something like OS/400 then they could rightfully claim everything.
 
user41796
Such a mind bending concept when the OS is a database and the database is the OS.
 
user55340
3:31 PM
Just checked with the sysadmin... yep, I can adjust the memory on the machine (he's primarily network/linux/postgres and currently working on repainting some of the walls).
 
user41796
@MichaelT the sheer absurdity of that last statement is awesome. "Our regular sysad is too busy painting so he can't fix the server." Makes me kind of curious what he's painting the walls with
 
user55340
Its more of a "I'm the one charged with working on this" having a fair bit of solaris and oracle admin experience back (~5 years ago) I'm not too far out of my league. And we're redoing the "dev cave" for which there's a timeline (redoing the cube walls this weekend).
 
user55340
Next monday we'll be out of the cave (further into the depths of the building... or outside). I've suggested reflectors on the laptops on the deck...
 
user55340
 
user55340
Granted, none of us are built quite that way... but we can get rid of our pasty white skin (northern wisconsin / minnesota isn't known for tanned bodies)
 
user41796
3:38 PM
@MichaelT Solid suggestion, but you need to go for the whole body tan. Just saying
 
user41796
I lived in Milwaukee for 4 years. It was really, really, really hard to get a suntan while there.
 
user55340
If you wanted to get a full tan, should have headed up to my old town... Mazomanie. Look it up.
 
user55340
Beach on the Wisconsin river...
 
user41796
Didn't have a car at the time, so didn't venture out from downtown all that much. I kind of wish I had checked out more of the activities up in WI but I also had a lot going on at the time too. Such is life.
 
How about the far more useful, "Programming kids" canonical post? I could use that one.. — Jimmy Hoffa 2 mins ago
I can't wait for the release of Child#... SayPolitely(StringBuilder statement) { return statement.Append(", thank you please."); }
 
user41796
3:50 PM
@JimmyHoffa unfortunately, that doesn't start showing up until after v20.
 
@GlenH7 ....then I suggest we skip some versions, get marketting involved and we'll start incrementing versions with fibbonacci
 
user41796
18 more close votes and I'll be tied with gnat. Woo hoo!
 
user55340
I broke the 1k close reviews recently for that badge...
 
user55340
4:08 PM
@GlenH7 you need to try to catch up to him on SO stackoverflow.com/review/close/stats
 
user55340
(and SO will likely break the 60k review queue today)
 
user41796
I generally don't do much on SO. The churn is just too fast there. I have answered a few questions, and I'll flag a few items. But I focus most of my effort on Programmers. Seems more rewarding / clearer feedback from my contributions
 
@GlenH7 Agreed. SO just doesn't have the fake-reward scenario I like
 
user55340
Same here... though I've been tempted to try to get to 3k rep there so I can help with the close queue.
 
user41796
Maybe I should build up my rep there through edits. I think the cap on rep from edits is 2k.....
 
user55340
4:11 PM
I've got more MSO rep than SO rep.
 
user41796
I think I've got 15 more SO than MSO at the moment. MSO scares me. Almost as much churn as SO as well as the comments and votes being thrown with wild abandon.
 
Yeah, I try to avoid MSO more than SO heh, it feels like usenet and I haven't stalked it long enough to join
 
user55340
535 rep there... a chunk of it from a migrated M.P.SE question.
 
user41796
Usenet had a degree of latency to it, since most folk at the time weren't always staring at their inbox. And SMTP servers were a bit slower to cycle through their queues. MSO is instantaneous .... [insert description here]
 
user41796
quality of content is often quite similar.
 
user41796
4:24 PM
The more I think about it, hitting 2k rep on SO through edits is a screwy enough idea that it sounds interesting enough to try.
 
user41796
Probably more productive than looking at the leaderboard and seeing which user is next for me to overtake in rep. Current bullseye is on Mike Brown with a mere 10 points left to match him. :-D
 
user55340
Heh... another messed up audit...
 
user55340
"Should this question be closed as: primarily opinion-based?" --- At what size of data does it become beneficial to move from SQL to NoSQL? [closed]
 
user41796
Yeah, I can't say I'm impressed with the current selection and suggested reasons behind the audits.
 
user55340
I had one audit that asked about the old close reasons.
 
user41796
4:31 PM
And it's counter-intuitive to say "leave open" on a good question that has been marked as on-hold for the purpose of the audit. I can't leave it open; it's already been closed (per the audit)
 
user55340
I'm still frustrated with the expected leave open on highly voted opinion polling that got asked on the weekend with no one casting a close yet on it...
 
user41796
I've hit those too. After I fail the audit, I turn around, hunt down the question, and then put a close vote on it. :-)
 
user41796
I happy to fix that problem, thank you very much
 
user55340
As do I... so no one else gets that as a review audit...
 
