Re: [Tex-music] Archive stats observations and mysteries
> >Don Simons wrote: > >> I wrote >> >> >>..it looks like every single file in the archive has been downloaded at >> >> least once, including of course many that are not linked from any of the >> web >> >> pages. Any good ideas how or why that could happen? >> >> And Christian replied >> >> >Just a guess: the counts are file accesses rather than more specifically >> >downloads. If that is true any file has been accessed at least once when >> >it was uploaded. >> >> I doubt that, because practically all of these files were originally >> uploaded before you started including the .zip's in the count. I check some >> more to see if it really does include all of the files. > >far more likely it's a crawler "signal" (there are masses of them, and >they skew the stats of the tex faq something rotten). > >robin Well, if we knew this crawler (or any others) hit ALL the files, we could correct the counts by subtracting from all sums the (small) number reported for the many files that obviously no one would want. --Don --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Don Simons wrote: > I wrote > > >>..it looks like every single file in the archive has been downloaded at > >> least once, including of course many that are not linked from any of the > web > >> pages. Any good ideas how or why that could happen? > > And Christian replied > > >Just a guess: the counts are file accesses rather than more specifically > >downloads. If that is true any file has been accessed at least once when > >it was uploaded. > > I doubt that, because practically all of these files were originally > uploaded before you started including the .zip's in the count. I check some > more to see if it really does include all of the files. far more likely it's a crawler "signal" (there are masses of them, and they skew the stats of the tex faq something rotten). robin --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Don Simons wrote: I wrote ..it looks like every single file in the archive has been downloaded at least once, including of course many that are not linked from any of the web pages. Any good ideas how or why that could happen? And Christian replied Just a guess: the counts are file accesses rather than more specifically downloads. If that is true any file has been accessed at least once when it was uploaded. I doubt that, because practically all of these files were originally uploaded before you started including the .zip's in the count. I check some more to see if it really does include all of the files. There is a webalizer FAQ at http://www.webalizer.org/faq.html It's section 5 explains the difference between 'hits' and 'files': 'HITS is the total number of HTTP requests that the server received during the reporting period.' Maybe webalizer's own activities involve http requests. --Don Simons -- Christian Mondrup, Archive Editor WIMA: Werner Icking Music Archive http://icking-music-archive.org/ --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
I wrote >>..it looks like every single file in the archive has been downloaded at >> least once, including of course many that are not linked from any of the web >> pages. Any good ideas how or why that could happen? And Christian replied >Just a guess: the counts are file accesses rather than more specifically >downloads. If that is true any file has been accessed at least once when >it was uploaded. I doubt that, because practically all of these files were originally uploaded before you started including the .zip's in the count. I check some more to see if it really does include all of the files. --Don Simons --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Don Simons wrote: Thanks to Christian for adding the .zip files to the webalizer. Now, with help from the filtering capabilities of Excel, I can quickly whittle down the list of 100K+ files to just the 202 .zip's in the software directory or below. When I do that, a VERY curious result appears: I haven't checked for sure, but it looks like every single file in the archive has been downloaded at least once, including of course many that are not linked from any of the web pages. Any good ideas how or why that could happen? Just a guess: the counts are file accesses rather than more specifically downloads. If that is true any file has been accessed at least once when it was uploaded. What exactly counts as a download in these stats? BTW, there were only 11 such files with>1 download (in ~4 days), with musixtex.zip leading at 14, and musixflx.pkg.zip and pmx2613.zip tied for 2nd at 5. Curiously, the 3 bitmapped musixtex font files were each downloaded twice; I would have assumed those were completely obsolete. --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music -- Christian Mondrup, Archive Editor WIMA: Werner Icking Music Archive http://icking-music-archive.org/ --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Jean-Pierre Coulon wrote >On Sun, 26 Jun 2011, Don Simons wrote: > >> There are also 177 hits to pages in the memorial folder. As best I can >> determine, there are no remaining links on the main page to anything in that >> folder (I looked at the source and they