The Official Status Thread

Office closes from Christmas Eve until new year anyway, and I'm going on holiday in the first week of January, so less than two weeks working here. They might give me garden leave but I'll probably need to do some knowledge transfer and clear unfinished business

Rhywden

Status: The car dock with Qi charging capabilities for my mobile has the annoying habit of pressing both the volume down and the camera buttons, due to the unfortunate positioning of the two plastic bars to the side.

So I kludged a bit.

Maybe I'll paint them black when I have the time

blakeyrat

And somewhere, someone's tongue remains undepressed.

dkf

Status: This one is for @ben_lubar… and I want to try to avoid ever reading it again.

Tsaukpaetra

@blakeyrat said:

someone's tongue remains undepressed.

Engine: I doubt the recipient is complaining...

FrostCat

@ben_lubar said:

Status:

@: You: Why are you traveling?
1: Leteng Ithbituthru, Surgeon: I'm going out for some water.
@: You: How are you feeling right now?
1: Leteng Ithbituthru, Surgeon: I've been well.</blockquote>

Leteng Ithbituthru, Surgeon meant that he had been to the well.

FrostCat

@dkf said:

FTFY. The first spoken words were probably something like “Damn you LLVM, can't you ever give a good error message?!” (Clang does things nicely. LLVM does not.)

Awww, you people who've never used the VMS ADA compiler and think you've seen bad error messages are so cute.

blakeyrat

@dkf said:

Maybe not everything everyone else does is shitty

Unlikely.

FrostCat

@Luhmann said:

Where is the nearest bridge?

Ask @blakeyrat, he lives under it. But at least he has good beer, if he wants it.

JazzyJosh

The Living Computer Museum in Seattle was pretty damn cool.

I think I borked some bits there though.

Eldelshell

Funny you mention this, but in our previous office we had this downstairs

Very classy place

Eldelshell

Status: got this user story today:

As a user, I don't want to have adults section when I am in LG

I really don't have any idea of why.

Weng

No porn on LG TVs obviously.

That said, some specs are just fucking stupid when rephrased as a user story.

Eldelshell

Yes, I'm able to understand that, what I have no explanation for is why are LG users not allowed to have pr0n.

Maciejasjmj

@FrostCat said:

Leteng Ithbituthru, Surgeon meant that he had been to the well.

I wonder if taking the joke and retelling it in a much more obvious (and less funny) way qualifies for a whoosh...

Weng

Presumably because LG says so?

Eldelshell

That would be even more stupid... I'll have to check on this.

cartman82

@Eldelshell said:

Status: got this user story today:

As a user, I don't want to have adults section when I am in LG

That's a pretty anti-climactic story.

locallunatic

@cartman82 said:

@Eldelshell said:
Status: got this user story today:

As a user, I don't want to have adults section when I am in LG

That's a pretty anti-climactic story.

FTFY

blakeyrat

Status: 3 alarm fire on staging server. (Stage is considered production at our company; it's for client data staging, not development staging.) On the virtue of literally no other back-end developers being in the office, I'm assigned the ticket.

I don't know the cause, but I already know the solution: the database is locked-up by some process which needs to be identified and killed. Basically: someone needs to run sp_who2 and kill the process everybody else is waiting on.

Since the ticket is mine, I guess the task is mine. Except:

I have no access to stage (nor do I want it)
Even with access, I have no ability to kill a DB process; the best I can do is communicate to the lead server admin which process needs to be killed

So. I've been waiting about 3 hours now to get a login to a server so I can run literally a single command in SQL. The first hour of that was to get VPN access to the prod VPN, and now we're trying to figure out SQL Server access. I have an account, apparently, but it is not allowing me in.

All of this so I can run a single SQL command, and based on the results send out an email. (Well, ok, there's a 1/100,000,000 chance that an actual code problem is involved but.) Oh and people are blocked on this, naturally. I brought up the whole, "why doesn't the lead server admin run the command?" with no results. THE PROCESS MUST BE FOLLOWED.

In the meantime I can't get any other work done, because I'm interrupted every 15 minutes with, "ok try THIS password, it'll work this time."

Tsaukpaetra

@blakeyrat said:

why doesn't the lead server admin run the command?

