Ranking system going crazy?

[#1]

That's good to know @jaread83 at least you won't have to spend hours or days doing housekeeping.

[#2]

---

@Anonymous wrote:

at least you won't have to spend hours or days doing housekeeping.

---

It should be quicker than when I put all the formulas in initially

[#3]

Are there any updates from Lithium on this fairly major bug @jaread83   Another two weeks have passed with no response.  Has it slipped off the Lithium priority list?

M

[#4]

Still being looked into. As I said before, I will update when I receive any news.

[#5]

I know you said you would update this post when you had something to post but this is getting beyond a joke. The OP is now 48 days old at the time of this writing yet nothing has been done with the exception of an acknowledgement that a fault exists and that itself was 32 days ago!

Surely the software development guys know what changes were made to the system, after all that is what a version control change log is for. As a software engineer myself I appreciate that bug hunting can be tedious work, but I also know if I were to take this amount of time to find a bug I’d be out of a job.

As Jeremy Clarkson would say ‘How hard can it be?’

[#6]

@Anonymous,

The problem here is you presume that there is change control. We've seen cases and read comments which imply that Lithium apply changes to the system of which Plusnet are not aware.

Such discipline would not lead me to conclude that Lithium has a strong source change control process. If they had, as you suggest it ought to be simple enough to identify what changes have been made to the rank (or is that wrank) management code module.

[#7]

@Townman, if what you say is correct then I am astounded and it would be interesting to read the SLA that Plusnet have with Lithium that allows them to do this. I can understand the need to do an urgent upgrade when an issue that affects security is found but I would still expect that to go via staging prior to being pushed to production.

My assumption regarding change control is based not only on common sense but that of an approach any professional software company would take, and I assume Lithium put themselves in this category. So for them not to have such an implementation is tantamount to, if not directly, professional neglect.

[#8]

I believe that Lithium have modified their process and called it "Chance"  Control - Push it live and hope it works? 

Never works in practice!

[#9]

@MauriceC - You could be so close to the truth there its scary.

[#10]

@Anonymous,

I understood the premise for your assumption - it was based on sanity.

As for code being tried in staging before live, well we know this issue has not been see on staging!

[#11]

Now that is odd, so has anyone taken the time to hash the files between staging and production to see if there are any differences in the source files? A quick script to hash the JS files shouldn’t be to difficult to do and would highlight the minutest of differences.

If any difference are encountered then a quick diff will tell you what is what.

This issue could be a JS event bubble not being cancelled at the appropriate time,  a wrongly placed return statement, semi colon, fall through you name it, but a hash will prove that the files are identical so eliminate the JS code as the source.

[#12]

@Anonymous,

I've already asked the question.

[#13]

Doesn't appear to be a response though, so Plusnet you can have this if it's useful.

sudo find /var/www/ -type f -iregex '.*\.\(js\|css\|html\)' -exec md5sum {} + > ~/filehashes.md5

With this you'll end up with a file called filehashes.md5 in the home dir of the current user whose content looks like this:

a92ebe9164a79fa3c78bddeb3a2ae849  /var/www/js/graph.js
41f6620ad18a9712efe5688de9e33163  /var/www/index.html
...

 

[#14]

The ranking bug has been fixed and I have got Lithium engineers to observe the Community for now to see if this is all working as expected.

I will be keeping my eye on things for now but so far I can see that ranks have been reapplied to everyone correctly (so far) so this is looking positive at the moment.

Thank you for your patience with this one.

[#15]

so is the spam PM filter working again now