seem to be commented out). Does >> anyone have any idea what's going on here? Could that many people have been >> both clever and inquisitive enough to look in the page's source? [...] > >A Google search with "http://icking-music-archive.org/Memorial"; gives 9 >results, most of them related to Daniel, e.g. >http://www.mail-archive.com/tex-music@sunsite.dk/msg02234.html >... Thanks to Christian for adding the .zip files to the webalizer. Now, with help from the filtering capabilities of Excel, I can quickly whittle down the list of 100K+ files to just the 202 .zip's in the software directory or below. When I do that, a VERY curious result appears: I haven't checked for sure, but it looks like every single file in the archive has been downloaded at least once, including of course many that are not linked from any of the web pages. Any good ideas how or why that could happen? What exactly counts as a download in these stats? BTW, there were only 11 such files with >1 download (in ~4 days), with musixtex.zip leading at 14, and musixflx.pkg.zip and pmx2613.zip tied for 2nd at 5. Curiously, the 3 bitmapped musixtex font files were each downloaded twice; I would have assumed those were completely obsolete. --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Christian Mondrup wrote: Jean-Pierre Coulon wrote: > > On Mon, 27 Jun 2011, Christian Mondrup wrote: > > > >>> /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/score > >>> s/ByComposer/Dowland.php > >>> > >> > >> Looks really weird. There are quite a few such cases. WIMA has no > >> such path; so I suppose URL accesses like that origin from buggy > >> links in other sites out there. > > > > But nevertheless this link works! > > Not for me. I get the usual 'not found' alert. > > With which OS and which browser do you get a working web page out of > accessing > /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByCo > mposer/Dowland.php? http://icking-music-archive.org/ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/Dowland.php returns a variant of the composers page where the stylesheet is broken - because the stylesheet uses a relative URL so I expect it tries to retrieve http://icking-music-archive.org/ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/ick-style.css instead of http://icking-music-archive.org/ick-style.css Google Chrome / Windows 7 ... David --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
On Mon, 27 Jun 2011, Christian Mondrup wrote: Not for me. I get the usual 'not found' alert. With which OS and which browser do you get a working web page out of accessing /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/Dowland.php? Sorry, I omitted to include http://icking-music-archive.org before it! I have Firefox. It works witj Internet explorer too. -- Jean-Pierre Coulon cou...@obs-nice.fr --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Jean-Pierre Coulon wrote: > On Mon, 27 Jun 2011, Christian Mondrup wrote: > > >> /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores > >> /ByComposer/Dowland.php > > > > Looks really weird. There are quite a few such cases. WIMA has no such > > path; so I suppose URL accesses like that origin from buggy links in > > other sites out there. > > But nevertheless this link works! I obtain a variant of the ByComposer.php > page. More likely a faulty script call sequence (somewhere scores/ByComposer is appended to the existing URL to filter it and via a particular sequence that's happened several times). The fact it returns a page suggests that there's a different bug in the PHP script for the page not validating its URL correctly. David --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Jean-Pierre Coulon wrote: On Mon, 27 Jun 2011, Christian Mondrup wrote: /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/Dowland.php Looks really weird. There are quite a few such cases. WIMA has no such path; so I suppose URL accesses like that origin from buggy links in other sites out there. But nevertheless this link works! Not for me. I get the usual 'not found' alert. With which OS and which browser do you get a working web page out of accessing /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/Dowland.php? I obtain a variant of the ByComposer.php page. Bye, -- Christian Mondrup, Archive Editor WIMA: Werner Icking Music Archive http://icking-music-archive.org/ --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
On Mon, 27 Jun 2011, Christian Mondrup wrote: /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/Dowland.php Looks really weird. There are quite a few such cases. WIMA has no such path; so I suppose URL accesses like that origin from buggy links in other sites out there. But nevertheless this link works! I obtain a variant of the ByComposer.php page. Bye, -- Jean-Pierre Coulon cou...@obs-nice.fr --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Jean-Pierre Coulon wrote: On Sun, 26 Jun 2011, Don Simons wrote: Other topic: who can explain such a line: 1 0.00% 127 0.00% /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/Dowland.php Looks really weird. There are quite a few such cases. WIMA has no such path; so I suppose URL accesses like that origin from buggy links in other sites out there. in the huge View All URLs page? http://icking-music-archive.org/usage/url_201106.html -- Christian Mondrup, Archive Editor WIMA: Werner Icking Music Archive http://icking-music-archive.org/ --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