5:42 PM
That's why I'm over here... SO is such a large beast
Guaranteeing that no one has ever asked your question is a little harder
 
6:15 PM
Modifying code that was written 4 years ago to the day
clean, readable and easy to work with
this guy deserves a medal. He's not even with this company any more
 
6:37 PM
Would it be weird to endorse him on Linked in for that?
 
6:51 PM
@Ampt Nah, I think that's the perfect example of when to endorse someone; if they wrote code that 4 years later stands up as readable clean and easy to modify, that's one hell of an engineer in my book.
 
@JimmyHoffa that was my thought process. The code was clean, commented just the right amount, and each piece was nice and small. It was really like a five star, 9 course meal
@JimmyHoffa yeah, and it was flippin delicious
 
user20683
@JimmyHoffa Today's XKCD
 
user55340
7:11 PM
@WorldEngineer ppst... xkcd is a whitelisted site.
 
user55340
 
user20683
good to know
 
user55340
Complete with mouseover.
 
user20683
@MichaelT the QR code works too
 
What's it do?
 
user55340
7:13 PM
HA!
 
user55340
Its another joke.
 
What is it though?
 
user55340
Its a URL to xkcd.com/1237/scan
 
user55340
(apparently, @JimmyHoffa got stuck...)
 
user20683
@MichaelT
 
user20683
7:20 PM
 
@WorldEngineer I what?
That's kind of creepy and sad and confusing.
 
user20683
@ThomasOwens cat is getting a bath
 
@MichaelT Nah, I just didn't want to install a QR code scanner on my phone off hand
 
user55340
There will be a side effect there of scratched and bloodied legs. Its not a pure thingy.
 
@MichaelT It's fine, it'll stay in the IO monad
the pure environment outside will stay kleenean
@WorldEngineer Where do you even happen upon an image like that?
 
user55340
7:23 PM
But kleen is a star, not a cat.
 
user20683
@JimmyHoffa Haskell Tutorial
 
user55340
 
user20683
@MichaelT We had that in the store for a while
 
@MichaelT Have you scrolled down to read the comic? I'm trying but uhh... what??
I'm just imagining some terrible programmer pun science fiction novel called "They came from planet Lisp"
 
user55340
7:35 PM
@JimmyHoffa Yes, I did scroll down... and... yes...
 
Heh I like the ruby gun, that's about right
 
user55340
Ruby doesn't need any help to get to that point.
 
user20683
gun.(shoot) unless surrender
 
Sounds like something the french would write
 
user20683
I've been cramming Ruby into my brain over the past few days, it's not so bad.
 
user20683
7:39 PM
@Ampt The French only surrendered like that once.
 
Yeah, I've got the Prag. Prog. Ruby 1.9 book and it's pretty nice
 
user55340
@WorldEngineer Can you explain why x?1:2 is a parse error?
 
@MichaelT I'm going out on a limb and going to say you need a space after x?
which is dumb
 
user20683
The colon
 
user20683
but I'm not quite sure why
 
user55340
7:41 PM
@JimmyHoffa Yep. "foo?" is the convention (only a convention!) for a method that returns a t/f value.
 
user55340
So it parses it as (x?)1:2 as "function x?" which doesn't exist and then breaks.
 
never heard of restarts in LISP before, that's pretty cool
(It's a monoid, in case you were wondering too... just saying)
 
user55340
I also "love" the difference between has_key? and include? and member? and key? in Hash.
 
user20683
unless still weirds me out a tad
 
user55340
(but then, I can rant about ruby for a bit)
 
7:47 PM
unless - for when you're too lazy to change the order of your code
 
until - for when you don't want to know for certain what you're code will do
 
user20683
@JimmyHoffa isn't until == "do while"?
 
user55340
unless cond is the same as if not cond
 
user20683
that's why I thought
 
yeah, which is the same as else. It basically allows you to reverse your if and else statements
 
user20683
7:53 PM
and yet the style guide says never to use them
 
if -> A
else -> B

is the same as

unless -> B
else -> A
 
user20683
unless it's a one liner
 
@WorldEngineer Is it? :)
 
user20683
@JimmyHoffa language dependent
 
user55340
The thing is, ruby pulled it from perl... and it was put in/used as part of the STMT unless (cond); and STMT if (cond); style... for orthogonality, it was added for the same way that if works.
 
user20683
7:54 PM
as anything else
 
Think about it, english is less specific than programming needs to be
 
user55340
Its rarely seen in the perl community in the unless (COND) block; form.
 
until doesn't work so well
 
user20683
until this happens do something
 
user55340
perl:
# define an array
@colors = qw(blue green brown cyan white);
print "@colors\n" unless scalar(@colors) < 5;
# prints: blue green brown cyan white
 
