Strange behavior of survey software spitting out (unnecessary) error messages
Closed, DeclinedPublic
Actions

Assigned To

None

Authored By

	• Purodha
	Apr 21 2011, 11:15 AM

Description

When I first came to the survey, I was distracted for a while during which I left the initial explanatory web page on my screen in order to do the survey later when I had an opportunity to do it.
Coming back, I hit a button to actually start it. I got a screen talking of cookies, and that my session had expired, and more such stuff. I am asking myself what the benefits are of starting a session at a splash screen, and came to the conclusion that is was utterly useless. If it's usesless, don't do it.

The "session expiry" screen offered a way to continue and start the survey, which I followed. Then I got a web page saying:

Wikimedia Surveys
The following surveys are available:

No available surveys

Please contact Wikimedia Operations Team ( software@wikimedia.org ) for further assistance.

[LimeSurvey Survey Software]

The Online Survey Tool - Free & Open Source

Now I decided, to write a bug report, I C&Ped the screen contents, used my browsers "back" button to get a copy of the "session expired" screen, too. This failed, I did not get to the previous web page back, but rather got a 1st page of the survey, as it appears to me.

This is so odd, that I am seriously doubting the survey results now. Whose programming skills are insufficient to the extent shown by the ovserved behavior of their software, obviously cannot be trusted to program decent survey evaluations either.

A quick glance on the (presumed) ast survey page reveals some more obvious flaws:

There is a red asterisk markers at fields - unexplained.
It asks for (illegal under my jurisdiction) personal questions such as income, marrige status, children, etc.
My mother tongue is missing from the list :-( despite its full support in WMF software and even having a Wikipedia, but
there are orthograpies (such as be-tarask. zh-hans) listed as additional native languages?

Apparently, survey makers did also not know enough of what they were doing, content- or meaningwise.

I am sorry. This is, where this survey attempt ends for me. Maybe a better one next year?

Version: unspecified
Severity: normal

Details

Reference: bz28637

Event Timeline

• bzimport raised the priority of this task from to Medium.Nov 21 2014, 11:35 PM

• bzimport added a project: Wikimedia-User-survey.

• bzimport set Reference to bz28637.

• Purodha created this task.Apr 21 2011, 11:15 AM

I forgot: I filled in the 1st page - there were only 4 answers I wanted to / could give (one question, at the end, was not answerable since it's radiobuttons should have been a set of checkboxes in the first place, most of my answers were not listed, and an "other, specify:" group was not provided. It took me very little time. When I hit the "continue" button, I got the "session expired" screen again.

That was the end.

(In reply to comment #0)

This is so odd, that I am seriously doubting the survey results now. Whose
programming skills are insufficient to the extent shown by the ovserved
behavior of their software, obviously cannot be trusted to program decent
survey evaluations either.

Apparently, survey makers did also not know enough of what they were doing,
content- or meaningwise.

Such ad hominem attacks insulting software developers have no place in Bugzilla. You're also assuming that the same person or people that caused this bug will also analyze the results, which is not necessarily true. In fact, this usually not the case for such WMF projects (I wasn't involved in this one so I can't say for sure).

You make some valid points in this bug reports, but they are made in an unnecessarily negative and frustrated tone.

Btw. I do not understand why current survey is not listed on homepage of survey.wikimedia.org.

Since when is software@wm.o a contact e-mail for ops? I thought the general server error pages are supposed to point to noc@wm.o

(In reply to comment #4)

Since when is software@wm.o a contact e-mail for ops? I thought the general
server error pages are supposed to point to noc@wm.o

Or a contact e-mail at all...for that matter. I've never heard of it before today.

(In reply to comment #2)

Such ad hominem attacks insulting software developers have no place in
Bugzilla. You're also assuming that the same person or people that caused this
bug will also analyze the results, which is not necessarily true. In fact, this
usually not the case for such WMF projects (I wasn't involved in this one so I
can't say for sure).

I am sorry. Being a deveoper myself, I was sarcastic about my own class.
I do assume that some rudimentary analysis is inherently part of the software collecting survey data, yes, likely because that's usual. I do not assume more overlap, however.

You make some valid points in this bug reports, but they are made in an
unnecessarily negative and frustrated tone.

Yes I was frustrated. The survey not working was only a little tile in the pile.
I apologize for letting it bleed trough into my wording.

User survey has been closed 2 years ago. No point in keeping bug reports for a product which is no more available :-]

(In reply to comment #7)

User survey has been closed 2 years ago. No point in keeping bug reports for
a product which is no more available :-]

survey.wikimedia.org certainly used to exist. We can see remnants of it here: https://www.google.com/search?q=site%3Asurvey.wikimedia.org.

I believe the user survey is re-run periodically, so I wouldn't be surprised if these same issues re-emerged in the future. As survey.wikimedia.org currently appears to be completely inaccessible/broken, this bug can likely stay closed for now.

LimeSurvey was uninstalled entirely and the service shut down. From what I understand, nobody plans on ever using LimeSurvey again. So yeah, WONTFIX sounds appropriate.

Strange behavior of survey software spitting out (unnecessary) error messagesClosed, DeclinedPublicActions