Page 2 of 4

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 12:44 am
by louis94
Arbogast wrote:Soucerforge? Windows binaries? Can you expand on this?
What does Sourceforge offers me?

Software can be distributed in two forms: source code and binary (.exe) format. On Sourceforge, there are Windows binaries and source code. No Linux binary. This is because all Linux distributions (think versions, like Vista/Seven/8) use packages (much like the Android app store). The people making distributions download the source code from Sourceforge, compile it to make a binary and put the result on their own server. Then users download the binaries from that server.
Most distributions don't provide statistics on how many times a package has been downloaded. Debian provides some statistics, and the result is that about 1% of all Linux users have Freeciv installed. Accounting for ~20 million Linux users worldwide, that makes 200,000 Freeciv users. There have "only" been 100,000 downloads of the Windows client this year.

Arbogast wrote:I stand corrected, but still, can we -Windows players- could get a little more attention? Maybe I'm crying at a blue moon...

That's always the same story: many developers don't like Windows and use Linux instead. Then their users ask for Windows support. They do their best, but restarting your computer takes a while when Windows Update decides you have time for updates (and when you start Windows once a week/month/few months, there are always updates). Add to that the poor (free) development tools on Windows, and you get a picture of what's going on.
I don't think Windows users get less attention than Linux ones. Most problems aren't specific to Windows (except some encoding issues, or bugs from other projects Freeciv uses). The Windows-specific ones are taken care of. I can't see a way Windows users get less love. Maybe you can be more specific ?

Louis

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 1:59 am
by cazfi
There's also a tiny matter of money. I don't own Windows machine. I wouldn't need one for anything else than for development of freeciv windows port. So unfortunately Windows specific bugs are not as easily investigated as those bugs that affect also Linux (but despite Windows having some really weird quirks, bugs common to all ports make over 95% of the lot anyway).

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 2:01 am
by cazfi
Oh, and one big category of Windows problems are due to Windows specific gtk -bugs, for which we can do almost nothing. Gtk is a third party component we only use, not develop. http://freeciv.wikia.com/wiki/Gtk_minefield

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 2:53 pm
by Arbogast
Gentlemen, thank you. I get a better overall picture now. So, it's Linux that rules, fine.

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 4:11 pm
by AndreasR
Arbogast wrote:Gentlemen, thank you. I get a better overall picture now. So, it's Linux that rules, fine.


Freeciv-web had 412 009 sessions by 251 626 users last year. The Freeciv desktop clients had 150 274 downloads from sourceforge in the same period. So your statement above is incorrect.

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 4:25 pm
by cazfi
AndreasR wrote:
Arbogast wrote:Gentlemen, thank you. I get a better overall picture now. So, it's Linux that rules, fine.


Freeciv-web had 412 009 sessions by 251 626 users last year. The Freeciv desktop clients had 150 274 downloads from sourceforge in the same period. So your statement above is incorrect.


What do you mean by "users" in context of freeciv-web? Unique IPs or usernames taken?

Not that I see your logic in comparing freeciv-web sessions to windows desktop client downloads when talking about Linux usage.

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 4:30 pm
by cazfi
...and yes, I'm working on freeciv-web related changes as we speak.

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 4:37 pm
by AndreasR
cazfi wrote:What do you mean by "users" in context of freeciv-web? Unique IPs or usernames taken?

The number of unique users measured by Google Analytics is a very commonly used method for determining the number of users, and is sufficiently reliable in my opinion.
It is for example documented in the "Sessions vs. Users" section here: https://support.google.com/analytics/answer/1257084?hl=en

Re: Freeciv usage statistics for 2014

Posted: Wed Jan 21, 2015 4:38 pm
by AndreasR
cazfi wrote:...and yes, I'm working on freeciv-web related changes as we speak.


Glad to hear that! 8-)

Re: Freeciv usage statistics for 2014

Posted: Sun Jan 25, 2015 3:01 pm
by dunnoob
cazfi wrote:unfortunately Windows specific bugs are not as easily investigated as those bugs that affect also Linux
Naive thought, something working under Wine should work with any Windows, or do I miss too many clues at once? And if that's the case (obviously it must be the case, but I don't know why), what about a ReactOS VM? Disclaimer, I haven't tested ReactOS, I only have a Windows 2000 VM. and that's not "free" for distribution.