7:56 PM
do something until FF
does it do something for FF also?
 
user20683
until (I_have_a_real_job) do job.hunt()
 
user55340
24
Q: Usage of ‘if’ versus ‘unless’ for Perl conditionals

JSBձոգչWhat are some guidelines for the best use of if versus unless in Perl code? Are there strong reasons to prefer one or the other in some situations?

 
user55340
28
A: Usage of ‘if’ versus ‘unless’ for Perl conditionals

friedoIn Perl Best Practices, the advice is to never use unless. Personally, I think that's lunacy. I use unless whenever there's a simple condition that I would otherwise write as if( ! ... ). I find the unless version to be more readable, especially when used as a postfix: do_something() unless $sh...

 
user20683
@JimmyHoffa do while will execute for all the things
 
user55340
while (<$fh>) {
    next unless /\S/;
    # ...
}
 
user55340
7:57 PM
Thats what its "intended" for.
 
@WorldEngineer I know what do while does, but semantically until is a poor choice, also until is do while not
or while not
unclear which
@WorldEngineer when do you choose a do while vs. a while?
 
user20683
@JimmyHoffa I don't, Python doesn't have do while
 
heh
 
user55340
The python philos is "one obvious way only"
 
user20683
you'd choose do while when you want to make sure it executes at least once
 
8:01 PM
Good, honestly there's no reason to have both genuinely, do while makes one uncommon scenario a little cleaner but really it's meh
 
user20683
basically python has while to represent conditional loops and for to do iteration
 
you mean like for-in ?
 
user20683
yeah
 
user55340
 
user20683
for beers in Jimmy_Hoffa:
drunk = drunk + 1
 
8:05 PM
Python has no ++?
 
user20683
@JimmyHoffa there is but one way to add two numbers
 
user20683
it has += IIRC
 
user20683
yeah
 
Increment is not the same as addition, it's an atomic single variable modification; no other variables are involved
 
user20683
@JimmyHoffa true
 
8:07 PM
addition is a binary operation, increment is a unary operation, they're different enough
 
user20683
yeah, increment is also unstable at the best of times
 
user20683
bear in mind that the default implementation of Python is in C
 
user20683
and thus is subject to some of its peculiarities.
 
user20683
9:46 PM
@manishearth, blog cleared, publication at 22:00 GMT
 
user41796
@WorldEngineer but we want it NOW!
 
user20683
Now = time.time() + 5
 
Whatever pushes that interviewing trash off the blog's frontpage is good in my book. Morons is speaking about how to get past the most wrong-headed hiring processes it seems, it all mostly boils down to "Be a type A and anyone will hire you!" Suggesting the phone screen only serves the purpose of "getting to know the hiring manager" betrays the fact that he has no sense of fizzbuzz pre-testing which is a standard, and the majority of phone screens act as technical screeners more than anything
 
user20683
@JimmyHoffa bear in mind he's a consultant
 
user20683
and he's mostly talking about consulting
 
9:57 PM
he's a salesmen it sounds like
 
user20683
I think he does a fair bit of database work
 
Though I'll bite, there's much more sales involved in being a consultant
those posts should have maybe said that though
been titled somethign about getting consulting jobs
I've seen people turned down in interviews for being all type-A salesmen, because a real engineer was doing the interviewing to whom big talk is a huge turn off
 
user20683
@JimmyHoffa true enough
 
user20683
Feel free to write a refutation
 
user20683
:)
 
10:00 PM
pfah, I'm so lazy I might already know perl
@WorldEngineer did you sign up for the haskell center beta?
 
user20683
@JimmyHoffa not as yet
 
user20683
or I might have
 
user20683
I don't recall
 
user55340
0
A: Why don't we just set up a robot to automagically add 5 close votes to every question over the course of, say, 5 hours?

MichaelTOne of the hardest things to realize about Stack Exchange is that we are contributors to a site, but we are not the revenue stream for the business model. That lines from the next person to visit the site. Much of the site is designed around this - optimizing the search engine and being the pla...

 
10:25 PM
Manishearth on July 12, 2013

So you’ve been using Stack Exchange for a while now and, like most other programmers, expect to be able to customize the website. I’ve seen many feature requests on the main meta, and a large fraction of them never get implemented because:

It’s not a feature that everyone will like — Stack Exchange tries not to provide too many customization options, so features that will be annoying to some won’t be implemented

It’s a rather minor feature that may be only useful to some people (or will be used rarely), and not worth the developer time …

 
 
1 hour later…
user41796
11:54 PM
Would someone double check my math, please? I added up 18090 close votes off the top 20 reviewers which evaluated to 88% of the 20724 votes recorded. Mostly making sure I approached the summarization correctly.
 

« first day (1043 days earlier)      last day (3957 days later) »