Adventurous Computing

  • Home
Twitter RSS

Disassembling Pioneer SE-50 headphones

Posted on 20 December 2011 by cyryl
No comments

Today’s story is a photostory !
Here’s how to disassemble Pioneer SE-50 headphones, as I found this task quite difficult, mainly because no info available on the internet on how to open the can from the back.
As you can see Pioneer signs on the side of each can are removable. Take them off to uncover the screws.
Research made possible by kabanosy – best multitool ever.

Some more info on the headphones:
Manual scan
Produced between ’68 and ’72
Two speakers per can !

Photos taken after replacing the cable and before pots rejuvenation.

IMG_0011
IMG_0014
IMG_0015

IMG_0017
IMG_0018
IMG_0024

IMG_0026
IMG_0027
IMG_0029

IMG_0032
IMG_0033
IMG_0037

IMG_0038
IMG_0039
IMG_0040

IMG_0041
IMG_0044
IMG_0047

IMG_0048
IMG_0049
IMG_0050

IMG_0051
IMG_0053
IMG_0055

IMG_0057
IMG_0059
IMG_0060

IMG_0061
IMG_0062
IMG_0064

IMG_0068
IMG_0074
IMG_0076

IMG_0077
IMG_0079
IMG_0080

IMG_0082
IMG_0083
IMG_0085

IMG_0086
IMG_0089
IMG_0091

IMG_0095
IMG_0096
IMG_0097

IMG_0101
IMG_0106
IMG_0107

IMG_0108
IMG_0110
IMG_0111

IMG_0112
IMG_0113
IMG_0115


Categories: hardware | Tags: headphones, pioneer, repair, repairing, se-50, soldering, teardown

Configuring Tor non-exit relay

Posted on 6 October 2011 by cyryl
No comments

I’m for privacy. I’m for cryptography. I do provide Tor relays for the good of all people.
Here’s how to configure a classic non-exit Tor relay on your machine. In my case I got it running on the VPS server, with 2 IP addresses and some bandwidth quota applied.
I wanted the Tor traffic to be easily recognizable from outside as different from the ‘normal’ traffic coming from my server. Second requirement was to make Tor not use my whole traffic quota up.

Here’s my config, with some notes

cyryl@serv:~$ cat /etc/tor/torrc
SocksPort 0 # what port to open for local application connections
SocksListenAddress 127.0.0.1 # accept connections only from localhost
RunAsDaemon 1

DataDirectory /var/lib/tor

ORPort 9001 #switches Tor to server mode
ORListenAddress tor.cyplo.net
OutboundBindAddress 91.213.195.28 #what IP address use to direct the outbound traffic
Nickname cyplonet
Address tor.cyplo.net
ExitPolicy reject *:* # no exits allowed
AccountingStart day 09:00
AccountingMax 2 GB

MyFamily cyplonethome, cyplonet
Categories: blog

Ubuntu 11.10 Beta1

Posted on 22 September 2011 by cyryl
No comments

What do I think ? Easy to install, hard to disable Unity, some programs crash sometimes and some things are weird. Despite that I have the faith in the Ubuntu guys that the system would be ready when 11.10 comes.

As whining itself never got anybody anywhere, I’ve reported all of these here:

#855901
#834425
#855945
#855917
#855919

I’ve even got one OOPS. However, since I’m writing this post under 11.10, it seems usable, even the OOPS was not of disturbance as it got nice GUI window displayed and no system crash then.

Categories: linux, open source | Tags: 11.10, bug, crash, oneiric ocelot, opensource, ubuntu

More RouterBoard debugging

Posted on 15 August 2011 by cyryl
No comments

I just wanted to install OpenWRT on Routerboard 433UAH. And as usual I ended up having lots of cables. And yes, awful soldering, I know ;)

IMG_4030
IMG_4045
IMG_4047

IMG_4048
IMG_4049
IMG_4051


Categories: hardware | Tags: buspirate, jtag, routerboard 433uah

Having fun on weekend

Posted on 30 July 2011 by cyryl
1 comment

Categories: hardware | Tags: mikrotik, openwrt, routerboard 433uah

How to hire a programmer, a.k.a. tech recruitment explained

Posted on 24 July 2011 by cyryl
3 comments

Today’s post is not a recipe at all. Maybe a little. More of viewpoint than a recipe.

