Skip to content

Very low success rate

edited February 2013 in Need Help
Hello, I have strange problem with GSA SER.

I scrape sites by myself using sb, and after I scrape them, I'm adding all to SER by Import URLs (identify...
After adding, I'm obviously trying to test the list and so, i'm creating project and select only social bookmarks.

And my problem is that with db of Hotaru CMS circa 2.5k, phpdug ~800 albo pligg 7818
i get only 22 verified links!

this is waaaay below acceptable level. Could anyone help with that?

(i use captcha sniper for captcha solving)

Comments

  • SvenSven www.GSA-Online.de
    Post some logs please else we can not help you.
  • what log do you wish to get?
  • SvenSven www.GSA-Online.de
    The one in the log window.
  • how you access log window? cuz I can't locate it?
  • the bigger box on SER
  • ok here it goes, I hope you wanted this one: https://www.dropbox.com/s/vv20z5neft780xp/gsa.log
  • "http//www.link.here" <-- what's that and where its coming from or did you edit the log before uploading?

    apart from that i see all kind of errors
    - captcha answer was wrong/not solvable
    - no form at all
    - missing form fields on the URL

    please visit some sites that are marked with "unable to find suitable URL" in your log with your browser and check if you can register and upload content to them. if thats the case than send Sven some example URLs so he is able to fix each engine.
  • yeah, I edited it.

    sure thing that there are some failed links but why do they get in the db if ser should filter them?
    still, i get 10k+ lines in 45k+ long log giving 'matches engine (engine name here)' and still so low success.
  • OzzOzz
    edited February 2013
    well, SER just identified a site as Hotaru, but it doesn't know if its open for registration for example.
    SER only can tell you that if its actually try to create an account on that particular site.

    apart from that a reason that SER isn't able to create accounts on this site is simply that the script of SER don't know how to handle that site because it was modified by the administrator (with fields that are unknown to SER but necessary to create an account) or the engine was updated globally by the platform provider (like Hotaru) so it needs to be fixed by Sven.
    if thats the case than send him some examples where you could register and post to by hand (with your browser).
  • edited February 2013
    Yeah sure, but that's not what I wanted to ask, because when I add urls from file it adds sometimes (not rarely) links that doesn't even have ANY cms on them.

    nevertheless, after filtering i'm capable of adding 10k link do SER, and sure, you're right that never 100% of db will work but success of 24/10000 is a bit low
Sign In or Register to comment.