Poll: Report your Windows version

[sledgehammer999]

WINDOWS USERS ONLY NO ACCOUNT NEEDED FOR VOTING

This is a poll that will help us determine our non-Win10 userbase. The new Qt6 supports only Windows 10. We plan to eventually support it and drop Windows 7/8/8.1 support. We want to figure out how big our non-Win10 userbase is, so we decide how soon (or later) we should drop Windows 7/8/8.1 support.

Vote and let others know about this poll.

Which Windows version do you use modern qBittorrent releases with?

(click on the relevant progressbar to vote)

Windows 7 or 8 or 8.1

---

Windows 10

[FranciscoPombal]

@sledgehammer999

This poll UX is terrible. The whole thing looks like some sort of image link to an external polling site, so I accidentally cast my vote for the wrong option. I suppose this might also happen to others. EDIT: at least you could place each bar further apart.

Also, I don't think this is terribly relevant. You will have a lot of selection bias for the users who visit GitHub. IIRC, you have access to FossHub download data. Why not just check that?

[sledgehammer999]

you have access to FossHub download data

They don't have Win ver breakdown.

[sledgehammer999]

I did some visual updates.

[FranciscoPombal]

@sledgehammer999

They don't have Win ver breakdown.

Bummer. Do they at least have browser user-agent data? We can extract the Windows version from that (it's OK if they only have raw data/logs, I don't mind writing a small script to process it myself). Reasonable assumption I'm working with to guarantee validity of the conclusions taken from the data: only a very small number of people change their browser user-agent.

[sledgehammer999]

Also, I don't think this is terribly relevant. You will have a lot of selection bias for the users who visit GitHub

As I said on my other comment there will proper news entries linking here along with the next releases. Also I created a forum topic. And I plan to announce to the IRC channel. Word will spread(hopefully).

Do they at least have browser user-agent data?

The stats in the dashboard are simple. Countries and OSes (with no version breakdown). I will ask them if they have those and if they can give me such info.

[xavier2k6]

The stats in the dashboard are simple. Countries and OSes (with no version breakdown). I will ask them if they have those and if they can give me such info.

Can same be done with sourceforge?

[FranciscoPombal]

I did some visual updates.

I did some more, hope you don't mind.

[xavier2k6]

@sledgehammer999 @FranciscoPombal I'd make the Windows 10 Header include build 1809+

[FranciscoPombal]

@sledgehammer999

As I said on my other comment there will proper news entries linking here along with the next releases. Also I created a forum topic. And I plan to announce to the IRC channel. Word will spread(hopefully).

A valiant effort, but I'm afraid that will not improve the situation enough. The reality is that as with all reasonably sized projects and communities, the vast, vast majority of users are simply pure users/lurkers that have never touched the issue tracker, let alone the forum, IRC, subreddit, or Discord.

@xavier2k6

Can same be done with sourceforge?

I think on another thread we had reached the conclusion that the number of downloads on SF was minimal and with a bias for more legacy users?

[xavier2k6]

I think on another thread we had reached the conclusion that the number of downloads on SF was minimal and with a bias for more legacy users?

Ah, You're right I believe.

[FranciscoPombal]

@sledgehammer999

@sledgehammer999 @FranciscoPombal I'd make the Windows 10 Header include build 1809+

I actually think this is a mistake - users that are not yet on 1809 will not be represented, despite having a very easy upgrade path. It is not necessary to split up Windows users like that IMO. This will create a bias for non-Windows 10.

[sledgehammer999]

A valiant effort, but I'm afraid that will not improve the situation enough. The reality is that as with all reasonably sized projects and communities, the vast, vast majority of users are simply pure users/lurkers that have never touched the issue tracker, let alone the forum, IRC, subreddit, or Discord.

You're impossible! Here you use the self-reported windows version on issues on Github as an example/gauge of windows usage among our users.

In any case, I believe that upon new releases users who install them go to our webpage to see the changelog. There they will read about the polls too. We can't have any other better way for poll exposure.

[sledgehammer999]

I actually think this is a mistake - users that are not yet on 1809 will not be represented, despite having a very easy upgrade path. It is not necessary to split up Windows users like that IMO. This will create a bias for non-Windows 10.

That's a good point. Going from one build to another is easier than going from Win7 to Win10. So I'll re-edit my initial post to remove the build numbers.

[sledgehammer999]

In any case, I believe that upon new releases users who install them go to our webpage to see the changelog. There they will read about the polls too. We can't have any other better way for poll exposure.

Also for poll validity: Judging from download stats we should have a couple hundred thousand voters to at least consider it a good representative sample.

[FranciscoPombal]

@sledgehammer999

You're impossible! Here you use the self-reported windows version on issues on Github as an example/gauge of windows usage among our users.

You have to take the context into account... I never claimed it was a good measure or that I was satisfied with it. It was only anecdotal evidence in response to your claim, which was not substantiated at all (All I am saying is that it is a sizable minority and we shouldn't ignore and drop support yet. - no source), and as a point of comparison with the other more-or-less flawed measures (statcounter and steam).

I believe data from fosshub is the only reasonable source for the data we want, that's why I'm insisting so much on it. All other alternatives are suboptimal for various reasons.

[FranciscoPombal]

@sledgehammer999

Also for poll validity: Judging from download stats we should have a couple hundred thousand voters to at least consider it a good representative sample.

OK, assuming a lot of users vote here when they see it on the news page, that is an acceptable second-best in case the fosshub data does not exist.

[sledgehammer999]

I believe data from fosshub is the only reasonable source for the data we want, that's why I'm insisting so much on it. All other alternatives are suboptimal for various reasons.

For clarity: I have asked them. I am waiting for a reply.

[xavier2k6]

I actually think this is a mistake

Honestly, I think by not having the build version is a mistake as we are informing the users that Qt 6 will not support any Windows OS below Windows 10 1809 (We are being honest & upfront)

I've said my piece & you can decide what's best.

[FranciscoPombal]

@xavier2k6

Honestly, I think by not having the build version is a mistake as we are informing the users that Qt 6 will not support any Windows OS below Windows 10 1809 (We are being honest & upfront)

I've said my piece & you can decide what's best.

The issue is that some users would not be represented in the poll. And when the time comes, it's a very small amount of effort for those who are not already on 1809 to upgrade to it. Plus it's not like they don't already have an incentive to upgrade - those versions are EOL and Windows and other applications are surely already notifying them to upgrade at this point, or lose features/functionality that they need.

[ArcticGems]

Maybe show this poll on the homepage as well?

[thalieht]

Uhh i just voted three times... this is pointless. One overzealous user to keep win 7 support and it's over.

[sledgehammer999]

Uhh i just voted three times

Were all 3 counted? For me it doesn't update the numbers even if it tells me "thank you for voting" again and again.

[thalieht]

Were all 3 counted?

No... Someone else must have voted at the same time, sorry.

[FossHub]

Hey guys,

Although I emailed @sledgehammer999 30 minutes ago and explained to him, we found a quick solution. We will send the additional data that you guys are looking for in a few minutes. For privacy, I will send the data to him and choose to share them with the other team members.

Thank you!

[sledgehammer999]

Here is the breakdown that I was informed with.

Period: January 2020 - March 2021 Unique visitors: 11,099,388 OS version	Percentage
Windows 10	86.00%
Windows 7	10.74%
Windows 8.01	2.46%
Windows XP	00.37%
Windows 8	00.30%
Windows Vista	00.13%
Windows Server 2003	00.01%
Windows NT	00.00%
Windows 2000	00.00%
Windows 98	00.00%

[sledgehammer999]

This can be closed now. And thanks to @FossHub for the quick response.