Forum logins are broken (again)

This happened yesterday with firefox, and today with rekonq.

I’ll describe what happened today.

My prior session had timed out. So I closed the browser (to delete session cookies). And then I opened rekonq.

I went to the main forum page, and logged in.

I clicked “New Posts” and got the list.

Then I clicked “Settings” so that I could check for PMs. I got a page “You are not logged in”

I went back to the main forum page. It showed me as not logged in.
I clicked “Login”. Then I got the same page back again. This repeated, a seemingly infinite loop.

Something seems to be wrong with the redirect.

Yesterday, I had the same problem with firefox.

Right now, I am using firefox. I logged into the bugzilla first, then went to the main forum page, and that seemed to work.

Someone emailed me about login issues and the re-direct.
But I haven’t noticed anything.

I’m using Chrome
But right now I tried Firefox and it was fine

I just now consecutively used Chrome, FFox, Konq and I.E. 9.
Logged on here no problems.

On 06/08/2013 05:46 AM, swerdna wrote:
>
> I just now consecutively used Chrome, FFox, Konq and I.E. 9.
> Logged on here no problems.

which i think means OP should delete cookies from opensuse.org,
novell.com, and netiq.com…then restart the browser and try the
login again…


dd

The login/session maintenance is indeed sometimes flaky especially if we are had logged in and return to the forum after few hours. The forum software is unable to decide whether we are logged in or not.Probably at those points there is a conflict of session between “forum” software and attachmate login service .

I clear cache and restart my browser to restore normality.

No, it doesn’t mean that at all.

Rather, it means that somebody should fix the forum software so that it works reliably without having to mess around with deleting cookies.

Or perhaps you are saying that only geeks and nerds should use opensuse, that people who don’t have the technical skills to delete cookies should stick to Windows or Ubuntu.


I’ll grant that deleting cookies helps, but it should never be necessary. That it helps is a clear sign of broken forum software.

> Rather, it means that somebody should fix the forum software so that it
> works reliably without having to mess around with deleting cookies.

nice attitude today!! what happened, did someone steal your ball?

> Or perhaps you are saying that only geeks and nerds should use
> opensuse, that people who don’t have the technical skills to delete
> cookies should stick to Windows or Ubuntu.

no, i didn’t say that. but i could say:

If the shoe fits, wear it.

and if you are not willing to do what it takes to log into the forums
using http, you could use nntp and avoid lots of faults and
slowdowns… easy how-to here:
https://forums.opensuse.org/faq.php?faq=novfor#faq_nntp

> I’ll grant that deleting cookies helps, but it should never be
> necessary. That it helps is a clear sign of broken forum software.

clear sign? really? couldn’t it also be a clear sign of broken
browser software?

or a transient and momentary fault in any transmissions that jumbled
one (maybe two?) of YOUR cookies and not everyone else in the world…

on the other hand, i agree that all software whether distributed or
used by openSUSE should be 100% perfect…and, if not then it should
not be distributed or used…

and, then we could all move to the perfect software of . . . of who?


dd

As I went to reply to thread, my SeaMonkey session restored from whenever finally timed out (not surprisingly). I closed browser, re-opened it, and signed in and now replying to thread, so far so good. However I have recently experienced the problem you describe, the only differences being SeaMonkey and here it happens from time to time, I would say infrequently. What browser is deployed doesn’t appear to matter.

When it happens, cookie deletion is usually the only remedy, as a last resort. It’s likely that I am using the service provided by the web interface at the edge of its intended capabilities. I often restore sessions sometimes after many elapsed hours, and don’t always close the browser as advised. The staff have tweaked the system to give us a lengthy period of “staying logged in”. I assume there has to be limits and mechanisms for security reasons (both mine and the services’). For all those reasons I can manage the infrequent occurrences that require cookie deletion and re-login.

It’s only really frustrating when one loses technical input that required painstaking preparation. There are precautions one can take, such as copying the input somewhere (clipboard even) before sending, and doing the re-login in a separate tab to preserve the original input.

