The Whiteboard

user55340

00:14

@JimmyHoffa aww... Yannis locked this one...

user55340

12

At my first workplace we were using Digital Standard MUMPS on a PDP 11-clone (TPA 440), then we've switched to Micronetics Standard MUMPS running on a Hewlett-Packard machine, HP-UX 9, around early 90's. Is still MUMPS alive? Are there anyone using it? If yes, please write some words about it: a...

gnat

11:17

@MichaelT well not quite "Yannis", he's been acting on a flag :)

Jul 16 at 14:43, by gnat

I think the only safe way for them is to keep unilateral / not clearly motivated / documented mod actions at minimum, basically mostly act on flags (btw that's why I don't want to be a mod myself)

Thomas Owens

You'd be surprised at how many times mods have asked for the ability to "act like a normal user". Like cast a non-unilateral close or delete vote.

Including myself.

user41796

13:17

@Macuser #include <string> incorporates the string.h header file and associated routines into that module. std::string indicates the namespace that you're using for the class / object.

user20683

14:01

0

Q: Where can I find TDD Red, Green, Refactor Reminders?

My office would like to get reminders for programmers to help they stay true to "Red, Green, Refactor". There is a famous hat which does this, here: http://blog.gdinwiddie.com/2012/12/26/tdd-hat/ However, we are interested in an object we can put on our desks. Do any such devices exist? Wher...

tdd

user20683

O_o

Thomas Owens

I've heard of andons and other visual things, but what the hell?

user55340

You need to take action when people do things in TDD - papercut.com/blog/chris/2011/08/19/who-broke-the-build

Thomas Owens

Andon (manufacturing)

Andon (アンドン, あんどん, 行灯) is a manufacturing term referring to a system to notify management, maintenance, and other workers of a quality or process problem. The centrepiece is a signboard incorporating signal lights to indicate which workstation has the problem. The alert can be activated manually by a worker using a pullcord or button, or may be activated automatically by the production equipment itself. The system may include a means to stop production so the issue can be corrected. Some modern alert systems incorporate audio alarms, text, or other displays. An Andon system is one of ...

user55340

►

Thomas Owens

14:06

No YouTube at work. :(

user55340

> Retaliation is a Jenkins CI build monitor that automatically coordinates a foam missile counter-attack against the developer who breaks the build. It does this by playing a pre-programmed control sequence to a USB Foam Missile Launcher to target the offending code monkey. Check out the video to see Tom take one in the back of the head all because of a missing semicolon!

user55340

> Honestly, would you work in a dev team with a lava lamp build notifier? What next? Nyan Cat mouse mats? Real coders work under the threat of Retaliation! -- Matt, Coding Machine

user55340

An aside... I got pissed off at SlashDot last night. One article was about people trying to overthrow copyright in Finland and people were "woo! down with copyright"... the next article was about a company that used 3d printed CC-NonCommerical designs to advertise their commercial 3d printer and people were "bad! must follow license!" -- don't they realize these are the same issue?

user55340

For ref - yro.slashdot.org/story/13/07/23/0013230/… and hardware.slashdot.org/story/13/07/23/0115242/…

user41796

@WorldEngineer Erm, and you haven't closed that already why? :-D

user41796

14:22

@MichaelT I work with too many engineers (myself included). We'll have changed out the recoil spring on that one and amp up the rate of fire. It would seriously hurt by the time we were done with that "build monitor." Fortunately, I'm in a high walled cube, so I'd be safe from that sort of system. Then again, we'd probably just put it on a track hanging from the ceiling to make sure everyone could be shot.

user55340

I used to work in a vaulted ceiling pole barn. We had a 10' divider in the middle. Placing it on top of there could hit anyone easily. Where I currently am, the cube walls are an 'E Ⅎ' with two of us in each open spot. There are many options for coverage.

user55340

(that said, I'm still working on getting CI to be part of the workflow here)

user41796

@MichaelT I was happy enough with nightly builds. Our code base is so dang large that "continuous" is a bit of a misnomer.

user55340

We could do it... its a matter of the time and willingness. The ant scripts are a bit convoluted. This could be a problem (I'm teaching myself gradle)

user41796

my current challenge is to coffee or not to coffee. :-D

Jimmy Hoffa

14:35

@MichaelT I'm on the fence about this, maybe if everyone on a team had one pointed at them, and all of them fired when that teams portion broke; but I'm not a fan of singling someone out in any way, it seems fun at first; but it can become in a bad way when the culture is built around the concept and management starts twisting it.

