Maintainers' Survey 2021 results...
...can be found at https://en.opensuse.org/Maintainers-surveys_2021. Let me know if you find any typo or error. A post on news.opensuse.org summarizing the key results should follow. Thank you very much for your time and help to everyone who participated.
Some of the results reporting / data handling I find a little strange. For example live in Western Europe (peaking at 57.7% on the UTC+1/2 timezones, with a handful in the Americas (6%) and Far East Asia (6%) left me wondering where the remaining 30% were. Scrolling down it seems those numbers come from 4 timezones, all other timezones having been removed because they were "below 5%" (if I interpret it right). A third of data for this item are being filtered out just because they are spread across more time zones? On 11/8/21 05:26, Adrien Glauser wrote:
...can be found at https://en.opensuse.org/Maintainers-surveys_2021. Let me know if you find any typo or error. A post on news.opensuse.org summarizing the key results should follow.
Thank you very much for your time and help to everyone who participated.
Yes, they are not reported on this page because they spread evenly (under 5% so there is less than 5% variation between them) across other time zones. In other words, there is a huge cluster of people within the +1/+2 time zones. That's not much of a surprise but it's still good as a confirmation. I've tried to make the results as minimal as possible, because it's quite a long document. So very marginal results have been omitted -- it's always made explicit in the documentation -- for the sake of brevity.
It seems to me it make would make so much more sense to report the totals within ranges of timezones for the americas, asia, europe, oceania, etc (whatever the categories are decided to be) rather then to use only the top timezones. On 11/8/21 10:25, Adrien Glauser wrote:
Yes, they are not reported on this page because they spread evenly (under 5% so there is less than 5% variation between them) across other time zones. In other words, there is a huge cluster of people within the +1/+2 time zones. That's not much of a surprise but it's still good as a confirmation.
I've tried to make the results as minimal as possible, because it's quite a long document. So very marginal results have been omitted -- it's always made explicit in the documentation -- for the sake of brevity.
There you go: Answer Count -11 (AO01) 0 -10 (SQ001) 1 +8 (SQ002) 12 +9 (SQ003) 5 +10 (SQ004) 2 +11 (SQ005) 1 +6 (SQ006) 3 -2 (SQ007) 1 -4 (SQ008) 5 +7 (SQ009) 5 +5 (SQ010) 4 +4 (SQ011) 1 +3 (SQ012) 10 +2 (SQ013) 94 +1 (SQ014) 21 0 (SQ015) 8 -1 (SQ016) 1 -3 (SQ017) 12 -5 (SQ019) 5 -6 (SQ020) 5 -7 (SQ021) 3 -8 (SQ022) 0 -9 (SQ023) 0 Total(gross) 199 For my information, what is the important information that you get there but not from the report?
Em seg., 8 de nov. de 2021 às 13:49, Adrien Glauser < adrien.glauser@gmail.com> escreveu:
There you go:
Answer Count -11 (AO01) 0 -10 (SQ001) 1 +8 (SQ002) 12 +9 (SQ003) 5 +10 (SQ004) 2 +11 (SQ005) 1 +6 (SQ006) 3 -2 (SQ007) 1 -4 (SQ008) 5 +7 (SQ009) 5 +5 (SQ010) 4 +4 (SQ011) 1 +3 (SQ012) 10 +2 (SQ013) 94 +1 (SQ014) 21 0 (SQ015) 8 -1 (SQ016) 1 -3 (SQ017) 12 -5 (SQ019) 5 -6 (SQ020) 5 -7 (SQ021) 3 -8 (SQ022) 0 -9 (SQ023) 0
Total(gross) 199
For my information, what is the important information that you get there but not from the report?
Hi, using some approximations like this: Zone Count Percent Europe/Africa ( -2 UTC to +3 UTC) 135 67.8% Asia/Oceania (+4 UTC to -10 UTC) 34 17.1% Americas (-3 UTC to -9 UTC) 30 15.1% Regards, Luiz
World regions like "Europe", "Asia", "Oceania" are difficult to work with because they evoke different things in different people ("Is Russia in Europe?", etc) and some combinations of Northern/Southern regions along the same timezone make more sense than others ("Europe/Africa", in this particular case, does not make much sense). For these reason we steered away from these fuzzy notions. I guess the alternative would have been to just copy-paste the data; but since this is more a digest than a copy-paste (after all, the point is to give the gist of the results to the reader who does not want to get sucked into small details -- for small details people can check the data set directly), it seems to that me presenting this particular datum in a way that emphasizes the cluster-like aspect of the distribution does not betray the story plot. That said I reckon that this is the result of a comprise, and people don't always like to compromise on the same things. If / when I figure out a good way of exposing some of the raw data directly, without requests, I will do it.
