Replies (26)
Please login or register to join the discussion.
I have the same error, started yesterday. I have tried logging out and back, to no avail.
I had a few yesterday ... and eventually found a solution.
If you look at the URL displayed in your browser (where it starts accountingweb.co.uk/ ..) you will find a # character (after, on this thread, any-answers/http-error-403).
That is then followed by some text that indicates a specific 'place' on the page - such as comment-1040885
BUT when you encounter the 'failed thanks' pop-up box, you will see that loads of spurious looking characters have appeared after the # character.
If you delete those (starting with the # character and everything after it) then the page will re-open without any focus (i.e. at the top of the thread) ... but you can now scroll down to the comment you liked and hit Thanks successfully!
WHY any of this is necessary (and why it only affected a random selection of threads) is a mystery that Aweb might like to explain?
BTW for me the issue was more prevalent (but not universally so) on Articles rather than Any Answers threads - but that might just be randomness in action.
It doesn't seem like anyone at Aweb is reading this thread (or is bothered by the problem) ... but, just in case someone cares, my further research shows that:
* ANY post on page 2 (or probably later) of ANY thread cannot be Thanked.
I've tried multiple different threads, of varying vintage, and you always get the same refusal ... with a similar but differently worded rejection if on a mobile.
[This problem is not amenable to my 'solution' above - so the problem is more wide spread than originally thought.]
Presume they'll get to it at some point.
They've a full diary at the moment, what with the Olympics and Team Bonding sessions; not to mention new guidelines to draw up.
And now you can't even 'Report' a comment - at least if it is on page 2 of a thread!
Viz https://www.accountingweb.co.uk/any-answers/client-asking-for-a-meeting-...
... where new 'member' https://www.accountingweb.co.uk/profile/keishalang-0
... has posted a link where (apparently) you can cast spells (!) to "help you get back your ex-husband/ex-wife"
So respected members can't be trusted to respond to specific Articles ... and when allowed to do so may find their comments arbitrarily vaporised (without notice) ... BUT we can no longer even Report obviously dangerous (and probably fraudulent) new members.
What was it Wes Streeting said about the NHS?
https://www.gov.uk/government/speeches/statement-from-the-secretary-of-s...
Ive just tried to report that too and it says 'forbidden'.
Mods should be on it without the need for us to do it anyway.
Maybe we should just let Aweb become a dumping ground for spam. ;-P
And now there's another one (again on a page 2):
https://www.accountingweb.co.uk/any-answers/how-can-we-fix-hmrc
... pure advertising/garbage from 'new member': The Success Talks
But we can't Report it!
Spammers moving faster than Aweb to exploit the new deficiency?
That's reassuring. I trust that the problem can be swiftly resolved. Huge thanks to @FactChecker for devoting so much time & effort to highlighting the problem, as well as coming up with a (provisional) fix in the meantime.
Hi Richard,
Any update?
As far as I can tell, the common factor is when the URL contains a # character - but that's not conclusive (and TBH I'm more interested in a solution than the problem).
Our Dev team are on the case. Unfortunately, like a lot of cases, it's not as simple as turning the system on and off again and then it's back working. It just might take some time to dig deep into the product.
Interestingly (and I use that word in its loosest sense), your link generates the page:
"Forbidden
You don't have permission to access this resource"
- rather than the 'HTTP error 403 occurred' one!
The 'link back' to what you were trying to do can be seen within the path of the URL (at ../flag/like_comment/..) so it shows that the two problems we've all been reported are highly inter-related.
The clues are set out in my above comments ... and will be tied to a deliberate change (as in 'update') performed by whoever looks after your software (or less likely the server from which it operates) over the night of 24th/25th July, when the fault was first reported.
That really should make it fairly easy to find for any software investigator ...
FWIW it really is having a debilitating impact on the site .. as it's not just 'Thanks' that are disabled (in the circumstances I've set out), but ANY form of interaction (including 'Report' for dangerous new 'members') in those same scenarios.
Any news, Richard?
It may not be simple ... but it's now 5 weeks since the issue was first reported and we've heard nothing.
The issue is easy to reproduce ... any scenario where the Sift software inserts certain characters (so far I believe ? or = or #) within URL in the browser title-bar.
This means ANY page 2 (or 3 etc) of a thread - and any page to which you've been directed by selecting the link within those Emails informing you that someone has thanked you (even for someone else's post that catches your eye) - and a few other scenarios.
Usually in software, if you've got details of the problem plus a description of how to reproduce it then you're most of the way there to finding the recalcitrant code or routine - at which point it's just a matter of testing a fix and then scheduling it being applied.