seven

I just started to promote my side project[0] and I guess I am in a more or less comparable situation. I work alone and my time for this is limited.

First I focused on getting the product usable and that the documentation is somehow understandable with some code examples. My design skills are very limited, and since I build something for a technical audience, I though that the look of my website would not matter that much.

That was until Zach from headlinr[1] made me change the layout. Since I have a more or less decent looking website I get significantly more sign-ups and the logs show that visitors take more time to explore the content.

So with this new knowledge, I would vote for nice and shiny with minimal functionality first. Get something online as soon as possible. Just to get indexed etc. but do not underestimate the rejection caused by bad design.

[0] http://template2pdf.com/

[1] http://headlinr.com/

I don't run a startup, just a one man show.. but anyway:

I use sipgate.de for access to the public phone system. I connect to them via asterisk which connects to some other PBXes and my devices. Whenever somebody calls my german line (or my extension at an other PBX), the call ends up in my asterisk. I try to ring my mobile first. If this is unsuccessful, I tell the caller via recording that I will try some other devices. If that fails, the caller can leave a voice mail and I get a notification via email. (xmpp soon..) My setup is still missing caller-id filtering fun, but it was easy to build and runs stable since a few month. If you have a server somewhere, perhaps a simple asterisk with a sip provider is all you need. I can't wait until I find a usable sip client that supports zrtp on my android. I have high hopes for jitsi. (https://jitsi.org/)

Just indicate your role in those projects.

Imho you should switch 1) and 4) :)

Is there any chance that you could freelance for your current company.. or at least work half time?

About your dream project: Try to get steady validation from the beginning. Find your first user first. I just started a little side project and without encouragement and a tiny amount of power users, I would have stopped before I really started. Marketing to developers is incredible hard (at least for me). Prepare for frustration.

Getting into the right flow is one problem I face myself on a regular base. What helps me a lot is to just relax a bit (days!) and trust myself to do the coding somewhere in my brain as background job. After being afk for some time I can't wait to get back and make somehow progress. Just do your job for another two weeks and really use your free time to get some rest and have some fun. Perhaps your outlook changes a bit after that.

What was the result from going to startup meetings? Never been to one, planing to visit some meetings in my area, but no clue what to expect.

Take a look at: http://argus.tcp4me.com/ It easy to use and quick to configure. Polls via snmp and stores in rrds. At first glance I did not like it.. but imho it beats other products in terms of simplicity.

If you are looking for something bigger, then zabbix (http://www.zabbix.com/) might be an option. It is not that easy to get running, but has capabilities like dependency based notifications. For example: "Only alarm me of missing ping responses if the router of that box is reachable." I think it is a fantastic piece of software although I had to build an argus-like monitoring interface for it, so that less trained staff is able to work with it.

Typo is fixed. Yes, my design skills are very limited. In case more people find the service useful I will for sure get somebody to make the page look nicer.

Thanks for your comment.

Let me make this clickable: http://a.fancyresu.me

Really fancy! :)

I just started to work a bit more on my side project Template2pdf (http://template2pdf.com/). It is an API to enable developers to produce nice PDFs out of templates that their customers can modify on their own.

It runs on Debian with ruby/rails, java, postgresql and some system tools.

I also do some work on a SAAS to build wireless hotspot systems. I don't own anything of it, but is an interesting field once you hit scalability problems. We use all kinds of technology but the core components would be linux/freebsd, freeradius, postgresql and ruby.

Yes, I read that somewhere as well. It is still a bad tactic.

At first you would want to encrypt the information. Then you would like to have some method of transfer that does not stick out and could hide the intent to transfer encrypted data. Perhaps with the help of some steganography tool. And you would make sure that nobody identifies you in the process by using tools like Tor. This "I hope they don't notice" method is just gambling.

"Escola Dom Bosco, Luanda, Angola" will show you the location of the school in google maps.

Which company are you referring to? There were some companies involved in putting together the financing for this charitable cause. This school is not searching for employees. If you want details, please ask for them.

I did not see that while proof reading. Neither the author nor me are native speakers. Google translates the single words into something that is understandable and would make sense. For Germans at least. :)

I guess the right words would be: pocket money

@all: Thanks for giving this topic a bit attention. In case you have more questions about the offer, let me know. I will find out. Also happy to answer questions about working in Luanda. Perhaps you would like to stay after the 3 month. :)

It does not.

Teaching python requires some experience.

In theory malicious software could do that in some cases.

Lets take videos as an example. If you are using a media player with an error that could be exploited to execute program code that is embedded inside a prepared (read: modified by the malicious software) video file, this could be one vector of re-infection.

Teach them how to use Tor.

https://www.torproject.org/

If you use pam_cracklib[1] or disable password based login for ssh, the only worry left over is working log rotation.

