[okfn-help] open Milton

[Iain Emsley]

> Yes this seems exactly the way to go. We should probably start trying to 
> spec out exactly what is in the core and not so we can reduce 
> duplication and refactor as cleanly as possible.
>   
Given that we now provide packages, I think that we really only need the 
xapian indexing facility and the concordance since the metadata.txt file 
builds the db. The user no longer needs to use the gutenberg file. 
However the other files, like websetup and cache are used in the 
background.

A thought on the concordance which I think can be coped with in 
building. I think it might be useful if there was a document / 
readme.txt file on how to "adapt" the metadata file to allow the user to 
build their own concordance/search functions to index individual 
materials - unless its worth doing a command line version which takes in 
a text command during building?

Just wondering if an archive copy of the initial project should be kept 
before we redevelop the project.
>   
>> At that point all the user need do is drop the relvant packages in the  
>> data and
>> templates, run db create, db init and paste serve your-name.ini and  
>> one running project (well, you need to install it first)!
>>
>> I think that would also fulfill the componentization part of the Open  
>> Knowledge (though the ref currently escapes me - sorry).
>>
>> Just so you know I'm currently looking at the idea of using LinkedData  
>> [1] to provide machine interfaces if somebody does use this on a  
>> webserver. Okay, I also wanted to explore how the web of data/data web  
>>     
>
> Interesting way to go. Originally I was thinking that people would 
> simply install the shakespeare package themselves but more and more 
> stuff is becoming 'online services'. By the way if you do this it might 
> make sense to implement this in the shakespeare 'core' rather than 
> simply in Milton.
>   
I've got  a working copy of shakespeare here (I think  I need to 
download the new one though which you've been working on) which I'll use 
as a core for any development.
>   
>> can be used to serve Open Knowledge projects. I'm curious to see if it  
>> can lead somewhere meaningful. I was hoping that if we can  
>> find/develop a framework of ancillary texts/audio/video for  
>> consumption / generation. Any thoughts (apart from it could become  
>> Frankenstein's monster.)?
>>     
>
> I think developing a framework ab initio is hard. My suggestion would be 
> to start storing/referencing this stuff and then see where this leads us.
>   
Poor phrasing on my part (mainly from typing as thinking). I was 
thinking more of trying to put a page or set of pages which link into 
known open sources of material as a reference source which can be built on.
>   
>> Perhaps an IRC meeting would be useful, not just for this but also the  
>> Text Mining stuff that's begun to sprout. I think that there are a  
>> load of avenues that could be explored and it would be useful to  
>> create a public roadmap.
>>     
>
> Good idea. We can use the trac instance we have:
>
> http://knowledgeforge.net/shakespeare/trac/
>
> Current tickets are at:
>
> http://knowledgeforge.net/shakespeare/trac/report/1
>
>   
All best,

Iain
>
>
> ------------------------------
>
> _______________________________________________
> okfn-help mailing list
> okfn-help at lists.okfn.org
> http://lists.okfn.org/cgi-bin/mailman/listinfo/okfn-help
>
>
> End of okfn-help Digest, Vol 18, Issue 5
> ****************************************
>
> No virus found in this incoming message.
> Checked by AVG - http://www.avg.com 
> Version: 8.0.138 / Virus Database: 270.6.4/1615 - Release Date: 16/08/2008 07:11
>
>
>
>   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/okfn-help/attachments/20080823/3ba10faf/attachment.html>