How to get the right people for the tech jobs employed ?  The most common way nowadays seems to be the one of paying lots of cash to the recruitment company to do this for you. How it ends ? There are few things that annoy me more than the people from the recruitment companies. I am not saying that all of them are annoying, no. Just the very most of the ones who contacted me. I find questions they ask fundamentally flawed. They appear to more like a pattern-matching mechanisms than the actual people. Maybe it’s because I am biased, I don’t quite get the idea of hiring non-programmer to hire a programmer.

Here’s what I think the tech recruitment should look like. It won’t be cheap. And yes, I conducted a lot of interviews in my life.

Start by describing the position. Who are you looking for, what would they do for you, how would they bring money to your company. Use words, not techie talk. I found most of the job descriptions online do not match the real expectations of the employer. You just want a good programmer, a person who thinks about what her does. It really doesn’t matter if they know Haskell and you write in Erlang. I doesn’t matter if they think in asm when you’re looking for Forth. I doesn’t matter if they are Linux nerds and you’re Windows-only, if they know how a computer works. As long as they think about what they do and know what’s that for and agree with your goals that’s perfectly fine. If you translate your desire for a good programmer into a bulletpoint list of tech words you dehumanize it. Don’t look for skillsets, look for people I say.

How to check whether someone is a good programmer then ? How to score people against such a vague job description ? You don’t score people I say, you just sort the set of people.
I start by a rather loose conversation. Tell me about the point in life you’re in, tell me something about what fascinates you in the programming job and finally what’s your motivation for changing the job and why do you want to work with us ? Somewhere in the middle I ask about the projects, what technologies you do like what not, why not. Do you feel comfortable teaching others ?

Then the talk gets momentum. We do focus on computers finally, why should it slow down, really ? If someone is fascinated by The Machines you should already noticed that, to be honest. Same with boring people or annoying people. If you do not want to work with them then do not hire them. Even if they’re brilliant.

How does the computer work then ? How does this happen that I have this text file full of strange writings and suddenly they make the computer do stuff ? Why it’s not a good idea to get the modern pc to fly a space shuttle ? What’s a stack and what’s a heap ? Is the stack memory really different from the heap memory ? Why so.

And the WHY. The most important question for me is the mighty why. Not only should they understand how something works, but more importantly why it works like that. I really enjoy working with people who are microwave-curious than the hex-feared ones. How does the microwave oven heat things ? Does the DEADBEEF scare your ? Maybe it needs some heat treatment then.

Talk with people, make notes. What annoys you, what impresses you. Do the exercises with them. Ask them how would they solve a simple but a real life problem. Design the solution with them. The more lifelike the work scenario the better. Do the things you’d normally do with your work colleague.
And order the set of candidates. Work as the simple max function. Take the first one and the second one,. Decide which one seems better suited for you. Then take the next one and compare her to the current max.

And please hire people only when you’re ready to hire them.

UPDATE.
After writing this post I’ve stumbled upon this 37signal’s art . Worth reading.

Categories: social | Tags: recruitment

Speeding up Eclipse/Aptana

Posted on 23 May 2011 by cyryl
2 comments

As some of you know Aptana is Eclipse. And Eclipse is a Java-based IDE which means it’s not particularly a speedy one. However if you have a decent amount of ram, like 4gb, it’s fairly easy to speed the Eclipse up.

Go ahead and find eclipse.ini or aptana.ini or other file which contents look similar.

These settings are from my Linux box, I know that MacOS might be kinda scared by so high values, try lowering the Xmx and/or others in such a case.

--launcher.XXMaxPermSize
512m
--launcher.defaultAction
openFile
-vmargs
-Xms128m
-Xmx2048m

view raw gistfile1.txt This Gist brought to you by GitHub.

These settings are for the 4gb ram box, try to find the ones which suit you best.
MaxPermSize stands for the maximum amount of the memory to be used by the Java internals, Xms gives the amount of heap allocated on the VM start and Xmx is the heap size limit. Start with upping Xms value as it’s often too small which causes the Java VM to make lots of heap resizes on the app start.

Categories: open source | Tags: aptana, eclipse, java, memory, settings, speed, speedup

Installing MacOSX Server 10.6 on VirtualBox

Posted on 7 May 2011 by cyryl
No comments

It all started with my work assignment of installing MacOSX Server 10.6 on the XServe rack. Of course nothing went really well. The universe keeps on picking up the most faulty scenarios for me, I think it’s just less boring. Briefly speaking, after debugging some hardware problem with the CPU temperature meter I managed to install the base OSX server system there. Hardware boy inside me is not amused with Apple’s ideas there.

