The Official Status Thread
-
Status: Episode 10
-
work smarter not harder.
Why "Work Smart, Not Hard" is the Worst Advice in the World
Work smarter, not harder? Don't tell it to Dirty Jobs host Mike Rowe, who meets some of the hardest-working people in America. In fact, he argues, that mantra is the opposite of the attitude we need to beat this lousy economy.
This popular mechanics article only make sense if you take the "work smarter" part as getting a random degree.
A heavy-equipment technician with real-world experience can earn upward of six figures. And the training program is free! But still the positions go unfilled? In a state with 9.6 percent unemployment? What's going on?
Finding opportunities like this count as smart in my book.
-
Additionally, "work smarter not harder" doesn't refer to ignoring trades in favor of knowledge work.
It refers to driving efficiency. Ironically, trades have the efficiency thing down, and knowledge workers by and large do not.
-
The biggest problem with the mechanical trades is that workers are expected to buy their own tools. So instead of being in hock to regulated financial institutions for 30k in tuition, they're in hock to Snap On or Mac Tools for 30k.
Heavy equipment tools are even more expensive than that.
It's basically the equivalent of my employer saying "Go buy an MSDN Enterprise membership for yourself. And a Resharper license. And a SQL Server CAL. And a computer. And we aren't paying for any of it".
-
Status: looking at front page code (that is, the code that powers the front page, not code that was on the front page)
-
Status: I wrote a Dockerfile for NodeBB that is 27 lines long. Without comments and blanks, it's 10 lines long. It works without any further setup.
tdwtf-staging/nodebb/Dockerfile
``` FROM ubuntu:latestRUN apt-get update
install a bunch of dependencies
RUN apt-get install -y build-essential curl git imagemagick python libkrb5-dev mongodb
install nodejs
RUN curl -sL https://deb.nodesource.com/setup_0.12 | bash - && apt-get install -y nodejs
download nodebb v0.8.2+
RUN git clone -b v0.8.x https://github.com/NodeBB/NodeBB /opt/nodebb
download more dependencies
RUN cd /opt/nodebb && npm install --production
we use supervisor to keep nodebb running
RUN npm install -g supervisor
set the working directory
WORKDIR /opt/nodebb
expose the port
EXPOSE 4567
start it up!
CMD service mongodb start && supervisor app
</details> I haven't started writing a Discourse docker build script, but luckily, @sam wrote one for me. Let's see how big it is. Well, we've got an image named `discourse`. Let's check it out. - https://github.com/discourse/discourse_docker/blob/master/image/discourse/Dockerfile Ok, but it seems that one uses an image called discourse/base. Let's check that out as well. - https://github.com/discourse/discourse_docker/blob/master/image/base/Dockerfile But wait, there's also a 54-line Ruby script that actually builds the images. - https://github.com/discourse/discourse_docker/blob/master/image/build.rb ...which needs to be run as root, negating any safety obtained by doing things inside containers. And there's also a 624-line bash script that you need to build a container that can actually be run. - https://github.com/discourse/discourse_docker/blob/master/launcher And that uses a 109-line config file: - https://github.com/discourse/discourse_docker/blob/master/samples/standalone.yml Which pulls in five additional config files, for a total of 616 more lines: - https://github.com/discourse/discourse_docker/blob/master/templates/postgres.template.yml - https://github.com/discourse/discourse_docker/blob/master/templates/redis.template.yml - https://github.com/discourse/discourse_docker/blob/master/templates/web.template.yml - https://github.com/discourse/discourse_docker/blob/master/templates/sshd.template.yml - https://github.com/discourse/discourse_docker/blob/master/templates/web.ratelimited.template.yml What exactly is the point of a system that safely builds server images and then lets you share and execute them for reproducible results if you're going to use it as root and not publish your images, instead choosing to build them all locally and not cache anything?
-
Status: trying to understand how Ben is smart enough to build all this stuff - and not yet smart enough to realise that the Discourse crew are a bunch of wastrels pretending to be smart enough to write something useful, but actually aren't.
-
The thing is, this forum is running Discourse. And I'm not going to be able to set up a local development version of TDWTF if I can't run Discourse locally.
