Fortunately we already have three viable choices that are already in
windup so nothing would have had to be added new:
1) Lucene
2) Berkeley DB
3) Titan
But I don't think it really matters much at this point.
On 02/13/2015 11:40 PM, Ondrej Zizka wrote:
I get your point - why write it when there's a DB.
1) It's another 1.5 MB jar added to the app, introducing more classes to
scan on startup for Providers, Models, etc.
2) The data producing module would have to be rewritten to produce
whatever H2DB consumes.
Changing this would be just few lines change.
Ondra
PS: Don't get me wrong, I like all databases - my thesis was implemented
in SQL :)
And also, this is how I dealt with csv data in linux shell:
https://code.google.com/p/csv-cruncher/
Ondra
On 13.2.2015 05:38, Jess Sightler wrote:
> H2 DB may be a good choice as well, as it supports pluggable filesystems
> (including read-only zip format):
>
http://www.h2database.com/html/advanced.html#file_system
>
>
> On 02/12/2015 10:22 PM, Ondrej Zizka wrote:
>> 1) If the data in the file were put as binary, the size would reduce
>> another ~ 25 %.
>> The same if stored to the database (if we stored it there).
>> Is it worth making the file not much readable as plain text?
>>
>> 2) Doing the same in data structures would reduce memory consumption
>> around 45 %.
>> But since it looks we will use the the binary, it doesn't matter much now.
>>
>> Ondra
>> _______________________________________________
>> windup-dev mailing list
>> windup-dev(a)lists.jboss.org
>>
https://lists.jboss.org/mailman/listinfo/windup-dev
_______________________________________________
windup-dev mailing list
windup-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/windup-dev