Slash Boxes
NOTE: use Perl; is on undef hiatus. You can read content, but you can't post it. More info will be forthcoming forthcomingly.

All the Perl that's Practical to Extract and Report

use Perl Log In

Log In

[ Create a new account ]

barbie (2653)

  reversethis-{ku. ... m} {ta} {eibrab}

Leader of [] and a CPAN author []. Co-organised YAPC::Europe in 2006 and the 2009 QA Hackathon, responsible for the YAPC Conference Surveys [] and the QA Hackathon [] websites. Also the current caretaker for the CPAN Testers websites and data stores.

If you really want to find out more, buy me a Guinness ;)

Memoirs of a Roadie []
CPAN Testers Reports []
YAPC Conference Surveys []
QA Hackathon []

Journal of barbie (2653)

Wednesday January 07, 2009
04:46 AM

A Little Perspective

[ #38228 ]

Recently some have berate me for not doing enough and spending more time on the CPAN Testers Reports website. Unfortunately I'm getting used to that. However, one individual rather irritated me by claiming that the site doesn't work for 95%+ of the world. Now while I'm quite sure 95%+ of the world doesn't care about CPAN Testers, I'm also quite sure this person has never spoken to the several thousand users of the site. However, it did make me wonder what the audience of the site used. Now the best data representation I have available to me is Google Analytics, which isn't perfect, and doesn't give me the ability to provide the following data based on unique visitors. As it is the following is based on 7,628 individual visits (from 6,266 absolute unique visitors) between 6 December 2008 and 5 January 2009.

Firstly we have the top 10 operating systems, followed by the top 10 browsers, and lastly the top 10 browsers AND operating systems

1.  Windows                         4,878   63.95%
2.  Linux                           1,683   22.06%
3.  Macintosh                         948   12.43%
4.  FreeBSD                            32    0.42%
5.  (not set)                          24    0.31%
6.  SunOS                              19    0.25%
7.  OpenBSD                            14    0.18%
8.  iPhone                              9    0.12%
9.  NetBSD                              8    0.10%
10. iPod                                5    0.07%

1   Firefox                         4,201   55.07%
2.  Internet Explorer               2,071   27.15%
3.  Safari                            499    6.54%
4.  Chrome                            276    3.62%
5.  Mozilla                           261    3.42%
6.  Opera                             198    2.60%
7.  Konqueror                          55    0.72%
8.  SeaMonkey                          37    0.49%
9.  Camino                             21    0.28%
10. Mozilla Compatible Agent            3    0.04%

1.  Firefox / Windows               2,329   30.53%
2.  Internet Explorer / Windows     2,071   27.15%
3.  Firefox / Linux                 1,338   17.54%
4.  Firefox / Macintosh               466    6.11%
5.  Safari / Macintosh                437    5.73%
6.  Chrome / Windows                  276    3.62%
7.  Mozilla / Linux                   229    3.00%
8.  Opera / Windows                   133    1.74%
9.  Konqueror / Linux                  53    0.69%
10. Safari / Windows                   46    0.60%

It surprised me to see iPhone and iPod in the operating system list, but pretty much everything else is as expected. Assuming these numbers, the self proclaimed end-user who emailed me, actually represents just 27.15%. In reality I currently target Firefox, and from the numbers above, that would seem reasonable. I currently don't have access to a Windows or Mac desktop, and thus I'm not able to test on IE or Safari, so if there are problems I would be glad of help to try and fix any problems.

However, please bare in mind that everything about CPAN Testers is a volunteer effort and berating me for not doing enough or ordering me to buy new hardware and/or software to get the site working for you, doesn't help and allows me to completely ignore your requests as the ravings of a lunatic. What does help is identifying where the errors occur, for example the line and file of a javascript error or by sending me a screenshot, or perhaps even sending me a patch if you're able to. Some have sent helpful pointers to problems and I've been able to resolve some problems within a few minutes.

CPAN Testers Reports gets between 200 and 400 visits a day, with over 10,000 hits in the course of a month. It might not be a mega site, but it does get used, and I would prefer for everyone to use it effectively. If it isn't working effectively for you at the moment, please remember that constructive critism is useful, claiming numbers that are not represented above and providing nothing helpful are not.

One person I would like to highlight and publicly thank though, is David Westbrook. Last week David asked whether I would be happy for him to improve the test suite for the code that builds the CPAN Testers Reports website. I was delighted for the offer, as the test suite is rather sparse. Over the weekend David sent me a very notable patch, which not only consisted of many more tests, but also picked up a couple of subtle bugs and provided me with a more complete todo list. David's efforts will be intergrated into the next release over the course of the next few days, and are a shining example of how to help improve CPAN Testers. Thanks David, I owe you beer ;)

The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
More | Login | Reply
Loading... please wait.
  • The subject says it all. Rest assured that there is plenty of people who appreciate your effort. It's just that if something is working as expected (or even better), nobody typically talks about it much.
    Thanks for keeping up the great work!

    • I know most people appreciate the changes, progress and direction, that wasn't what I was getting at, but thank you for saying so :)

      Unfortunately some people expect me to understand "it's broken" with no relevant information such as error messages or even the page they were looking at. When those same people tell me I should buy a new desktop or laptop, or tell me to target an audience that is irrelevant, I have to wonder why they bother to the send the email in the first place. Ranting at me isn't going to

      • There are these people who think they deserve to get everything. I cannot say "without giving anything" as some of them are even writing Open Source code but most of them don't.

        The best you can do is to disregard them or maybe to give a short reply, "patches are welcome".

        Most of them won't bother you again.

        Some might learn from that.

        Just make sure you hear the silent majority who appreciates your work.

  • Thanks for the work that everybody puts into CPAN Testers. I - at least - find it bloody useful. Both as a user and a CPAN author.

  • WFM on Safari. If you like, I could see if I could give you VNC access to one of my Macs for testing it. Although it'll be *hideously* slow. Apple need to hurry up and let people virtualise OS X.
  • ... but i suspect the used data is the best representation. Why not use the webserver logs instead?
    People explicitely block m/.*oogle.*(ytics|syndicat).*/ for shorter page load times and better privacy.
    Would be nice to know the actual percentage of people trying to protect themselves, but that's a different story...

    Anyways, thanks putting together these numbers!