Saturday, September 28, 2013


I am working in a project in which every participant tends to over-engineer its part.
Most of the guys are quite young and they want to impress with their design pattern skills - which is not bad at all if they would choose the correct DP for the problem tackled - but somehow they miss the goal of some DP and apply them in every possible situation only to have the usage ticked.

Then there are some expert developers or ex-engineers that are now trying to apply the dev-ops paradigm. So they are always trying to create something that CFEngine and Puppet or Chef are already doing. It literally paralyzes my mind when I assist to some discussions about reinventing the wheel. They are - with all respect to their development skills - illiterate sysadmins and they do not want to read about existing proven working solutions and try to resolve everything by  themselves.

That's not okay in my opinion because there are so many corner cases that one can miss with an in-house solution. Any of the above tools does a better job on the long term than a custom made solution by the fact that it is extensible and there is documentation about it - while the in-house is quite a black box brainchild of a excellent dev/terrible sysadmin.

Friday, September 27, 2013


Best logo:

DroidCon Romania -


I am generally against organized football. I like it as a sport but not as a business. So I am extremely proud and happy that an amateur team from the Birsana village of my native region, met the en-titre champion team of Romania played excellent. That's the beauty of the game!

Going slightly mad

Emperor's New Clothes

In a software project, if some of people are misleading others because they seem trustful due to their position and/or their seniority, the project will go in a bad direction and legitimate critical thinking will be discouraged. 

What would one expect from a project in that people are misleaded? Useless effort and pain...

Be fair, be critical, be realistic - do not accept "Emperor's New Clothes" as a fact.

Wednesday, September 25, 2013

Rosia Montana

I do not know much about Rosia Montana... But I have seen the scorched earth of Rosia Poieni...
Gold mining is a short term solution for inhabitants... 10 years, 20 years... but after? How will Rosia look? Will that place - unchanged from Roman Empire era look the same... Will it be able to sustain life?

I do not know...

But I know that children are born in this country and they deserve a future. As much as we deserve a decent present.  

Wednesday, September 18, 2013


1. questioning, probing, testing.
2. disbelief, atheism, agnosticism.
One of the things that really annoy me is the dependency of people on frameworks and things they know in software.
I am mostly questioning people on why do we need new buzzwords in projects, what value they bring and why precisely those buzzwords and no other. I am amazed that they always answer by the book, not deliberately, but rather they repeat the hype...

"We need Spring" (although we have no beans but a SPA)...
 "Err... because of IoC :)"
 "But why Spring with Vaadin and not Guice or other?"
 "Err... Spring is good because it gives you beans and Guice is from Google hence is bad"
"Really... I was under the impression that it is POJO based"
"No, beans, beans are good."
"Why are beans good?"
"Because they have interfaces"
 --No Comment - Discussion with a Scrum Master with over 15 years of experience

 "DEB packages are a solution for our deployment problems."
"What problem would they solve?"
"I am spending too much time configuring the system at deployment"
"And how will the DEBs help you?"
"Err... Reduce time to deploy?"
"In fact you need a DEVOPS solution as Chef or Puppet, no?"
"No, those are not ok."
"I heard about them that are bad."
"With them you can deploy directly our WAR files on thousands systems at once on different configs without scripting... Isn't is a cool feature? Especially if we manage to sell and deploy our system."
"Err... Nooo, too much overhead"
"Well what if we will deploy on RedHat? They use RPM, you know?"
"Never in RH... It is evil..."
 -- No comment - Discussion with an Architect

Sincerely... The arguments are weak and based on marketing hype, rumours or hunches... But those are basically people that take decisions on software projects - so there is no surprise on the quality of software.

Saturday, September 7, 2013

20 Years since Assembly'93

The source of 2nd Reality demo (which run on in i386) PC has been published.

I have spent half of the high school trying to imitate this magnificent demo.

Java apocalypse



In my recent vacation I have been close to this tormented region.
Here is a movie about what happened there... Food for thought.