On 11/9/21 03:19, Adrien Glauser wrote:
There you go:
Answer Count -11 (AO01) 0 -10 (SQ001) 1 +8 (SQ002) 12 +9 (SQ003) 5 +10 (SQ004) 2 +11 (SQ005) 1 +6 (SQ006) 3 -2 (SQ007) 1 -4 (SQ008) 5 +7 (SQ009) 5 +5 (SQ010) 4 +4 (SQ011) 1 +3 (SQ012) 10 +2 (SQ013) 94 +1 (SQ014) 21 0 (SQ015) 8 -1 (SQ016) 1 -3 (SQ017) 12 -5 (SQ019) 5 -6 (SQ020) 5 -7 (SQ021) 3 -8 (SQ022) 0 -9 (SQ023) 0
Total(gross) 199
For my information, what is the important information that you get there but not from the report?
Interesting to me with knowledge of who is in what timezone and our broader community I can get a rough idea of who is in "Far East Asia" vs Who is in Australia most people in +9 are probably in Japan maybe with a few in Indonesia where as the people in +10 are probably in Australia (but who knows they could be in Russia). But then timezones still aren't a great measure because at the start of the survey Sydney was at +10 and now its at +11, maybe if we are using the results to determine region next time we can ask a better question. Id be interested in the split between North, South and Central America for example. For me the areas where we have just a handful of contributors are more interesting then the ones where we have a lot (I already have a decent idea of the places where we have alot of contributors). -- Simon Lees (Simotek) http://simotek.net Emergency Update Team keybase.io/simotek SUSE Linux Adelaide Australia, UTC+10:30 GPG Fingerprint: 5B87 DB9D 88DC F606 E489 CEC5 0922 C246 02F0 014B
That's why we went for timezones + hemisphere: it's less identifying than country (some folks suggested asking for country was too identifying) and much more precise -- and less fuzzy or error-prone -- than regions/continents if you can cross over the two data, i.e.: - hemisphere: Northern --- +1: % --- +2: % ... - hemisphere: Southern --- +1: % --- +2: % ... When we designed the survey I thought I'd be able to do that, but I don't see that button now. If you know where it is, I am all ears!
On Tue, 09 Nov 2021 09:04:30 -0000, "Adrien Glauser" <adrien.glauser@gmail.com> wrote:
That's why we went for timezones + hemisphere: it's less identifying than country (some folks suggested asking for country was too identifying) and much more precise -- and less fuzzy or error-prone -- than regions/continents if you can cross over the two data, i.e.: - hemisphere: Northern --- +1: % --- +2: % ... - hemisphere: Southern --- +1: % --- +2: % ... When we designed the survey I thought I'd be able to do that, but I don't see that button now. If you know where it is, I am all ears!
Personally I would be interested in native language of the people: English vs Spanish vs German vs Dutch vs … IMHO it is more informative than country -- H.Merijn Brand https://tux.nl Perl Monger http://amsterdam.pm.org/ using perl5.00307 .. 5.33 porting perl5 on HP-UX, AIX, and Linux https://tux.nl/email.html http://qa.perl.org https://www.test-smoke.org
Hello, Am Montag, 8. November 2021, 11:26:41 CET schrieben Sie:
...can be found at https://en.opensuse.org/Maintainers-surveys_2021. Let me know if you find any typo or error. A post on news.opensuse.org summarizing the key results should follow.
Thanks for your work on the survey! I wonder if some sections under "Demographics" miss the % sign. For example, I somehow doubt that 12.5 persons are 50+ - I've never seen a half person ;-) Similar question for the "good for..." section - it says "answers: 40", but the individual items have numbers > 40, so either the individual numbers are percentiles, or the number of answers is wrong. I'd also recommend to use a different formatting for the bullet lists with text answers - at the moment, each of the longer ones has a horicontal scrollbar, which makes reading them very annoying. Maybe <tt> instead of <pre> would be an option for the one-line answers? Regards, Christian Boltz -- What's about the flash plugin, there a licence agreement is displayed when installing (that's a good thing, not installing flash plugin makes web useable again ;-) ). [Manfred Tremmel]
On Monday 2021-11-08 21:28, Christian Boltz wrote:
Hello,
Am Montag, 8. November 2021, 11:26:41 CET schrieben Sie:
...can be found at https://en.opensuse.org/Maintainers-surveys_2021. Let me know if you find any typo or error. A post on news.opensuse.org summarizing the key results should follow.
Thanks for your work on the survey!
I wonder if some sections under "Demographics" miss the % sign. For example, I somehow doubt that 12.5 persons are 50+ - I've never seen a half person ;-)
That is probably for those that couldn't decide. Just like the series 1 + 1 - 1 + 1 - 1 + 1 ... converges to ½ ;-)
My pleasure :) 3rd sentence reads 'Values for "answers" are plain numbers. All other values are percentiles, unless mentioned otherwise.' But if you missed it, that's my bad (doing without the % was supposed to make the whole thing less cluttered and to save some typing ^__^'). I've added all percentiles now. Good suggestion about formatting; implemented that too.
participants (7)
-
Adrien Glauser -
Christian Boltz -
Eric Schwarzenbach -
H.Merijn Brand -
Jan Engelhardt -
Luiz Fernando Ranghetti -
Simon Lees