Or rather, singling someone out for doing things poorly is fine I guess, it's punishing them for it that I think can become terrible

Single them out; for help, not for punishment. Sure quality issues have to be handled at one point if continued, but you're just wasting money if you choose that option without first trying to improve things

@GlenH7 Engineering! Work out the trajectory and power required to accurately arc over walls.

user41796

@JimmyHoffa that's ballistics, not engineering. ;-)

Jimmy Hoffa

@GlenH7 Engineers should be able to calculate any of these things. How else will you know where the bridge support cables will land when they snap? Gotta make sure they aren't going to hit a building or housing development

Engineers calculate trajectories

user55340

@JimmyHoffa Its not punishment... any more so than email is. It is notification sooner than later. Emails aren't intrusive enough for bad code in the build. Something is broken. Fix it now. It costs more to fix it later.

user41796

nah, just make sure they never snap. :-) That's actually a big concern for military air bases for what should be obvious reasons. They really, really don't like any sort of development creeping up near their runways

Jimmy Hoffa

@MichaelT Yeah, I suppose, I still think it could in time be twisted horribly

user55340

14:43

Software testing

Software testing is an investigation conducted to provide stakeholders with information about the quality of the product or service under test. Software testing can also provide an objective, independent view of the software to allow the business to appreciate and understand the risks of software implementation. Test techniques include, but are not limited to, the process of executing a program or application with the intent of finding software bugs (errors or other defects). Software testing can be stated as the process of validating and verifying that a computer program/application/prod...

user55340

(The economics section)

user55340

If you introduce a bug in the mainline and leave (didn't check your email before you left), and I pull the code and trying to do anything with it, my code breaks. I've got to debug it - and find that you checked in code that broke the build. That takes much longer than if you check it in and realize the error within a few minutes.

Jimmy Hoffa

@MichaelT Oh I did builds for years, I know full well and agree completely on those points

user55340

Or even if I don't pull it, and the next day development continues... and then you have to spend much longer to find and fix the bug than if you had fixed it the previous day.

Jimmy Hoffa

I'm all for quality CI with a quick feedback loop for the engineers to know

user55340

14:46

Don't look at it as punishment. It is a push notification with some actual force behind it.

Jimmy Hoffa

Maybe it's the public function of it I don't like

user55340

(Jimmy is in favor of electric buzzers hidden in the chairs... bzzt)

user41796

I'm surprised that with a 300 point bounty, this question still has low views.

user41796

4

Q: Multiple scrum teams moving to single backlog

We currently have 5 scrum teams that work off their own product backlog for the past year. Each team works on their own dedicated system but underlying technology is the same .Net. There has been a lot of discussion on moving to feature based teams working off a single backlog. The reason is on...

agile scrum

Jimmy Hoffa

Works for me! :) I just think singling a dev out in front of the company as screwing up would lead to an adversarial culture

user41796

14:49

Oddly enough, it answers my musings on whether a larger bounty would have attracted more attention to my earlier question. Short version: apparently not.

Jimmy Hoffa

@GlenH7 I think some of us are just srumlogged at this point from all the scrum scrum scrum, scrum this scrum that backlog product owner and scrum that you hear about around the internet so constantly

user41796

I'm going to coin the term SDD right here and now. :-P

Jimmy Hoffa

@GlenH7 I'm going to coin the term Kicking Glen In The Shins Driven Development

user41796

I thought the rules were that we had to stick to single letters. :-( KGiTSDD is catchy but violates that rule.

user55340

@GlenH7 When you've got 43k rep, a 300 bounty is about the same as a 100 bounty on 10k rep.

Jimmy Hoffa

14:52

I'm more than happy to work in scrum, I like it, I just am kind of tired of reading questions that babble on in scrumlish using the unique scrum vocabulary

user41796

@MichaelT Yeah, I like how Karl will just toss about large value bounties for giggles. He's done it a few times before.

user55340

Delete vote increments max out at 35k (and I don't recall him being that active in delete votes)

Jimmy Hoffa

@MichaelT this is nuts, I never realized how much rep you can get just by being timely on a given piece of info, I've never had this happen before but I get like 30+ rep every day since I wrote this, I think I've never answered a question that got me significant continuous rep more than a day after I wrote it

22

Q: Where does paypal's 92233720368547800 number come from?