Because they're offline.
Good luck, we have a turnaround time on issues like these of about three hours, 2.99 hours of it are "waiting to see if it will complete".

dkf

@FrostCat said:

Awww, you people who've never used the VMS ADA compiler and think you've seen bad error messages are so cute.

I think you've misunderstood what I'm talking about.

Here's an example:

Assertion failed: (C1->getType() == C2->getType() && "Operand types in binary constant expression should match"), function get, file /unimportant-build-location/llvm-3.7/work/llvm-3.7.0.src/lib/IR/Constants.cpp, line 1875.

This rather long line is the only message printed. No other output was produced at all; the process instead keeled over. The bug was nowhere near to that point, needless to say, and the LLVM source was not in that location any more. Nor was it easy to determine what was actually wrong from looking at the stack trace; it was spat out rather deep inside a lot of other complicated code.

Clang produces good error messages. It genuinely tries to figure out what you might have meant and point to where you might fix things; it's even right about 80% of the time. LLVM is not the same thing; it's tools for people writing compilers. That just hits an assertion check (not an exception throw; that would be too good) somewhere deep inside and dies with a message that it is hard to relate to your own code. Or maybe you get to the point where you've actually made a bunch of functions; verifying your intermediate code is pretty much a mandatory step, and the verifier tends to either tell you which function has the error in it, or what that error is, but not both at the same time.

You can also get runtime crashes from your code just being wrong but type-correct, such as dereferencing a pointer that's actually null but not provably so statically. But that's where the compiler author ought to worrying about things; it's fair game that if you put bad code in, you get crashes out.

fbmac

@blakeyrat said:

So. I've been waiting about 3 hours now to get a login to a server so I can run literally a single command in SQL. The first hour of that was to get VPN access to the prod VPN, and now we're trying to figure out SQL Server access. I have an account, apparently, but it is not allowing me in.

Here I would need at least 6 months to get all those accesses, with PHB asking me everyday if I have solved the problem, while I know he really expects me to sneak into the datacenter and press the reset button in the server.

blakeyrat

Conclusion to the DB access saga:

The problem solved itself.

(Coincidentally when the other interactive user on the DB server was disconnected-- hmm!)

blakeyrat

Status:

Just ran across this comment on Hacker News that's a pretty good summary of IT:

https://news.ycombinator.com/item?id=10770064

Never underestimate what sorts of systems are possible to create and force to continue running by current organization management practices. I personally worked on an extremely (and unnecessarily) large system that had no documentation. When I started, and asked for something that at least explained what the different end executables did, I found that no such document existed. When looking at a problem report which said "system X has problem Y", finding out what code was responsible for system X required talking to large numbers of different team members until you could find someone who knew where system X lived.

This system survived because no one in management could bear to pull the bandaid on doing significant rewrites and instead preferred to pay much larger amounts spread over longer time periods to continue accumulating technical debt. It basically made it take 6 months to a year before a new hire could be useful, but once someone was familiar with some corner of the system they mostly just stayed there. Keep in mind that most software engineers don't approach their job with passion. Most of the people I worked with didn't even want to touch a computer when they got home. They didn't spend their weekends contributing to open source projects, or keeping up with the latest techniques. They kept their head down and collected their paycheck. That is what the majority of companies actually want, and it's what they get.

That system has since (after a billion-dollar-plus contract to create a replacement system) moved to a 'devops' atmosphere. No automated testing or continuous integration, no, that would threaten the large test team. DevOps to them means what I suspect it means at many non-startup places. It means developers developing on the production environment. It means 10x as many problems cropping up, but management is thrilled because problems are fixed in hours instead of weeks. Development no longer consists of understanding anything, or designing anything, it's just an endless series of adrenaline fueled hasty patches while everyone keeps their fingers crossed that the data doesn't get so corrupted that everything grinds to a halt.

Arantor

I really need to print this out and staple it to the desk divider on the boss' desk.

ben_lubar

Status: I now have the recipes for the components of one of the components of one of the components of one of the components of one of the components of one of the components of the DISCOMACE.

FrostCat

@Maciejasjmj said:

I wonder if taking the joke and retelling it in a much more obvious (and less funny) way qualifies for a whoosh

No, and fortunately we don't have a "killed the joke by explaining it" badge.

FrostCat

@dkf said:

I think you've misunderstood what I'm talking about.

