...one other thing: looks like the main page pageviews on Hebrew mobile web Wikipedia is (assuming the correct URL) 18.8K in the previous day's log.
(Note 1% sampling here, so multiple times 100.) $ zcat mobile-sampled-100.tsv.log-20140808.gz | cut -f9 | grep ' http://he.m.wikipedia.org/wiki/%D7%A2%D7%9E%D7%95%D7%93_%D7%A8%D7%90%D7%A9%D7%99' | sort | uniq -c 188 http://he.m.wikipedia.org/wiki/%D7%A2%D7%9E%D7%95%D7%93_%D7%A8%D7%90%D7%A9%D7%99 With a rsimplified definition of a pageview, that's maybe something like 3% of total pageviews on Hebrew mobile web Wikipedia. $zcat mobile-sampled-100.tsv.log-20140808.gz | cut -f9,11 | grep ' http://he.m.wikipedia.org/wiki/' | cut -f2 | grep -i 'text\|json' | grep -iv javascript | sort | uniq -c 4 text/html; charset=iso-8859-1 80 text/html; charset=utf-8 5918 text/html; charset=UTF-8 -Adam On Fri, Aug 8, 2014 at 12:09 PM, Adam Baso <[email protected]> wrote: > Some thoughts inline. > > > On Thu, Aug 7, 2014 at 5:07 PM, Amir E. Aharoni < > [email protected]> wrote: > ... > >> I don't have a strong opinion myself, but the current tendency is to show >> all the boxes rather than just a few, as it is now. >> >> This begs the question - is it really good for mobile readers? >> > > Depending on user agent support for it, text-overflow adjusted with media > queries for screen dimensions may be one way to tighten up stuff if screen > real estate is the issue. > > For example, is the mobile main page very popular at all, or is there much >> more traffic to the articles? >> > > I don't know about main page pageviews relative to total pageviews, as the > formal definition of a pageview is I think being refined, but it looks > like, for example, en.m.wikipedia.org/wiki/Main_Page was accessed about > 9.9 million times in the prior day's log. [1] > > >> Is there a difference between mobile main page traffic in the app and in >> the mobile web? >> > > Mobile web main page pageviews are proportionally larger than mobile app > traffic from what I understand, and that seems to be the case at first > glance. [2] In the Wikipedia for Android app, startup behavior is to go to > the main page for the language of the user (first time use has an account > splash screen). In the Wikipedia for iOS app the startup behavior > (post-first time use account splash screen) is to go to the main page for > the language of the user, and subsequent app launches go to the last > reading location. The user can change the language from the in-app > settings. Both the Android and iOS app now have a "Today" button in the > navbar now (like the "Home" button on the mobile web), too, to get the user > back to the main page in the current language. > > >> Are there any known good practices for mobile main page design? >> > > One is speed. Careful inlining of some ResourceLoader stuff and getting > the additional *main page* <img> tags pointing at the same domain as the > HTML of the main page might help. > > -Adam > > [1] $ zcat mobile-sampled-100.tsv.log-20140808.gz | cut -f9 | grep ' > http://en.m.wikipedia.org/wiki/Main_Page' | sort | uniq -c > Then multiply times 100 for the specific row. > > [2] A hackish grep, with the result * 100 suggests about 394K Main Page > loads, assuming I didn't mess up. > > $ zcat mobile-sampled-100.tsv.log-20140808.gz | cut -f9,14 | grep > WikipediaApp | grep 'wikipedia.org/w/api.php?' | grep > 'page=Main+Page\|Main%20Page' | grep -c 'sections=0' > >
_______________________________________________ Mobile-l mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/mobile-l