I have not deleted my cookies for longer than I can remember (years)
Even when the consensus of opinion has been to do so. Cache - yes, but not cookies.
Just FYI

On 06/09/2013 05:26 AM, caf4926 wrote:
> Cache - yes, but not cookies.

interesting!

hmmmmmmm…i wonder how a browser’s cache content might affect site
log-in . . .


dd

Hmmm, I wonder…

If I delete all cookies, I tend to do cache at the same time, but not for any particular reason other than spring cleaning. However, to me it seems pointless knowing that stuff without also understanding the user’s logout/login habits and level of access.

If browser cache is not properly synced with the “online” page then you would sometimes notice that sometimes “new” replies to a thread will not appear. This is because the browser instead of getting new content from the online webpage, it will “serve” the webpage from it’s own cache(this cache is probably from previous visit to the same thread.)

I have “set” my browser to check / compare the cache content each time with the online webpage so that i see the latest content.

http://paste.opensuse.org/images/41065481.png

If… maybe… etc, but dealt with by using the “reload” browser button, and that action hasn’t reset the login issue in question AFAICT.

Maintaining session objects/cookies and browser cache are entirely different subjects as far as i know.

I believe that the issue is not with the software of the forum itself but “how the forum software interacts with the attachmate/net iq authentication”. I think the root cause is sometimes the “session” is “live” in forum software(may be due to browser cookies etc) but “killed/expired” at attachmate/ent iq end (authentication server session has expired).

On 06/09/2013 11:26 AM, vazhavandan wrote:
> If browser cache is not properly synced with the “online” page then you
> would sometimes notice that sometimes “new” replies to a thread will not
> appear.

hmmmm…i had assumed (wrongly?) that what is served to the reader
asking for “New posts” would be done by ‘looking’ at the date/time
stamp of the last log off and then serving all posts since that
time–and not by looking in the asking user’s cache…

and, that date/time stamp might exist in either in the user’s cookie
or in the server’s memory/log/something–and, again i wouldn’t expect
that to be pulled up from the log-off page in cache (if it existed
there)…

still, i don’t see a connection between log-in problems and browser
cache–which is not to say that cache is not the problem, instead
just that i don’t know how it might be…(and would like to learn)…


dd
openSUSE®, the “German Engineered Automobile” of operating systems!
http://goo.gl/PUjnL
http://tinyurl.com/DD-Caveat
http://tinyurl.com/DD-Hardware
http://tinyurl.com/DD-Software

So you were just digressing wrt reloading web pages to the cache, a pretty basic browser operation. :slight_smile:

i probably missed the “log-in” in next line.:wink:

The two systems may use different time limits and possibly different session cookies, but I’m not sure. The cookie clearance has always worked, even if only deleting the o’suse + a’mate + novell + etc. cookies. I still think the issue is mostly seen by those of us who “push the envelope” on a system set to disconnect general users on reaching a certain session length either with or without activity. The forum, and Attachmate’s front end, just do not match on requirements and mechanisms. If we logged out/in and closed browser followed by starting new browser session a bit more frequently, I suspect we wouldn’t be affected by it.

Here’s a little bit of data that may or may not help someone lots smarter than
I to understand the problem:

I was experiencing the very same problem described by the OP, but I only had
the problem on one PC and only with one browser. The problem occured when
using Chrome on my primary PC, not with Firefox or Konqueror. Other PCs did
not have any issues. I first noticed the issue on June 4. There had been no
recent changes to browser software here for some time, so I suspected the new
issue was triggered by a change in the forum software.

I tried clearing cache, deleting cookies, restarting Chrome, even rebooting…
no luck.

Then I realized that the one and only PC with the problem was running a
different version of Chrome (28.0.1500.36, a beta version) while the others were
all running the stable 27.0.1453.110. I’ve no idea how or when the change to
the beta had happened, but after I switched back to the stable version the
problem went away.

Obviously my situation is different than the OP’s in many ways, but I offer the
notion of switching browser versions to resolve it in hopes that the data might
trigger an “ah-hah moment” for someone far brighter than I.