Surveys in the German Community about Testing, Stable, Bugreport

[Lastwebpage]

Hello,
we have made two surveys in the German forum, here are the results:
--------------------
Poll 1: Which version do you use?
current stable: 6 (10.3%)
current testing: 47 (81%)
older stable: 1 (1.7%)
older testing 1 (1.7%)
SVN 3 (5.2%)

Overall: 58 votes
http://forum.miranda-im.de/index.php?topic=10050.0
-------------------
Poll 2: How often you make bug reports?
Testing: regular (>20) 4 (9.1%)
Testing: often (5-20) 4 (9.1%)
Testing: sometimes (1-5) 12 (27.3%)
Testing: only in IRC or Jabber channels 7 (15.9%)
Testing: never 6 (13.6%)
Stable: regular (>20) 0 (0%)
Stable: often (5-20) 0 (0%)
Stable: sometimes (1-5) 0 (0%)
Stable: only in IRC or Jabber channels 1 (2.3%)
Stable: never 3 (6.8%)
I don't know how and where I can make a bug report. 7 (15.9%)

Overall 44 votes
http://forum.miranda-im.de/index.php?topic=9703.0
=======================================

I think the three important results are representative for all Miranda users:
1) about 90% use the testing version
2) less than 50% make bug reports, less than 10% regular
3) no one make bug reports for the stable version

=======================================
I am not sure about the conclusion and consequences,
but I find the results interesting.

Peter

[Sami]

I want express my opinion because it's easy to misinterpret the results.

Originally Posted by Lastwebpage

I think the three important results are representative for all Miranda users:
1) about 90% use the testing version

90 % of the people who read forums regularly and/or are active enough to answer.

2) less than 50% make bug reports, less than 10% regular
3) no one make bug reports for the stable version

Firstly, active people tend to report the easy-to-find bugs very fast. Less active users don't test features that rigorously in order to find other bugs.

Scondly, all users can't report bugs. In that case the devs would be seriously incompetent. I would rather look at the number of people who never do anything (20.4 %).

Thirdly, my observation is that people are trying in the bug tracker, albeit many bugs get resolved status instantly.

=======================================
I am not sure about the conclusion and consequences,
but I find the results interesting.

Peter

My conclusion is that dev builds are prominent in the forums, whereas the bug tracker is not. It's interesting considering the desired state and means to get there.

[Schelm]

I agree with you. Such a survey only reflects the situation of a community about a piece of software. That does never reflect the overall situation!!!

By the way: It could help to add an link to the bugtracker in the forums like you did with the core mailing list. I guess most people download a dev build and don't read the dev site completely.

[nowotny]

Originally Posted by ASSami

My conclusion is that dev builds are prominent in the forums, whereas the bug tracker is not. It's interesting considering the desired state and means to get there.

Well, my conclusion is exactly the opposite... more bugs are fixed after a report in the bugtracker, and faster, then after a report on the forums...

[Sami]

Originally Posted by nowotny

Well, my conclusion is exactly the opposite... more bugs are fixed after a report in the bugtracker, and faster, then after a report on the forums...

I agree with you on that. I didn't say anything about what is useful where because developer's needs do not depend on what users vote.

[koobs]

Lastwebpage, thanks for the survey results. Although the sample size is small and biased to a particular subset of the miranda users, im nontheless very interested in "concrete" numbers about:

- How many active users use Miranda daily, weekly, monthly, yearly.
- What OS are they using, and in what language
- What is the breakdown of versions in active use.
- What plugins are in use (with details and aggregation by name, version, type (unicode/ansi))
- What network configurations are in use (proxy, direct, upnp, etc)

This would help immensely in the following areas:

- Knowing how widely deployed Miranda really is.
- Making informed choices for supporting previous versions.
- Allocating resources/people to upto date translations
- It's just VERY cool to know these things

From a technical standpoint, acquiring the information is relatively trivial and we have the likes of miranda-vi already. From a privacy and political standpoint these issues are traditionally very controversial, and can risk the confidence of our users.

Edit: I felt it was worthwhile thinking about this in more detail. http://www.miranda-im.org/2009/04/11...rivacy-debate/ incase anyone misses it on the front page