-
And I'm not going to be able to set up a local development version of TDWTF if I can't run Discourse locally.
Good job all the Docker shite is done for Dischorse then.
-
I don't want to run a script as root that generates a container by running scripts in containers that it immediately deletes. Something about that is just intrinsically wrong.
-
Ok, so buried inside that mess is "by the way, you need to configure an email server".
I recall something called mailcatcher that I used back when I was testing Discourse originally.
Turns out the latest (and also every other) mention of mailcatcher on meta.d assumes that you're running Discourse outside of a container directly on your development machine.
Apparently nobody runs Discourse outside of production.
-
Apparently nobody runs Discourse outside of production.
Certainly not the Discodevs.
-
Status: NodeBB has about 1000000 copies of this in its log:
Starting child process with 'node app' 23/11 00:05 [164] - info: NodeBB v0.8.2 Copyright (C) 2013-2014 NodeBB Inc. 23/11 00:05 [164] - info: This program comes with ABSOLUTELY NO WARRANTY. 23/11 00:05 [164] - info: This is free software, and you are welcome to redistribute it under certain conditions. 23/11 00:05 [164] - info: 23/11 00:05 [164] - info: Time: Mon Nov 23 2015 00:05:19 GMT+0000 (UTC) 23/11 00:05 [164] - info: Initializing NodeBB v0.8.2 23/11 00:05 [164] - warn: You have no mongo password setup! 23/11 00:05 [164] - info: [database] Checking database indices. 23/11 00:05 [164] - info: NodeBB Ready 23/11 00:05 [164] - info: NodeBB is now listening on: 0.0.0.0:4567 crashing child 23/11 00:05 [164] - info: [app] Shutdown (SIGTERM/SIGINT) Initialised. 23/11 00:05 [164] - info: [app] Database connection closed. 23/11 00:05 [164] - info: [app] Web server closed to connections. 23/11 00:05 [164] - info: [app] Shutdown complete.There's nothing else. No information about why it's crashing, just that it is.
-
Status: I'm approaching the limit of what you can do in Fallout 4 without allying with one of the factions-- just had to revert a quest with one of them that made me enemies with another. (Even though I had no intention of actually doing the quest. Bit of bad design there from Beth.)
Still only have 2/3rds of a X-01 set... not sure where to find the arms. But I painted it in the shark paint job. Here's its ass because I didn't screenshot its front:
Also slightly curious if anybody in the history of Fallout has ever allied with the Brotherhood of Steel. Even though that's an option in every single game. (Except New Vegas, IIRC.)
The way the faction quests interact with each other in Fallout 4 is ... amazing. And so far I haven't hit any bugs due to completing faction tasks of multiple factions in essentially random order. Huh.
-
Also slightly curious if anybody in the history of Fallout has ever allied with the Brotherhood of Steel. Even though that's an option in every single game. (Except New Vegas, IIRC.)
Guilty.
Allying with the Enclave in FO3 just seemed wrong to me. They had fucking Vertibirds, they didn't need my badass self.
-
A wife asks his husband which is a computer programmer to go to the market and buy a dozen eggs, and if there is milk, buy 2.
Husband comes home with 2 eggs, because there was milk there.
-
The thing is, this forum is running Discourse. And I'm not going to be able to set up a local development version of TDWTF if I can't run Discourse locally.
Did you check the Discourse Developer Install Guide (Vagrant)?
-
strong textStatus
wtf discourse?strong text
-
-
Status: Cooking lunches for the week.
Pork curry! Because I wanted curry and I had a bunch of pork in the fridge.
Plus the office is almost empty all week so there will be minimal butthurt over the smell!
-
Status: discovering that if you Discourse on a brand new iPad Air 2, Discourse is about usable, as opposed to being mostly (badly) read only due to shitty performance on a two year old first-gen iPad Mini.
The extra RAM and CPU really help.
-
Status: Local dev environment is up and running!
You can probably get to it if you try hard enough.
-
May the force be with you
-
Status: Eagerly awaiting a discussion where this xkcd is relevant:
-
As long as you're not Discoursing on Chrome on that iPad. Unsupported by Jeff.