After that I needed to set up VMs with 10.6 64bit and 10.4 32bit on top of that. Why so if already having 10.6 as the base system ? Simply I wanted to avoid constant reinstallation of the base system, as the machines are meant to be used for development and testing. Also the XServe hardware was needed instead of classic blade system to meet Apple license requirements.

Having the internet read twice I decided to give VirtualBox a try. My way of thinking was that I am already familliar with that piece of software and know it as a easy to use one. In addition to being pretty idiot-proof it also enables more enterprisely features like operating from the commandline only.

So what’s the deal ? Bring up VirtualBox GUI, click on new virtual machine creation button, choose MacOSXServer, insert the iso file made from the installation dvd and volia ? Not even remotely close. It appears that Apple has embedded some kind of valid processors db into the system and checks the processors present against it on every boot. Fine. But I am running this thing on the Apple hardware. The only configuration I managed to install and launch was single-core. Both 32 and 64bit guests seem to be working fine. But what’s the point of having many cores there if only one can be used per a VM ?

As with many other interesting problems I took it home. Some time after that I found brilliant blog pair by prasys and nawcom. It seems that the latter one produces what’s called ModCD which allows booting with the processor checks disabled.

Recipe for amd64:

  • create new MacOSX vm in VirtualBox
  • change it to non-EFI type
  • boot ModCD
  • swap iso image for the one with the MacOS install
  • press f5
  • type -force64
  • press enter

For me the installer crashed on the last step before reboot, but then it rebooted successfully and continued on with the installation.

Don’t forget to donate to nawcom if you can.

Categories: freedom, mac, server | Tags: apple, mac, macos, macosx, processor, server, virtualbox, xserve

How not to blog

Posted on 27 March 2011 by cyryl
1 comment

Today’s post is definitely not a tech one. Or maybe just a bit. It’s about blogging, from the other side of the keyboard. PEBKAC-style. Fun-fact: Did you know that ‘kac’ in Polish means ‘hangover’ ?

To the point. I realised that I have dozens of posts in the state of started but definitely not finished. Why is that so ? Is it that I am superbusy with my new work ? Kinda. Is that that I am the guy responsible for the installation of the sound system in one of the new Wrocław’s clubs ? Sounds pretty serious but still not a perfect excuse.

The thing is that there is no right excuse for not doing something you want to do. Do it or don’t. Don’t pretend to.

The funny thing is that the smallest possible posts, like this, brought me the most views. I don’t wanna loose you guys because I don’t feel like blogging recently.

Mine unfinished tech posts won’t probably make it because I just don’t remember what were these small tech tricks I used to solve that problems. Not even wrote them down as I was so enlightened by the superiority of my solutions.

Don’t do that. Take notes about everything which you would possibly need and start working on the sentences and wording. Rigth now. Go. Hit publish and let the world know.

Thanks go to TC & Mark Suster.

Categories: blog

Netbeans 6.9 for Rails 3 on Ubuntu 10.10

Posted on 29 January 2011 by cyryl
No comments

I am a lazy person. I do like things to just work, run out of the box, name it. I like Ubuntu for it’s packaging system and ease of installation of various software. However using Netbeans from the default repo could cause you a headache when developing Rails 3 apps. These just won’t run.

Let us start with installing the newest version of the IDE

sudo apt-get install netbeans
#then go to the Tools->Plugins->Available and install all regarding Ruby

Now go ahead and try running some Rails app.

/var/lib/gems/1.9.1/gems/activesupport-3.0.3/lib/active_support/dependencies.rb:239:in `require': /var/lib/gems/1.9.1/gems/activesupport-3.0.3/lib/active_support/cache/mem_cache_store.rb:32: invalid multibyte escape: /[\x00-\x20%\x7F-\xFF]/ (SyntaxError)

you say ? Here’s a quick fix

#edit /usr/share/netbeans/6.9/etc/netbeans.conf
#=> append -J-Druby.no.kcode=true to the the netbeans_default_options and volia
# the whole line im my case goes like that:
netbeans_default_options="-J-client -J-Xss2m -J-Xms32m -J-XX:PermSize=32m -J-XX:MaxPermSize=200m -J-Dapple.laf.useScreenMenuBar=true -J-Dsun.java2d.noddraw=true -J-Dsun.java2d.pmoffscreen=false -J-Druby.no.kcode=true"

Long term solution ? Wait for Netbeans 7.0 as the devs promised it to be fixed there

Categories: ruby | Tags: linux, maverick, netbeans, ruby, ubuntu
Previous Entries
© Adventurous Computing. Proudly Powered by WordPress | Nest Theme by YChong