There has been a story in the news about a man whose Paypal account was accidentally credited with $92,233,720,368,547,800. Where does this number come from? What sort of programming bug is likely to give rise to this number?

errors numbers payment

I've been doing it the stupid way, just trying to write quality answers where I can, apparently I should be hunting down zeitgeist questions

user41796

@MichaelT I don't think he's all that active in close votes either. But he's got 700+ answers to his name

Jimmy Hoffa

@YannisRizos - I was just moving my puny little self out of the way of any collateral damage from the diamond hammer coming down. ;-) — GlenH7 1 hour ago

@GlenH7 thanks, now I need to hear the song and I'm at work with no access to music other than that I can buy online and the beatles can't be bought online. mumble mumble

it's going to be stuck in my head all day

user41796

15:01

►

user41796

disclosure: I haven't listened all the way through on that one.

user41796

And try to tell me it's not an apt analogy in this case... :-)

Jimmy Hoffa

15:15

@GlenH7 Quite apt.

Jimmy Hoffa

15:32

...now back to the aqualung album I had to put on my phone after talking about it the other day

user55340

I was reading somewhere that any time you have an angry employee that means they have a way of doing something that is better than the way it's being done now, so if you have 100 angry employees then you have 100 ideas for improving your business. If you have just 10 angry programmers then you probably have more than 100 ideas. — jzx 42 mins ago

Jimmy Hoffa

@MichaelT I had to remove the star when I noticed the abuse of 'then'

user55340

There sometimes things you can do there... and its by design to avoid possible misinterpretation of strings.

user55340

Nope, can't html tag it here. You can in actual posts (though I don't think comments will do).

Jimmy Hoffa

Oh I just learned something new; I instinctively hit ctrl+up arrow to bring up the same thing I just typed, some muscle memory from another app; I'm not even sure where I know that from. but it brings up your last message for edit

It's how I navigate some things in emacs, that's gotta be where I got it

oh shit, you know what that means???

P.SE chat is just an emacs extension.

user55340

15:38

^ESCks/Neat/Cool/

user55340

rats. No vi mode for chat.

Thomas Owens

Don't need ctrl.

Just press up.

Jimmy Hoffa

@ThomasOwens Don't you have like a project to manage or something? Nobody like's a wiseass, you know.

Thomas Owens

@JimmyHoffa I'm on lunch. Then I get to go play "IT, give us things we want"

Jimmy Hoffa

@ThomasOwens Oof, that game sucks, remind them to pull their pony tails back before the meeting so they can hear you

user55340

15:41

@ThomasOwens Blame wheel - "They didn't give IT what we want"?

Jimmy Hoffa

@MichaelT No it's more like "The CI processes broke it!"

→ 1 message moved to Hide! Hide!

Jimmy Hoffa

16:20

Is it that wrong to like letting exceptions just...happen, I mean, if I need to do something, what's wrong with just doing it, and if an error occurs then it bubbles up, instead of checking if I can do it before hand, at which point throwing an exception anyways

I guess there's cause for claiming it's a performance improvement to check ahead of time before you go spin a big yarn and run into the error at a later point

but then is that guard premature optimization; it's not difficult to retrofit guards

My philosophy on exception handling and data validation is a bit controversial I suppose, if I wrote a language; I would make it impossible to throw an exception in a catch block, I think that is total bullcrap, everyone does it and it drives me nuts, catching exceptions at every layer rethrowing them so they get caught and rethrown a layer above... just let the damn exception happen.. oh well, I don't suspect to win anyone over on that, everyone believes so much in guard statements because

they're pretty and easy to unit test

nevermind all the constant duplication, but oh wait, we can add another technology to solve that by using AOP, which is just turning what left alone would be no problem into 2 problems: proliferation and magic

user41796

16:50

@JimmyHoffa I could see an argument in favor of a re-throw if you're bundling additional diagnostic info within the exception or if you're forcing the logging of additional details at that point. It's a pretty weak justification though

user41796

I think some of the existing try / catch semantics came out of early Java exception handling. You'd get "nasty" warnings about such and such method will throw an unhandled such and such. So coding conventions were created to make the warnings go away but not actually deal with the exception itself.

