Lisa,

with the information you provided, I would need to build a complete script to try to reproduce the problem. I don't have time for that. Please provide a complete - yet minimal - webtest script that was working with older WebTest build but now fails with recent ones.

Cheers,
Marc.
--
Blog: http://mguillem.wordpress.com


Le 15/10/2010 19:34, Lisa Crispin a écrit :
Hi Marc,
I posted a link to the page with the problematic button on our public
demo site, along with the WebTest step, but if that doesn't work for
you, I will try to get one of my teammates to code something I could
send. I'm pretty sure it's HtmlUnit because one of my teammates ran into
the same problem with Selenium 2.0 which also uses HtmlUnit.
thanks
Lisa

On Fri, Oct 15, 2010 at 10:40 AM, Marc Guillemot <[email protected]
<mailto:[email protected]>> wrote:

    Lisa,

    you don't need to post to HtmlUnit mailing list: I'm HtmlUnit
    committer too.

    Concerning the regression you encounter, the information you
    provided wasn't enough to have an idea of the problem. Would you
    provide a simple webtest script working with an older WebTest build
    but not with the latest one, then I would have a look at it.
    Otherwise it simply takes me too much time to try to reproduce your
    problem and other well described bugs are waiting too ;-)

    Cheers,
    Marc.
    --
    HtmlUnit support & consulting from the source
    Blog: http://mguillem.wordpress.com




    On 10/14/2010 07:06 PM, Lisa Crispin wrote:

        We've used WebTest for almost 7 years. In general the tests are
        quite
        stable. There's one test that over the years randomly failed and we
        couldn't figure out why. I mean, this went on for years. Finally, we
        discovered the underlying bug, and it was a bad one, it was too
        bad we
        had not investigated the random test failure more thoroughly.
        WebTest
        was trying to tell us something, and we ignored it.

        That said, sometimes we have backward compatibility issues so it
        can be
        a bit unstable in that way. About a month ago we tried to
        upgrade to the
        latest release, but it (probably it was really HtmlUnit) stopped
        being
        able to click some of our buttons that have Javascript. I asked
        about it
        on this list and never got any help with it, so we just stayed
        at the
        old version. That bugs me, but I don't know what to do about it,
        I asked
        if we should submit a bug to HtmlUnit (I don' follow their list or
        anything) but didn't get an answer.
        -- Lisa

        On Thu, Oct 14, 2010 at 2:51 AM, mascis <[email protected]
        <mailto:[email protected]>
        <mailto:[email protected] <mailto:[email protected]>>>
        wrote:


            Hi,

            This is not a question for a single problem but rather a
        opening of a
            discussion.
            I have been working with Canoo Webtest for a half of a year now.
            Everything
            has worked quite well and these automated tests have been a
        great
            help. The
            only bigger problem is that Webtest don't seem to be very
        stable. What I
            mean is that the results that Webtest change sometimes even if
            nothing has
            changed in the web page I test.
            Sometimes Webtest loses cookies used for logging, sometimes
        it gives
            strange
            JS errors and next time I rune the same test it passes,
        sometimes it
            gives
            error for not founding a certain web page even if resulting
        page shows
            correct page etc.
            We use Bamboo to run tests in certain times every day and
        for some
            reason
            this seems to make things harder. Bamboo makes tests run a
        bit slower so
            chance of missing logging cookie is bigger and for some
        reason those
            strange
            JS errors (
        http://old.nabble.com/A-strange-Java-Script-error-to29894316.html
            this  for
            example) occur more often.
            This is quite annoying since I can't rely that once I make
        test pass
            it will
            pass every time I run it. It also takes my time to go trough
        reports and
            look if there is real errors or something that Webtest just
        gives
            without
            anything being broken.
            We have a lot of JS and AJAX etc. "advanced" things in our
        web page.
            I don't
            know if that makes those errors appear or what. Does anyone else
            have same
            kind of things or is it just me?

            --
            View this message in context:
        http://old.nabble.com/Webtest-and-stability-tp29960485p29960485.html
            Sent from the WebTest mailing list archive at Nabble.com.

            _______________________________________________
            WebTest mailing list
        [email protected] <mailto:[email protected]>
        <mailto:[email protected] <mailto:[email protected]>>

        http://lists.canoo.com/mailman/listinfo/webtest




        --
        Lisa Crispin
        Co-author with Janet Gregory, _Agile Testing: A Practical Guide for
        Testers and Agile Teams_ (Addison-Wesley 2009)
        Contributor to _Beautiful Testing_ (O'Reilly 2009)
        http://lisacrispin.com
        @lisacrispin on Twitter



    _______________________________________________
    WebTest mailing list
    [email protected] <mailto:[email protected]>
    http://lists.canoo.com/mailman/listinfo/webtest




--
Lisa Crispin
Co-author with Janet Gregory, _Agile Testing: A Practical Guide for
Testers and Agile Teams_ (Addison-Wesley 2009)
Contributor to _Beautiful Testing_ (O'Reilly 2009)
http://lisacrispin.com
@lisacrispin on Twitter



_______________________________________________
WebTest mailing list
[email protected]
http://lists.canoo.com/mailman/listinfo/webtest

Reply via email to