[0] http://www.deer-run.com/~hal/sysadmin/pam_cracklib.html

[1] http://lani78.wordpress.com/2008/08/08/generate-a-ssh-key-an...

You are right about "end-to-end verifiable". I misunderstood that.

About understanding what is happening, I have to say: No, I do not know every detail on how we do it in Germany. But I would say that I have a pretty good overview and trust in the system. We had minor election related scandals, but never on a larger scale. Everything happens under the eyes of many. Those many consist of volunteer. When I go to vote, I know some of them personally. If somebody does not know how our system works, it is at least possible for him/her to find out and fully understand. There are some complains about our system, but none are about the risk of large scale fraud.

I don't know much about on how you do it in the US. I heard that it is more complicated than ours, and that using a computer would help a lot. But as I said, I would prefer to get the job done manually. It is possible. It worked before.

About online shopping. It is a trade-off.

I know I should not trust my computer. I know the current state of how we use ssl and our trust in different CAs is pretty bad. But I am still doing it, since I hate to go shopping in real live. For me it would be much more risky to get into my car and drive to a shop. I would not want to trust my computer and the current state of how we use crypto with everything in any case.

I was more about bribing people to make mistakes during the implementation. And do you know for sure that there is not yet a quantum computer in some kids basement? :)

In some cases like "xor" with a random key, you can proof that the crypto is unbreakable. But the way people use it is often faulty. There are certain requirements that apply to the key and without them, the method is not secure any more. In other cases like RSA, there is no real proof that it is working. It is hard to proof that nobody will figure out an highly efficient way to do prime factorization. All we know is that many people tried and failed.

Public inspection of those computers is not really possible. How do you trust the hardware? It is more easy to show that something does something. But it is hard to proof that something is not doing something. We would use computers to do the calculations that prove that our votes are ok.. so we need to trust those machines again..

The beauty of pen and paper is that every person is able to fully understand what is happening. Additionally you do not need to have a strong trust in the implementation of the cryptography or someone discovering a flaw in it.

Large scale voting fraud on pen and paper systems is probably very hard, since you need to bribe a lot of people.

For a large scale attack on computer based voting, you would need to bribe way less people.

If you can proof to yourself after the election that your correct vote went into the system, you can proof this to other people too. "end-to-end verifiable" sounds good at the beginning but makes buying votes possible.

F*ups are just more likely while doing computer based voting.

While paper and pen voting is expensive, it is for sure something I am willing to pay for.

Take a look at https://www.owasp.org/ and review your source code for common vulnerabilities.

A very helpful tool for testing: http://portswigger.net/burp/proxy.html

Went from Trac to Redmine for code centric projects and just started to use Teambox for projects that involve non-coders. My main requirements are: Option to host it myself and the ability to read and modify the code.

Call the company via phone. Most Germans speak English, so that should work.

You might want to take a look at http://gpl-violations.org/ the guy who runs it is also a German.

I was a bit bored during this weekend and created "Bored Bookworm". http://www.bored-bookworm.com/ It has not yet generated any money, but my brother promised me to buy some book through it. :)

Hi.

I'd like to get some early feedback about my mini-project.

http://datajones.com/ - mobile data collection

Use Datajones to create custom forms, to collect data via mobile phone or tablet, even while being offline. Export the collected data as csv file for further processing in your spreadsheet application.

It is very early beta, but already functional without any major known bugs. Tested on Firefox, Chrome and Opera. Since I do not have a modern phone, I was only able to verify functionality on a friends HTC Android. btw: This friend happens to be my first customer. :)

Love to hear your thoughts, ideas for improvement etc.

Many thanks!

Some recommendations for connectivity, that helped or would have helped me during my trips in Africa:

Setup some sort of vpn service that you can use in case of restricted internet access.

If you run your own mailserver, open up a port beside 25, since some provider will block this one.

An ip over dns setup is something you should consider as well.

Make backups and use encryption.

Have fun!

I am struggling to give a good answer to this since some minutes...

There is no special trick. I just realized that it makes no sense to worry about stuff I can not influence. And that if I have bad mood without obvious reason.. I could have good mood without obvious reason as well.

I have no step by step instruction. It is just a decision you have to make for yourself.

Focus on the good things instead on the bad. No matter how big you fuck up, chances are very high that you are still better off than most people on this planet.

default bad mood => default good mood

pessimistic => optimistic

Off topic: Since I will move to Luanda/Angola in a few weeks, I would be happy to find some hackers from that area. Please contact me. You will find my mail address in my profile.

They do grow back. Just destroy the chip, then there is no way to read them out again.

This is not true for Germany. (Which makes this even more a security theater..)

Or build a RIFD_Zapper:

http://en.wikipedia.org/wiki/RFID_Zapper