Evopedia 0.2.1

Christian Reitwießner christian at reitwiessner.de
Wed Aug 12 10:09:51 CEST 2009


c_c schrieb:
> Hi,
> 
> Christian Reitwießner wrote:
>> unfortunately without any feedback, I have now released evopedia 0.2.1,
>> which fixes a single bug: the "near articles" feature should work now,
>> if it did not already. 
>>
>  Well, I saw your site and realised that the English dump is 7 GB. I have a
> 8 GB card - but this would leave me with almost no space for anything else.
> I wonder if there was a way of reducing the image sizes further by splitting
> them into topics (maybe science / computers / physics etc).
>  
>  Otherwise I'll just have to buy another 8GB card - but the card swapping is
> a little cumbersome. What do you think?

Please note that unfortunately, the English dump is for Evopedia 1.0
only. It is not much more than a squashfs-compressed version of
http://static.wikipedia.org/downloads/2008-06/en/wikipedia-en-html.tar.7z
(which is 14 GB and 7-zip compressed), where the special pages are
removed. There is some redundancy in the html files (for example there
is always the same header) but I thought that squashfs takes care of it.

So the bottom line is: I don't know if the size can be reduced
significantly. For Evopedia 2.0 I do the dumps myself (mainly because
those on static.wikipedia.org are more than a year old), but it takes
really long. At the moment I'm more trying to reduce the dump time and
not the image size. I have not tried to do a dump of the English
Wikipedia at all because it would take weeks.

If you look at the dump sizes for the German Wikipedia, the sizes are
1.8 GB (2008, Squashfs 3), 1.9 (2008, Squashfs 4), 2.5 (2009, Squashfs
4). I really don't know what caused this.

I can try to do the dump for the "simple English" Wikipedia next.

Kind regards,
Christan



More information about the community mailing list