-
Apparently nobody runs
Discourse outside of productiontest environments in the real world.Sadly, far more prevalent than you probably think.
-
Status: FUCKING TELCOS!
I estimate basic config of our PBX system (assuming no complicated queues, basic IVR if any, local network and phones already set up) to 30-45 minutes. I'm usually even faster than that, barring any weird demands.
It's never less than 2 hours. It was 4 today. Reasons mostly being, in order from most common to least common:
- I don't get the credentials and network setup info telco promised will be sent yesterday
- They send me the info but the user account is not actually activated
- They send me incomplete network setup info
- They send me wrong network setup info
- The info I get is correct, but they fuck up the network config on their side
- The info I get is correct, the network on their side is correct, but the tech that set up the router at client's location fucked up
- My boss agreed that the client can provide their own hardware and they manage to find the one motherboard that won't detect an addon card we need to install
- My boss agreed that the client can provide their own hardware and they manage to find the one network card that doesn't work properly on a Debian system
- I actually fucked up somewhere and it takes me an hour to figure it out.
Today it was a combination of 1, 3 and 5.
Bonus: After calling them about 5 they promised to escalate and call me back. They actually fixed it (MIRACLES! THEY HAPPEN!) and then never called me back, leaving me to twiddle my thumbs for an hour on-site.
Every day I'm more surprised I can even access the Internet...
-
All telcos are incompetent.
-
Someone should photoshop that into "Breaking news: 10 random people agreed 9-1 that Discourse is bad. No word on if the 10th person was Jeff."
-
I estimate basic config of our PBX system ... to 30-45 minutes. I'm usually even faster than that, barring any weird demands.
It's never less than 2 hours.
I think I found TRWTF.
-
Every day I'm more surprised I can even access the Internet...
That's better than being surprised that you can't...
-
Status: WTF SQL server. Why throw an error for the hourly transaction log backup maintenance plan when the problem was you couldn't write to the history file because the weekly maintenance was still finishing with cleaning it up? For that matter, why throw your hands up and kick an error there at all, rather than, oh, I don't know, waiting for the lock to clear.
Grumble I wish I would have used the Ola Hallengren scripts to begin with... Oh well, at least my problems will go away when we finish switching to Veeam in application aware mode, right? What's up with the laughter over there, guys?????
-
Yeah, expected that... That's not what I tell to clients, that's my own estimate on "if nothing goes horribly wrong, that's how long it would take".
I usually say it's at least 4 hours to the clients, because of the above. It's still infuriating though.
-
I'm just saying, if you estimate basic config as 30-45 minutes, and it's never less than 2 hours, your estimation process is TRWTF.
Also, a DCAP is supposed to be able to get to dial tone in 90 minutes or less
-
Also, a DCAP is supposed to be able to get to dial tone in 90 minutes or less
Assuming that the other side of the connection doesn't mess things up, and that the hardware provided is actually capable of doing the task, of course. If the setup has all gone Belgium-shaped, the 90 minutes are a complete fiction.
-
-
-
**Status:**strong text Installed SQL Express last week on a spare PC so I could install an ERP demo for testing. In spite of the fact that Windows 8.1+SQL Express is not a supported combination, and that SQL Express is explicitly non-supported, it still worked. But it SE ate 1.7GB of RAM. Guess I'll have to not keep it.
-
lubar.me
-
VB.NET develoeprs
Love between VB.NET developers is forbidden, because we do not want them to reproduce. So, if they want to get married, they have to elope. We call them "develoeprs" (it's phonetic).
-
Status: Epi-snowed 11
-
Also, a DCAP is supposed to be able to get to dial tone in 90 minutes or less
When does the timing start? Fresh machine? OS installed? Asterisk installed?
I can do it in 90 minutes from freshly built machine with no sweat ...
Assuming that the other side of the connection doesn't mess things up, and that the hardware provided is actually capable of doing the task, of course.
Yeah, that.