Maaaaaybe. Was the error produced by something like assert(), at runtime, a compiler bug, or something else? I'm guessing the former, although it could possibly have been the latter.

By contrast, what I meant was, as I've described before, certain basic errors in VMS ADA, like "forgot to instantiate the text io generic package for every single subtype you've defined in your program", will create an error that is like 5 lines long for every single subtype you've defined that you did remember to instantiate. It could span a couple of printed pages, and the text was a masterpiece of obtuseness.

FrostCat

@dkf said:

LLVM is not the same thing; it's tools for people writing compilers.

Also, until now I thought LLVM actually was a C++ compiler.

Polygeekery

Status: Every year around Christmas I come up with a shitload of great stuff to be recycled or just disposed of. Last year, it was a stack of gaming laptops. This year, it is a fucking fuckload of nearly new Macs. Mac Minis and iMacs, none of them more than a year old.

Backstory: Around a year ago a NFP that we work with had an Executive Director that had the bright idea to switch everything over to Macs. This was in spite of them using a shitload of software that only works on Windows. His plan was to use their XenDesktop environment to access the Windows apps. This was despite all of our advice to not do this.

Six months ago he was shitcanned by the board. Four months ago the new Executive Director apportioned funds to basically scrap all of the Macs. Yesterday he asked me to get rid of them all. He asked for $150/per credit on next year's invoices. So, in essence, I just bought an SUV load of Macs for $50 a piece.

Not a bad high note to end the year on.

HardwareGeek

Status: Big pot of homemade lentil soup on the stove, nearly done. Enough to last me a couple of weeks, maybe. For the first time ever, maybe, added enough liquid to start that it's coming out as soup, not stew.

FrostCat

**Status:**strong text The one Doctor Who Cheezburger macro I made just hit the front page again. Well, by the time I saw it it was on page two.

Dreikin

@Polygeekery said:

He asked for $150/per credit on next year's invoices. So, in essence, I just bought an SUV load of Macs for $50 a piece.

And how much do you charge for shipping?

dkf

@FrostCat said:

Was the error produced by something like assert(), at runtime, a compiler bug, or something else? I'm guessing the former, although it could possibly have been the latter.

It's an assert failure. That's how LLVM reports any bugs in the inputs you give it. When you're starting out to write a compiler, you get a lot of such bugs and much confusion.
@FrostCat said:

It could span a couple of printed pages, and the text was a masterpiece of obtuseness.

That sounds like some kinds of error you get in C++

The classic of the kind was when the compiler and the library supporting it disagreed on the version of the C++ standard they were supporting, giving utter confusion.

@FrostCat said:

Also, until now I thought LLVM actually was a C++ compiler.

It's a library for writing compilers. You give it input in a sort of virtual assembly language (which is a lot easier to write than real assembly) and it handles everything from there down to native code, including all the optimization. It has to be paired with a front-end that understands the programming language you care about: clang is such a front-end for C and C++ (and maybe some other languages too).

cartman82

@blakeyrat said:

Conclusion to the DB access saga:

The problem solved itself.

(Coincidentally when the other interactive user on the DB server was disconnected-- hmm!)

Frustrated-unable-to-help man does it again!

Up up and away! (into the bus home)

cartman82

STATUS:

Boss asks me for access to the production database. Wha? With some trepidation, I send him the SSH and db logins.

Cartman, these user passwords are encrypted!
Ummm... yes
So how do I decrypt them?
Ummm... you can't. That's kind of the point.
Oh. Never mind then.

loopback0

@Polygeekery said:

So, in essence, I just bought an SUV load of Macs for $50 a piece.

Not a bad high note to end the year on.

How much is shipping?

edit:

Onyx

Status: Got knick-knacks for people a the office. This is what the intern guy got:

Filed under: Yes, these pictures were taken using a potato

Jaloopa

Status:
Discovered that the Sage SDK has an Exception viewer, presumably because they obfuscate the exceptions that come out to the point where you're not even sure which method threw.

I open it up and it's not super obvious, but the Help menu item has a user guide.
Which throws an exception.
In the exception viewer.
Brillant.

Turns out it's in a pre-Vista mindset and tries to open the help file in Program Files, presumably with read-write privileges, and gets Access Denied if not run as an admin.
Brillant