Jimmy Hoffa

@GlenH7 Yeah, I think you hit the nail on the head there, how to actually organize and design your exception handling approach has been kicked down the road since ever because for some reason, the whole concept of it illicits all kinds of fear out of people

They think unthinking-blanket-paranoia is the appropriate approach to exception handling, and so they either avoid dealing with the subject or they just decide they should do everything ever at all, it could be due to it's relation to security which is an even more avoided topic by most devs due to the fear factor

@GlenH7 This is everyone's response to my position on it, but I just...don't see it, I've nearly never seen valuable extra information added for the logging purposes that couldn't be handled at the top layer where you should be catching these types of errors anyway. The top level knows what interaction caused this, logging a ton more than what happened, where, and what the top level interaction was is how people end up with these horrible useless logs that actually serve only political CYA

the typical barf-logs people end up creating of everything serve next to zero value in diagnosing problems half the time

@GlenH7 thecodelesscode.com/case/73?topic=logging

user41796

17:10

@JimmyHoffa it can be difficult to balance, no doubt. I've seen the lower level additions helpful for tracing type scenarios that we just couldn't catch in the debugger. But your broader point is sound - whoever has the most amount of contextual information should handling the logging or error correction.

Jimmy Hoffa

@GlenH7 My point isn't he with the most context should handle it, my point is he who lives atop the stack (unless a lower level can recover from the error) should be the only one to do anything about it. When I see try { } catch { throw new someException(); } at every level of the stack this isn't code duplication anymore, this is code putrification...

user41796

@JimmyHoffa well, then I disagree with you. :-) it's all about context, imo. The level with the most context has the greatest chance of being able to correct the issue. Otherwise object responsibility starts to bleed all over the place. And yes, I'm making an explicit correlation between available context and likelihood of responsibility. Top level handling can cause an object without knowledge or responsibility for the problem to try and resolve the problem.

psr

The argument also might be that the errors you throw are effectively part of your API, since code calling you can see it. But this really only matters if the code calling you will do something different with a MyWidgetException vs. a null pointer exception, which is quite a dubious assumption in practice.

user41796

said another way, if a class makes a mess while providing a service to another class, then it's the first class's responsibility to clean up the mess it made. Making the 2nd class clean up passes responsibility away from where I think it belongs.

user41796

@psr I think that's a great corollary consideration. Taken to an extreme, all class methods are an API. What do you want to foist upon your poor callers when things go badly? And we know things will go badly at some point.

Onwuka

17:22

In the US, is all software code owned by USDA in the public domain? USDA isn't federal government, is it?

user41796

@Onwuka No; it depends. Much work & research provided by the USDA, DoD, NIH, Dept of Energy, etc... are public domain. But they do have individual projects that they may choose not to release.

user41796

The US gov't is in essence funding basic research in many of these cases, and they are a very significant supporter of developing the world's intellectual capital.

Onwuka

:10478932 yeah their website says "Some materials on the USDA Web site are protected by copyright, trademark, or patent, and/or are provided for personal use only."

user41796

The big question is what your target usage is

Onwuka

@GlenH7 will they release source code for internal applications? it is not like it will be useful for anyone else...

user41796

17:27

@Onwuka generally no. The cynic in me says that's because there isn't a form available to make that request with. In theory, you could try a freedom of information request but that means you need to be a US citizen and you need to be in a jurisdiction where you would have a reasonable claim to the information. I'd say you're still at no.

user41796

A lot of the internal apps have proprietary and / or sensitive information wrapped up in them. So it's against the public interest to release that information.

Onwuka

@GlenH7 thank you :) I don't really want it. I am just curious.

user41796

They may also be bound by contracts signed with suppliers. Some vendors provide proprietary frameworks. The contract will keep you out of that code too.

user41796

Protip 1: Stock silverlight charts do not respond well to having ~7k points thrown into them.
Protip 2: Performance tweaked silverlight charts do not respond well to having ~7k points thrown into them.
Protip 3: Sometimes you really don't need to prove what you think will be obvious.

psr

17:57

@Onwuka I know the Veterans Administration hospital systems have been successfully requested under the Freedom of Information Act, and all the source code is kept public except for the security and billing stuff. But I'm sure the USDA doesn't own every bit of software it uses.

Thomas Owens

18:20

@psr If you're referring to VistA, it's...eh.

Yeah, you can get it via FOIA, look at it, and even use it. But...would anyone really want to is the question.

Onwuka

Um, let's pretend I didn't say that

user20683

@Onwuka Said what?

user20683

:)

Thomas Owens

Wasn't it a command-line app at one point?

psr

@ThomasOwens I am, and it is.

Thomas Owens

18:22

I'm pretty sure it was.

The government doesn't like to throw things away.

Onwuka

@WorldEngineer thank you for that

psr

@ThomasOwens I think EPIC, which is growing pretty fast, used a lot of that code to get themselves started.