-
Status: If anyone wants my TDWTF staging script, here's how to run it:
git clone https://gist.github.com/130e7439401cd123ccdc.git tdwtf-staging cd tdwtf-staging ./oh_god_whyHere's the relevant section of my nginx config:
server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name discourse.local.lubar.me; location / { proxy_pass http://172.17.0.3; proxy_set_header Host $host; } } server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name mailcatcher.local.lubar.me; location / { proxy_pass http://172.17.0.3:1080; proxy_set_header Host $host; } } server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name nodebb.local.lubar.me; location / { proxy_pass http://172.17.0.2:4567; proxy_set_header Host $host; } } server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name tdwtf.local.lubar.me; location / { proxy_pass http://192.168.1.100; proxy_set_header Host $host; } }IP addresses may vary. Here's how to get them:
$ docker inspect -f '{{ .NetworkSettings.IPAddress }}' tdwtf-staging-nodebb 172.17.0.2 $ docker inspect -f '{{ .NetworkSettings.IPAddress }}' tdwtf-staging-discourse 172.17.0.3The
192.168.1.100is the address of the machine running IIS.
-
Here's the relevant section of my nginx config:
server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name discourse.local.lubar.me; add_header X-Clacks-Overhead "GNU Terry Pratchett"; location / { proxy_pass http://172.17.0.3; proxy_set_header Host $host; } } server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name mailcatcher.local.lubar.me; add_header X-Clacks-Overhead "GNU Terry Pratchett"; location / { proxy_pass http://172.17.0.3:1080; proxy_set_header Host $host; } } server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name nodebb.local.lubar.me; add_header X-Clacks-Overhead "GNU Terry Pratchett"; location / { proxy_pass http://172.17.0.2:4567; proxy_set_header Host $host; } } server { listen 0.0.0.0:443 ssl; listen 0.0.0.0:80; server_name tdwtf.local.lubar.me; add_header X-Clacks-Overhead "GNU Terry Pratchett"; location / { proxy_pass http://192.168.1.100; proxy_set_header Host $host; } } ```</blockquote> Fixed your config for you. <small><small><small><small>It was missing `X-Clacks-Overhead` headers
-
Back last I took the test, the timing was for a fresh machine with OS installed and the Asterisk source code downloaded (their testing machine stock image, basically). And in the DCAP test scenario specifically, you have to get DAHDI configured for 1 line (physical handset), get a SIP extension configured, and get those two to be able to call each other, plus out through a SIP trunk to the testing server IVR. Oh, and no Internet usage except to voip-info.org (I assume the official Digium wiki is allowed now too).
Last time I attended Astricon, the "first to the dialtone" competition under those similar circumstances winner was around 7 minutes. When I took my DCAP test, I spent a while screwing around setting up Voicemail and some other services, just because I could.
-
Oh, yeah, that's perfectly doable with enough time to play solitaire in 90 minutes. I couldn't do it in 7, mostly because I have no experience with DAHDI (simply hasn't come up, we mostly install systems for people who have to get rid of ISDN these days, not the other way around), and because I recheck my config multiple times and work relatively slowly (too much time wasted on stupid typos in the past).
Filed under:
canreivnite=yes
-
Status: Wondering what take home pay for a traveling consultant must be. Hiring one who will be commuting from the other side of the country and staying in a hotel all week. Billing rate is 125...
I figure he must be pocketing a mere 60k after subtracting travel expenses and agency profit.
-
**status:**strong text My alarm rang me today and told me that the power had gone in the house. Went home early, and then I had to go out and dig up the connection box for the water pump from the snow and wipe away all the moist it had gathered and that triggered the residual current circuit breaker. Then back in to light a fire to get the heat back up to reasonable, because of course this happens on the first day this season when it is -15 degrees Celcius. Expletive!
On the plus side, I got to work from at home during the afternoon in front of an open fire and surrounded by two cats and with my feet buried in the fur of my dog. He snores, and so does one of the cats.
Filed under: one of these days I need to fix that damn connection box properly
-
Snoring cats and dogs are a plus?
-
Status: Is there something wrong with Java 7 and Java 8 that companies aren't upgrading to it?
Got an email last week and another today from vendors of our client that are saying they can't connect to the new server we migrated our upgraded applications to, so their applications can't connect to ours. Doing some legwork and it appears to either be an SSL/TLS issue (not as likely) or SNI (which we are using for the client's clients), which appears to be fixed with Java 7.