On Sun, 26 Jun 2011, Don Simons wrote: There are also 177 hits to pages in the memorial folder. As best I can determine, there are no remaining links on the main page to anything in that folder (I looked at the source and they seem to be commented out). Does anyone have any idea what's going on here? Could that many people have been both clever and inquisitive enough to look in the page's source? [...] A Google search with "http://icking-music-archive.org/Memorial"; gives 9 results, most of them related to Daniel, e.g. http://www.mail-archive.com/tex-music@sunsite.dk/msg02234.html Once you've reached http://icking-music-archive.org/Memorial/Taupin/Statements.html does it require a tremendous intelligence to try out the /Memorial and the /Memorial/Taupin pages? Other topic: who can explain such a line: 1 0.00% 127 0.00% /ByComposer.php/scores/ByComposer/scores/ByComposer/ByComposer/scores/ByComposer/Dowland.php in the huge View All URLs page? http://icking-music-archive.org/usage/url_201106.html -- Jean-Pierre Coulon cou...@obs-nice.fr --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
Don Simons wrote: I've been browsing the stats for WIMA, available to all at http://www.icking-music-archive.org/usage/ In June, through 6/26, of 1366924 recorded hits, only 8075 or 0.59% were to the software pages. I guess I'm not surprised at that ratio; I can believe that 170 times more people want to download free scores rather than typeset music. But as best I could determine, there were no hits at all recorded for any .zip files, including of course musixtex.zip, pmx250.zip, and pmx2610.zip. I can't believe that no one at all downloaded any of those files. Does anyone have any idea what's going on here? Does the software simply not record hits to zip files? Or maybe not to zip.files if they are "save"d rather than "open"ed? Way back I configured webalizer to count accesses on a limited number of file types: IgnoreURL * IncludeURL *.pdf IncludeURL *.php IncludeURL *.html Now I've included .zip files into the count. IncludeURL *.zip But it'll take some time until the figures for zip files get comparable to those for the other 3 file types There are also 177 hits to pages in the memorial folder. As best I can determine, there are no remaining links on the main page to anything in that folder (I looked at the source and they seem to be commented out). Does anyone have any idea what's going on here? Could that many people have been both clever and inquisitive enough to look in the page's source? I believe the counts for the memorial pages are caused by someone out there linking to those pages. Even if the WIMA links have been commented out the content they're referring to is still there. I know at first I wasn't; I first found the missing link through my server access to the archive. I have gone there several times and did give one other person that link, but it doesn't seem like that could have accounted for so many hits. --Don Simons --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music -- Christian Mondrup, Archive Editor WIMA: Werner Icking Music Archive http://icking-music-archive.org/ --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
Re: [Tex-music] Archive stats observations and mysteries
I wouldn't be surprised if search robots would also scan comments in web pages. Easy to find out: just try to hit such a "hidden" page by searching for relevant keywords in the page on e.g. Google... -- Peter Vanroose, Leuven, Belgium. Don Simons wrote: > I've been browsing the stats for WIMA, available to all at > http://www.icking-music-archive.org/usage/ > > In June, through 6/26, of 1366924 recorded hits, only 8075 or 0.59% were to > the software pages. I guess I'm not surprised at that ratio; I can believe > that 170 times more people want to download free scores rather than typeset > music. But as best I could determine, there were no hits at all recorded for > any .zip files, including of course musixtex.zip, pmx250.zip, and > pmx2610.zip. I can't believe that no one at all downloaded any of those > files. Does anyone have any idea what's going on here? Does the software > simply not record hits to zip files? Or maybe not to zip.files if they are > "save"d rather than "open"ed? > > There are also 177 hits to pages in the memorial folder. As best I can > determine, there are no remaining links on the main page to anything in that > folder (I looked at the source and they seem to be commented out). Does > anyone have any idea what's going on here? Could that many people have been > both clever and inquisitive enough to look in the page's source? I know at > first I wasn't; I first found the missing link through my server access to > the archive. I have gone there several times and did give one other person > that link, but it doesn't seem like that could have accounted for so many > hits. > > --Don Simons --- TeX-music@tug.org mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music