Thomas Owens

I applied for an internship/co-op at EPIC. Also at SPARTA.

psr

I'm not sure what the EPIC software is written in today. Do you?

Thomas Owens

That was like 4 years ago, and I never got an interview. I think they were looking for Java and web experience.

user20683

18:25

@psr from what their website said, Python and Java but I can't recall.

user20683

I can ask Badcock about it.

user20683

he had an interview with them last year

user55340

@psr IIRC, its MUMPS too.

user55340

Epic Systems

Epic is a privately held health care software company founded in 1979 by Judith R. Faulkner. Originally headquartered in Madison, Wisconsin, Epic moved to nearby Verona, Wisconsin in 2005. Epic's market focus is large health care organizations. Epic offers an integrated suite of health care software centered on a MUMPS database. Their applications support functions related to patient care, including registration and scheduling; clinical systems for doctors, nurses, emergency personnel, and other care providers; systems for lab technicians, pharmacists, and radiologists; and billing syste...

user55340

Epic has a very young culture to it... I talked to some friends from college when I got back from my layoff and they warned me about it. They also warmed me that it was a "get them while they are young, burn them out in 5 years, and let them go"

user41796

18:36

@ThomasOwens USDA has a lot of Java code, afaik. I know a few folk who have been consultants there.

user55340

indeed.com/forum/cmp/Epic-Systems/…

user55340

> Friends don't let friends work for Epic

user55340

> It is commonly known that Epic doesn't hire you if you're "too old", judging by Judy Faulkner's comments that "it takes 3 years to turn someone into an Epic Person and it doesn't work on older workers"...or one ones with experience or maturity.

user41796

@MichaelT Sounds like Anderson Consulting (now Accenture) back in the day. They liked new hires straight out of college. In all fairness, they did pay OT. But their rates were a bit below market.

Onwuka

@MichaelT I interviewed with Epic. Someone else made an offer while they dragged their feet. I guess I dodged a bullet.

user55340

18:39

@GlenH7 Its one thing for a consultant shop to have bodies to throw at a problem... Taos (where I worked out of college was much the same). Its another thing for a company to try for a median age in the mid 20s.

user20683

I debated interviewing with Epic

user55340

> If you're 30 yo+ or worked for another company for more than a couple of years, I wouldn't bother applying. The odds are stacked against you for those 2 reasons. Now and then I'd see a new hire not directly out of college but it was rare. Basically what I'm saying is if you fit this criteria, I'd color your hair to hide any gray and I wouldn't emphasize past work experience. Then you might have a fighter's chance.

user55340

> Epic just wants young, impressionable folks roaming the halls of the campus. That allows for controlled costs in labor and it hides the drawbacks with the company because the newbies don't know any better. You would see some gray-haired folks at Epic but they're almost always employees that have been there for 10+ years.

user55340

It looks to be a nice campus, and the Madison area is quite nice... very googleish in parts of the culture, but also very... insulated from outside ideas.

user41796

@MichaelT snow can make for some great insulation.

user20683

18:50

@MichaelT Yeah, I was kind of "born older".

user55340

@GlenH7 I think its more "far away from the big tech places" while still "close to a great university". They are able to get college kids without experience (who want to stay in the area) without having to go to older, more experienced people have different ideas (good or bad).

user41796

I was trying to play off of the location and the frequency of snow up there. I failed. :-) Some of the Indeed reviews are hilarious though

user55340

@GlenH7 I did get the joke... and yes, it is. Though honestly, the past couple of years have been wonky for snow.

user20683

@MichaelT I've never understood fooseball as a perk.

user20683

or ping pong

user55340

18:58

First year here in Eau Claire...

user55340

user55340

See that big vertical line... yea, that was fun.

user55340

So I got a 2 stage snow blower... the next year...

user55340

Hmm. Used it twice.

user55340

18:59

This year, I was happy with it.

user55340

user20683

@MichaelT It barely snowed the entire time I lived in Germany.

user55340

This past spring, we kept expecting it to be over... and then we'd get another foot of snow.

