> Are you allowed to tell us more about the company? Number of employees?
We were around 15 people ( sales, design not included )
> It sounded like you got hired to do TCL without knowing TCL.
Indeed, in their job description, they just wrote that I had to be curious and willing to learn something unusual.
They didn't mention TCL.
> Why was it a secret weapon?
I would say flexibilty, it was quite impressive to watch how his very own web framework evolved.
He would check out new frameworks over night, if he liked some patterns/features etc. from other frameworks, he would just put it into his own framework over night.
The most impressive part, was probably the loose coupling of all the software components.
There wasn't some kind of an ORM, just pure handcrafted optimized SQL.
The secret weapon was, how he ignored other opinions.
Just a few examples, for some people OOP is the way to go, anything else is just unacceptable.
They are forcing themself into a corner with limited options.
I agree, OOP is awesome but it's not the solution for all our problems, in TCL we just could switch from OOP to imperative or to functional without breaking anything.
As already described above, plain SQL is already the best DSL for databases, we didn't have any abstraction layer on top of that, which we had to learn.
I don't agree with all his opinions but I do agree on the fact, that we software engineers tend to over engineer things.
I used to work for a company which heavily used TCL[0].
When I started working there, I really questioned his decisions about the stack and the language in general ( everything is a string )
After a few weeks I realized that, this stack was blazing fast in combination with the NAVI[1] server which is written for high performance web sites.
This guy wrote everything you could possibly think of in TCL.
That company is powering a ton of enterprise grade e-commerce web sites.
I still see it as a secret weapon and I'd probably use it as well if there wasn't my fetish for statically typed languages.
edit:
As a side note, I'd like to mention that this guy taught me that I shouldn't go with the hype but just choose a language/framework that helps me finishing my projects.
I'm a Java and a C# dev, Scala when I feel adventures or when I'm drunk.
I do all my work in IntelliJ and Visual Studio but I found a
few useful tools like VsVim which emulates all the keys in VS, there is also the IntelliJ plugin which brings Vim into the IDE.
Then you got the best of both worlds, I have to admit it's not the same as running Vim from a terminal but it's very close.
Other than that, if I had to write something in a dynamic language, I'd still prefer Vim with a few useful plugins.
I really tried hard to emulate an IDE inside Vim but that was a lost cause.
Vim with some todo plugins, Emacs orgmode, but there are also some lists which I need daily on mobile, for that purpose I use Wunderlist with abbreviated descriptions ( Yea I know I'm paranoid, but hey it's in the cloud and Microsoft owns it )
We were around 15 people ( sales, design not included )
> It sounded like you got hired to do TCL without knowing TCL.
Indeed, in their job description, they just wrote that I had to be curious and willing to learn something unusual. They didn't mention TCL.
> Why was it a secret weapon?
I would say flexibilty, it was quite impressive to watch how his very own web framework evolved. He would check out new frameworks over night, if he liked some patterns/features etc. from other frameworks, he would just put it into his own framework over night.
The most impressive part, was probably the loose coupling of all the software components.
There wasn't some kind of an ORM, just pure handcrafted optimized SQL.
The secret weapon was, how he ignored other opinions.
Just a few examples, for some people OOP is the way to go, anything else is just unacceptable. They are forcing themself into a corner with limited options. I agree, OOP is awesome but it's not the solution for all our problems, in TCL we just could switch from OOP to imperative or to functional without breaking anything.
As already described above, plain SQL is already the best DSL for databases, we didn't have any abstraction layer on top of that, which we had to learn.
I don't agree with all his opinions but I do agree on the fact, that we software engineers tend to over engineer things.