AdrianBruce wrote:jmurphy wrote:
... It's interesting to see a technical person admit what I feel has been intuitively obvious all along -- Ancestry, in trying to provide hints and trees and Member Connect, and suggested records, and all this other gunk which doesn't pertain to searching record collections and retrieving records, has bitten off more than it can chew ...
Sorry - I don't follow your logic at all. Assuming I understand your description of the Site Alerts page correctly, this doesn't imply that Ancestry has bitten off more than it can chew. It's simply standard practice - "Tell the customers what's going on". Because things are never perfect in real life and every so often, even with the most robust site in the world, things will go awry.
"Tell the customers what is going on" should be standard practice, I agree -- but Ancestry has been particularly bad about it, if the comments I've seen on the Ancestry blog are anything to go by.
My remarks that Ancestry has bitten off more than it can chew are a non-IT person's way of saying that their current setup won't scale up any longer -- that is my interpretation of this part of the blog post:
As I mentioned in my last post, the amount of data that we host on the site, along with the complexity of the services we offer (trees, search, records, hinting, etc.), has led us to a point where we began to experience performance and stability issues with our site. In the past, we’ve been able to add hardware to keep up with needs, but now we have to make smart and purposeful changes to our infrastructure, which are taking time for us to execute. - See more at:
http://blogs.ancestry.com/ancestry/2013 ... Hnv1f.dpuf
As for your other point:
AdrianBruce wrote:As for your comment about accurate indexes, the choice is not between an accurately indexed City Directory collection and an inaccurately indexed one. The choice is between an inaccurately indexed one and (virtually) no directories at all. It is simply not viable to index all those directories manually, so OCR technology needs to be used. If the original documents are small and the original films taken at small magnification (neither of which are Ancestry's responsibility, I suspect) then you're on a loser straight away. Interestingly, my gut feeling is that later documents and films are better, implying the original paper and / or film are higher quality, as the OCR system is the same in both cases.
I understand why OCR technology needs to be used, and its limitations. What I don't like is that if I find a place where the OCR has not worked, and it has simply skipped several lines of entries on a page, there is no way to do anything about it. For one of the large commercial databases I use as part of my job, every detail page about a product has a link on the bottom of the page that says "Notify our database administration group about errors in the data". There is a way to submit feedback to Ancestry if you can't retrieve an image, but that's it.
The only solution they have is for people to submit corrections to the index lines which are already there.
If I had a service like that, I would want to have a means whereby errors found by users could be reported to a forum, like this one, properly tagged with place and surname and publication date, so that users who were working on those families or in those areas could search for alerts for records which an ordinary search won't find. That to me would be better than the scenario they have now, where one of my husband's relatives was
"corrected" to have her mother's maiden name in a census index, as if it were her own maiden name, and gross mistakes in indexing can't be addressed at all.
Edited to add: it's nice to see that this thread might end up being useful as a catch-all for "I'm having technical problems with Ancestry" posts where we can help each other. Maybe there are similar areas on Ancestry's forums, but the forum software is so primitive, I don't often go there.