Onwuka

Is it common to write a ticket saying "there are a few tickets in the queue that are fixed but not updated in the queue"?

user55340

@Onwuka Kind of, though it is a task that needs to be done and a 'bug' in the process if people aren't doing it.

user20683

19:16

@JimmyHoffa

user20683

chimera.labs.oreilly.com/books/1230000000929/index.html

Onwuka

19:39

@MichaelT thanks.

user55340

@Onwuka It could probably be written in a better way that clearly spells out what the problem is so that it can be corrected... rather than as an observation.

user55340

The observation would be closed as a WORKSFORME or SOWHAT close reason. The problem behind the observation - "fixed tickets remain in the todo queue" (yea, thats kind of an observation too) indicates that something needs to be done to keep that from happening, be it process or tool enforcement.

user20683

3

Onwuka

@MichaelT yeah, I am afraid of being taken the wrong way especially if it is a new team

On another topic, I never understood branching in svn or tfs. I mean, just copy the folder over makes me ask... why bother with version control?

user55340

@Onwuka Let me tell you a story from my college days, when I learned the value of version control...

user55340

19:46

It was with RCS, a very old system today, but reasonable back then. I worked on a project with another guy in CS 538 - we had partners. Simple programs. We sat in the lab together, stubbed out the functions we needed and then divied them up and went home and wrote them.

user55340

When working on something, you'd lock it. This made us use lots of small files so we could have several locks on several files, but not interrupt the other person working. Opps, made a mistake, roll back... Or if the other person has it locked, branch the file and work on your own copy, and merge it back when done.

Onwuka

you mean like a partial class in c#?

user55340

We were done days ahead of other teams that didn't - they all had to be in the same place at the same time to work (we could work remotely). When we had a flaw, we could roll back and work from where we made the mistake.

user55340

This was old C. So we had header files with simple hard coded values for functions and procedures. The work was 'just' writing the functions and remove the hardcode.

user55340

But yea, we got done much earlier because we could both work without fear of messing up the other person, or if we did make a mistake we could go back to where it worked eaislly. Other people had drunken coders on single files... you save your working code, and your partner after a night of drinking and trying to code at 3am makes it a mess...

Onwuka

19:50

@MichaelT oh wow

user55340

Or someone (not drunk) saves over the file you just wrote out.

Onwuka

By the way, I never got beyond what are basically hello world in C.

user55340

When you get into branching, each branch has a role... and thinking of it in that way makes it easier to see why you would do what where.

user55340

vance.com/steve/perforce/Branching_Strategies.html

user55340

19:53

You've got your main code. Only have working code in the mainline. You don't check in broken code. But some times, you need to do high risk development (HRD). But you can't check that into the main line. You need another branch. So feature A gets its own branch, feature B gets its own branch... C does too.

user55340

So, you've got your development done. But A and B did some drastic changes that make the merge back into the mainline conflict and be risky... so you make an accumulation branch (off of the mainline), and then merge A and B into the Accum branch, and then merge that back to the mainline.

user55340

At any time, mainline is good and safe to work from.

user55340

Now, you need to package it for release. So you branch one for each release, do the tweaks there, merge to the the final release package, and then merge that back to the mainline.

user55340

bradapp.com/acme/branching/branch-creation.html (hmm, @gnat this one looks useful too)

Jimmy Hoffa

Just signed all the closing papers!

New research shows over 60% of the financial collapse's toxic assets were created by power drills.

user20683

19:57

@JimmyHoffa congrats :)

user55340

Just forwarded that to a cow-orker who closed on Friday.

Jimmy Hoffa

@GlenH7 My issue with this is you have f1 level logging something so you don't make the calling function log it, and function f2 logs for something else, and later down the road f2 needs to call f1, now f1 catches exceptions, logs, rethrows, f2 catches the rethrow and logs. This happens constantly. If everyone would maintain all the logging to ensure proper paths such that you don't have now f1 and f2 responsible for the same exception which causes duplication of responsibility and code

then i'd like the standard approach. but it's not that simple. software is just too complex, it's not that people are lazy or not good enough, we just aren't going to be able to maintain rational logging in this fashion, however if you make a consistent rule that exception handling and logging is only handled at the top, you will at least have a usable consistent system that will not decay over time

the alternatives of handling logging/exceptions wheresoever "it makes sense" decays horribly because the software changes and it makes sense in other places over time and you inevitably get this stuff sprinkled everywhere until there's no clear guidance whatsoever to the people working on it or maintaining it where or how they should be handling this stuff

</rant>

really, I'm done now, I promise.

(until next time)

user55340

@Onwuka so, does that help any? or did I just confuse it more?

