On 29 Jan 2009 10:41:21 GMT, Lucifer wrote:
>"mr.bill" <mrbill@invalid.ls> wrote in news:4980f1f5$0$31165$450c70f1
>@news.privacy.li:
>
>> On Wed, 28 Jan 2009 04:36:52 +0000, ridesaway@bloodymorningafter.org
>> wrote,
>>
>>>There evidently exists a CSV for this site of which I am unaware.
>>>I have "Home Lolita-Charming Angels-Chara Complete_mb_16359.csv",
>>>which indicates that I am complete, yet I do not have most of the
>>>files which you both seek. What CSV should I be using?
>>>
>>>--
>>>ots
>>
>> Bourne's csv differs in a few fundamental ways from CD's csv. It shows
>> some updated files and some authentic files not on CD's list. I'll be
>> releasing a new csv myself once I've processed what Servic posted.
>>
>> --
>> mr.bill
>>
>
>
>
> Thank you again mr bill, i look forward to seeing
> your re-released csv to make sure i'm complete on
> that site.
>
> i've noticed some little differences in
> a number of site csv's, but N@@dles and yourself
> seem to have the most reliable ones, CD did a great
> job on alot of sites, but there are little bug-a-boos
> in some csv's when you start comparing them to other
> peoples site rips and csv's.
These differences are what makes this work so challenging, yet rewarding.
There are darn good reasons why every entry made it onto a csv. Finding
files to match each entry can be tough, but once competing files are found,
it is generally a no-brainer to pick out the bad one, or to show there was
a reissue. Developing explanations for the errors is sometimes easy too,
for example, an image gets issued on its side, and the resave was
rectified.
For just the case of Charming Angels, I compared the work of bourne and
Count Draculol. CD's work was on-going, so he had indexed just 13,850
files. 81 of his entries were of bad or missing files. The situation was
actually a bit worse since 7 files got renumbered, but I disregarded this
since they still belonged, albeit under a different name. For the sets he
had indexed, he achieved an accuracy of 99.42%.
Bourne indexed 16,258 files of which none were incorrect if you allow the
two swapped images were a simple rename. However, he failed to index
chara-05 and tossed out a misplaced image, hl-32-113. It can be argued
tossing out the one image that was in a set where it didn't belong was
okay, but I prefer to keep these strays that LS occasionally issued.
Taking into account the missing 103 files, bourne achieved an accuracy of
99.37%.
These guys both did phenomenal work by anyone's standards. And don't
forget, one cannot index what one doesn't have access to.
> Thanx again for all your time & hard work!
>
>
> LCF
Lucifer,
I'll attach my latest hl/ca/chara csv for you & the group. I've begun
dating my work inside the csv just after my sig. It should help sorting
out the latest csv as this work is demanding constant changes as more amd
more comes to light.
--
mr.bill
|
|