Comments for post One year of Redis
http://www.artimplant.ro writes: I believe the real success can be determined by the subject of your blog / website. If you offer quality, you will have just quality visitors, instead of low quality one. And this seems to be the most important reason to offer quality.
Mario P. writes: Right. There is a chasm between "underground" security and professional security.
antirez writes: @Mario P. security at some point was no longer fun. It started to become a product, and there were no longer great new attacks to discover. Even worse, a security research to live have to do consultancy, and security consultancy were (are?) particularly boring: go in a big corp and try to figure how their network works to add layers of security, firewalls, proxies, ..., or on the other side source code auditing that is even worse :)
Mario P. writes: > I stopped the development of Hping when I quitted security...
Why you have quitted security?
antirez writes: @Justin: I'm now using Ruby instead of Tcl to write the same kind of code I used to write with Tcl mostly. I still use Tcl for "scripting" instead of bash or to write trow away networking code, as I think Tcl is still more practical in many ways for this problem domain.
The reason I quit Tcl is because the TCT (Tcl Core Team) has an idea about how to handle/evolve the language that are in my opinion not compatible to the need of evolution of Tcl. Compatibility with the past is taken in great regard even when changes to break with the past are absolutely needed (for instance Arrays should be first calss objects).
Also numerous design errors were made in the fields of namespaces, I/O, and so forth.
This wrong development path also lead to a very impoverished community, this means less libs, less documentation, less everything.
Basically Tcl needed his benevolent dictator with an unique, consistent, modern vision. The TCT can't work this way for a number of reasons, so Tcl is a language without hopes IMHO.
When I realized this I looked around me and started trying other dynamic languages, and I picked what I think it's the best, Ruby. I loved SmallTalk and Lisp and I found part of both, and a consistent simple design, in Ruby.
Justin writes: What made you quit TCL? Seeing your past projects (Jim, picol) you seemed very fond of it, and it made me check it out for myself. Did you ditch it for another language, or do you just not write the kind of code that you used to use TCL for anymore?
antirez writes: @all: thank you very much for the nice words!
@Peter: example, I want to build an appliance that runs a web application. I sell this appliance that contains Redis binaries internally. I think this qualifies to require source code distribution for the GPL license, while I consider this internal use.
Also I don't like the dual licensing strategy, it's something like the proof that GPL code is not really *free*. Just a point of view btw.
Peter writes: Your comments about BSD is based on a misinterpretation of the GPL. The GPL does not require you to release anything externally. All that it requires is that IF you release binaries externally, THEN you must release source. Anything you do in the privacy of your company can stay within the company. It is only once you decide to release binaries externally that the GPL kicks in.
The GPL is perfectly fine for internal development. This isn't ambiguous in the GPL -- what I stated is the interpretation of the GPL that the FSF subscribes to, as well as lawyers at every major Internet company. Google does a ton of internal development on a GPL codebase for internal use without releasing changes. Most of the other major Linux-based Internet companies do the same thing.
I favor GPL for projects like this for a number of reasons. The most substantial one is that it does give you the option of dual-licensing later, which can make a nice profit stream eventually. I do like "GPLv2 or later" or same for LGPL licenses, to avoid offending both the GPLv2 die-hards and the GPLv3 die-hards.
Note: I am not a lawyer, but I have studied technology law under a number of people, including Lessig, and am friends with RMS, know Eben Moglen.
Alex Popescu writes: Add to all this advise, one specific to NoSQL projects: (seamless) integration with existing frameworks. The presence around major frameworks will encourage people to try it out, which will result in new use cases and that would finally lead to more and easier adoption. I have posted a longer article on this subject just a few days back http://nosql.mypopescu.com/post/299775877/nosql-to-people. Right now, according to the data I can gather on MyNoSQL (http://nosql.mypopescu.com), it looks like Redis is one of the most actively tried NoSQL projects. So, keep it going!
:- alex
paulinohuerta writes: Redis arrives when RDBMS'es are not going anywhere.
How it arrives?. With ANSI C, a very simple installation (the first goal).
To have 16 GB of RAM in servers it'is quite common, in addition moderm web apps have created new problems domains in data storage and querying.
Redis is a alternative noexistent before.(IMHO second goal).
Part of Redis future is basically in efforts by disseminate key-value concepts
Non-relational stores are not relational
The key/value store not knows the abstraction on table, rows, ..
Redis is a evolution in the key-value store. To get data structures "shared and persistent" from my apps
adding only the network layer, it's simple, too much simple but, I don't never seen before.
List push operation with O(1) complexity, it's simple, that's is your conceptual niche.
Some Rules:
Don't try to implement everything that the MySQL driver implements.
Use the strengths of alternative store.
Nice work on Redis, Congratulations on the first year
and Thanks
Diego Sana writes: Congratulations on the first year and thanks for doing it. I think you should include one last paragraph: "Talk a lot and show commitment to the project". When i first heard of redis, i did a quick read in the page and followed you on Twitter, but didn't started using it. Then i saw you tweeting so much and so passionately about it that i decided to give it a try... and now i'm in love with Redis, which is the 2nd stuff from Italy i most like (1st is Pizza :D)
Kashif Razzaqui writes: Excellent work with Redis, I prefer it to Memcached :)
Andres writes: started looking at redis recently, great job!
Roman Vorushin writes: Redis is excellent open-source project! Thank you for your work and great write up!
András Bártházi writes: Thanks for your project, keep the speed of the development, and I believe that it will be one of the best known project of the year 2010.
antirez writes: @KS Sreeram: thx, fixed!
KS Sreeram writes: The link to Redis at the beginning of the article is incorrectly pointing to:
http://code.google.com/redis
home