Onwuka

20:17

@MichaelT yeah because you want to keep code checked in but you don't want to break the build for others

Parallel software development sounds so fancy :)

user55340

@Onwuka And, if done right, you can merge changes from the mainline back into your code...

user20683

@Onwuka It's not horribly fancy, just difficult.

user55340

user20683

Languages like Haskell and Erlang make it easier

user55340

Its just difficult until you realize that not doing it is more difficult.

user55340

20:19

@WorldEngineer parallel development (as in two or more people working on the same code - not parallel development - multiple threads running from the same code)

user20683

oh

user20683

it's also not horribly fancy, it's not particularly difficult and languages like Haskell and Erlang make it harder because most people don't know them.

user55340

@Onwuka That image is a branch for some developers who wanted to check in dangerous, broken, or incomplete code. that is an incomparable policy from the /devel_line - so branch.

user55340

Another model branching for releases...

user55340

20:23

You branch for each release, and merge back to the mainline with each release. Make sure you branch off of the mainline.

user55340

Bad things happen if you do this:

user55340

Jimmy Hoffa

@WorldEngineer You know, I could actually imagine conflict resolution for a line of haskell code could be a lot trickier and more common than in other languages... the semantic density and the fact that it takes so many fewer lines for an application means less places for changes to occur so more conflicts

user20683

@JimmyHoffa plus you've got some many possible approaches to solving problems. It makes it more variable than some languages.

Onwuka

@MichaelT it just seems archaic compared to how git "magically" transforms your working space

user55340

20:27

@Onwuka ever look at git flow?

user55340

nvie.com/posts/a-successful-git-branching-model

user55340

And you know what I see there? I see a release line (blue), a packaging line (green), a mainline (yellow) and HRF branches (pink).

user55340

its just tilted on its side...

user55340

Onwuka

20:31

makes more sense now

so branching and merging is still the same conceptually

user55340

They are the start and end of a process.

user55340

(though merging can be done to pull changes in the middle - its not a bad thing to keep up with the mainline in your development branch... minimizes merge conflicts).

user55340

Or rather, moves some of the merge conflicts into your branch when you can handle them more easily than when you try to move them to the mainline.

Onwuka

when I import a root level certificate in windows 7, where does it go?

user20683

@Onwuka That sounds like a question for Super User

user55340

20:44

1

Q: How can you import a root certificate to a machine level store in Windows 7

I have a service (Running as local system) that uses an SSL connection. Currently this connection fails because the remote host used a private CA to sign it's certificate. For previous operating systems, I used to use the certificate manager to import the CA cert into the local machine's Truste...

windows-7 ssl-certificate

Onwuka

20:58

very nice. you're awesome, Michael :D

user55340

@Onwuka Fear my google-fu!

user41796

@MichaelT hilarious comment to see as the last line in the chat room

user55340

@GlenH7 I try.

Jimmy Hoffa

@MichaelT Remember pre-google programming? You always had to do depth first searches as opposed to google giving you a more efficient breadth first search, times were tough man.

Onwuka

@MichaelT :D

Jimmy Hoffa

21:04

I used to actually open books to find API information, I kid you not!

user41796

@JimmyHoffa I was explaining to a coworker how back in the day, I worked on a medical machine that had a bare connection to the interwebs and was ftp'ing data to a system back in my professor's lab.

user41796

and yes, that's FTP no S.

Onwuka

If I applied for a job at Google, they could just pull my search history and deny my application :P I ask so many stupid things on Google search

user55340

@JimmyHoffa Yep. I've got stacks of books from those days.

Jimmy Hoffa

@GlenH7 haha yeah, then there's good old security standards of the era, if you could call them that

Onwuka

21:05

Sometimes I even Google the word Google :-\

user41796

Googol

__notoc__ A googol is the large number 10100; that is, the digit 1 followed by 100 zeroes: :10,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000. The term was coined in 1938 by 9-year-old Milton Sirotta, nephew of American mathematician Edward Kasner. Kasner popularized the concept in his 1940 book Mathematics and the Imagination. Other names ...

user55340

The thing I learned in my first job (tech support) - it isn't what you know, but knowing how to find out what you don't know.

psr

Which is much easier if you know less!

user41796

@psr nah, we were just smarter then. :-)

Jimmy Hoffa

@GlenH7 I'm going to pretend that's a than with a sarcastic blank space after it

purposeful cognitive dissonance ftw

user55340

21:10

@JimmyHoffa btw, dye your grey hairs... next thing you know, you'll be ~~reminding~~ telling me about how you used to program in an IDE that was just a text editor with a 'compile' button next to it... or using a separate ~~tty~~ window to type make in.

user41796

Ctrl-x Ctrl-q

Jimmy Hoffa

@MichaelT I didn't setup initd to give me 12 consoles so I had to C-z everytime I wanted to run make...

user55340

I was thinking about the days of the glass tty (vt100!), and pulling one over next to the one I was working on so I could make in one window and code in the other and keep both screens visible.

user55340

Each of the systems had an x11 head that you could log onto, but the vts were less commonly used. So you just grabbed two of 'em...

user55340

21:17

(note the lack of arrow keys on that terminal - which is why I'm quite ok with hjkl navigation in vi.)

