Serendipity Forums

Problem with Spam Protector plugin: One of my commenters could not comment in Firefox but could in IE.

I am using Firefox 2.0.0.11 with cookies enabled.

I then tried with Internet Explorer 7.0.6000.16575, and the comment went through at once.

It sounds as if it is a display problem.

Tried with the word "Test"

6 failures in a row, with the strings:
4HH
2FH
39B
CMY
ABEMV
4BFKM

In each case it was clear where to enter the string, but entering it and clicking "Submit comment" simply returned me to the same screen containing the error message:

You did not enter the correct string displayed in the spam-prevention image box. Please look at the image and enter the values displayed there.
Your comment could not be added because comments for this entry have either been disabled, you entered invalid data, or your comment was caught by anti-spam measurements.

(I like that term "anti-spam measurements"!!!)

Can anyone help? At the same time, is there any way simply to moderate comments without captcha?

(The 'anti-spam measurements' is a bit funny for English speakers because it should be 'measures').

TIA

Margaret

Btw, here is the entry in question (the first comment was entered by me on behalf of Victor):
http://www.transblawg.eu/index.php?/arc ... .html#comm

Margaret

Hi!

If it worked in his IE, it must be a browser problem. Either cookies are not enabled, or the HTTP Referrer is not send. Tell the visitor to try again by running firefox in safemode, without any extensions. I'm quite sure an extension or a config option of his browser is preventing access to cookies.

Regards,
Garvin

Thanks, I'll pass this on.
The thing is that I have not been using this feature on recent comments long enough to get any other feedback. I had some spam so I decided to change the settings.

Margaret

Feedback as follows:

Can't find a function "safemode".
I tried manually disabling the add-ons, but that did not seem to help.

I checked the FF "Error Console", which says:

Unknown property 'behavior'. Declaration dropped.
http://www.transblawg.eu/index.php?/serendipity.css Line: 2

This 2-line error message (without any further explanation) appears for each failed attempt.

I wonder if the developer can make sense of the message.
Until he (or I) finds a solution, I'll just have to switch to IE every time I want to comment.

I found the following in the file in question (Seitenquelltext):

img {
behavior: url("http://www.transblawg.eu/index.php?/plu ... havior.htc");
}
/* templates/bulletproof/style.css */
/**********************************************************************/
/* Template: bulletproof for Serendipity (http://www.s9y.org) */
/* Authors: Matthias Mees, David Cummins and Don Chambers */
/* */
/* Central Stylesheet */
/* */
/* http://s9y-bulletproof.com */
/**********************************************************************/<<

TIA for advice,

Margaret

Hi!

That "behaviour" stuff is unimportant, it's only a CSS issue which can be disregarded.

SafeMode for Firefox can be found in the documentation for firefox, can't look it up right now, sorry.

Regards,
Garvin

Many thanks for the prompt reply. That is helpful. I will look into it myself later and I'm sure we'll sort it out.
Why are error messages often so unhelpful? (rhetorical question) They can be so very precise and at the same time wrong.

Margaret

Latest state of play:

I found safe-mode in FF and tried another test.
Failed (with the same report as before).

I suppose it is encouraging to know that the "behavior" is unimportant. If only I knew how to tell that to my browser, so that my browser can disregard it, too.

Very strange.

Margaret

Zirbelholz wrote:Why are error messages often so unhelpful? (rhetorical question) They can be so very precise and at the same time wrong.

There's a whole branch of programming (Usability) and psychology (Human Factors) that deals with that issue. I know two professors who deal with it almost exclusively. The answer turns out to be "it's complicated". (Sometimes the answer is "people are inconsistently weird".)

In this particular case, IE was implemented incorrectly, but the MS developers didn't really read the CSS standards until too late. They couldn't just fix it, because that would break thousands of sites already using CSS that was incorrect, but worked in IE. (Others may also assign sinister ulterior motives, since those webpages are incompatible with other browsers, but I adhere to Hanlon's Razor.)

They provided a workaround: they implemented a CSS attribute called "behavior" that changes the way their browser behaves. If your webpage uses certain standard CSS attributes, you can tell IE to render them as expected by setting the behavior attribute appropriately.

Unfortunately, it's not a standard attribute, so standard browsers will flag it as an error. Fortunately, standard browsers ignore errors.

So, while it's an error, it will be safely ignored by FireFox. (Why am I so chatty today?)

I just posted a comment on your blog (to the Christmas tree post) using Firefox 2.0. It worked. There must be something wrong with his browser configuration in particular.

Perhaps he has a FireFox add-on that's getting in the way?

Thanks, yes. I did try it myself with Firefox, when I was logged out, and it worked.

But what can be wrong with his configuration?

Margaret

Hi!

Is he maybe using a cookie blocker, or some security setting so that cookies are completely rejected?

Regards,
Garvin

No - he said at the very beginning 'with cookies enabled'. However, I can ask him to double check, but I'll do that tomorrow morning. (He will post through IE - I told him I might enable registration, but on second thoughts I can't find that anywhere).

Margaret

Hi!

You could ask him, if he can install the "LiveHTTP Header" firefox extension. This allows you to track which HTTP Requests the browser makes, and which he recieves. This would allow us to "backtrack" what's happening, if he can provide the HTTP log. There we can also see if the browser accepts cookies or not.

Regards,
Garvin

I will pass this on now - I am going to England in a few hours' time, so I doubt I'll investigate it till the New Year.

Thanks again, Garvin.

Margaret