Jimmy Hoffa

@MichaelT ahhh, I've been meaning to force myself to use emacs navigation keys, I just haven't gritted my teeth to do it yet

it's going to be painful for a little while I"m sure

user55340

It helped I'm an old school rogue player... where those were the movement keys.

Jimmy Hoffa

I don't remember that... I played a fair bit of nethack, and I clearly recall it having some odd controls, I probably never associated it with vi because I always used the arrow keys..

(in vi that is)

user55340

Hmm... interesting article...

user55340

Arrow keys

Cursor movement keys or arrow keys are buttons on a computer keyboard that are either programmed or designated to move the cursor in a specified direction. The term "cursor movement key" is distinct from "arrow key" in that the former term may refer to any of various keys on a computer keyboard designated for cursor movement, whereas "arrow keys" generally refers to one of four specific keys, typically marked with arrows. Arrow keys are typically located at the bottom of the keyboard to the left side of the numeric keypad, usually arranged in an inverted-T layout but also found in diamo...

user20683

21:23

speaking of Vi, there is a script that lets you reply to messages with : and up arrow

user55340

With various variations. I'll have to look at the ESDF mapping instead of WASD for some games.

user55340

> Examples of games that use HJKL are the text-based "graphic" adventures like NetHack, the Rogue series, and Linley's Dungeon Crawl.

user20683

@MichaelT you might enjoy FTL if you're into Rogue-likes

user55340

@WorldEngineer I kicked it back when it was a starter.

user55340

Though I'm more a Dungeons of Dredmore fan.

user55340

21:29

MichaelT, Eau Claire, WI

594 1 3 11

dungeons-of-dredmor realms-of-the-diggle-god world-of-goo

user20683

@MichaelT That one was pretty good too

psr

Turns out red goo is flammable

user55340

@psr Yes, yes it is.

user55340

I'm still happy with my famous question for runes in DoD.

user55340

I've gotta play it a bit more... just getting back to it and trying to finish up my non-hardcore random build and get that achievement...

user55340

21:38

Hmmm...

user55340

+1 for extra lutefisk. :) — edsobo Dec 27 '11 at 14:07

user20683

►

user20683

This looks mildly interesting

user20683

if only because it shows Keanu Reeves doing something that resembles emoting.

psr

@MichaelT How can you live with yourself after playing a game with a horadric lutefisk box?

Jimmy Hoffa

21:40

@WorldEngineer Saying "dude" is emoting, isn't it?

user55340

@WorldEngineer dude.

user55340

The key to getting a good Keanu movie is to restrict his script to the word "dude".

user55340

@psr its either that, or go back to playing dwarf fortress...

user55340

> One of my Butchers just created a figurine of a Roc ripping an Elf apart. It's called Gorrothedos, "The Disembowelled Amazement".

user20683

The cast is actually Japanese which is nice.

user20683

21:46

always good to have the ethnicities mostly line up.

gnat

@MichaelT yes, that's indeed a good one. In my monster post, there is a reference to cmcrossroads.com/bradapp/acme/branching which builds rather heavily upon it. Draft version post also had a direct reference to it, but when I was forced to cut corners to fit into answer length limit, it went away. IIRC, in a battle for survival it lost to "Microsoft Team Foundation Server Branching Guidance" (by quite a little margin)

gnat

22:41

so much drama this time, oh my

"Hotness score" of this question is about 2/3 fake, as usual - 11 of 16 answers score less than 1/10 of top-voted one, but each of these meh answers brings solid 10 points to question score thanks to bug in the formula. Unnaturally high position at collider damages a reasonably okay question, as usual. - What should we do about this question? - Fix the freakin' bug in hotness formula that causes drama like that — gnat 19 mins ago

nothing surprising really

2 days ago, by gnat

when you know how collider bug works, it's like watching the train wreck in the slow motion

Jimmy Hoffa

23:03

@gnat this graphic alone has made each of your recent hotness complaints entertaining

Transcript for