9331 2016-03-23 13:59:54 +0000 searching on google almost always returns some outdated link 2017-06-20 21:20:04 +0000 1 1 5 Infrastructure Website web-content Production x86 Multiple RESOLVED FIXED Medium enhancement Future 1 igor.stoppa mhalstead jefro melissa mhalstead randy.macleod --- --- No (bug/feature does not impact docs) oldest_to_newest 60340 0 igor.stoppa 2016-03-23 13:59:54 +0000 Precondition: Just search on google for anything YP related. Ex: search for "yocto megamanual" Expected Outcome: the link should point to the latest stable set of documents / manuals. This should be the first link: http://www.yoctoproject.org/docs/latest/mega-manual/mega-manual.html Actual Outcome: the link hits whatever was the idea google had of the highest-ranking version and this is a self sustaining issue: the more google reports outdated links, the more people will hit them and confirm that they are good In the example, this is the first link: www.yoctoproject.org/docs/1.8/mega-manual/mega-manual.html Effects: This is really annoying when one is looking for help/explanation and hits obsolete pages, missing the improvements added in the meanwhile. After a while one learns to always check the url or add latest to the search parameters, but it's definitely not friendly towards newbies. Possible solutions: Dynamically add an header to the older docs, stating that it is superseded by another and provide link to it. Collapse all the current versioned urls to the latest and introduce a different url scheme for older versions (ex: add "archived" in the url). 60341 1 igor.stoppa 2016-03-23 14:16:35 +0000 Correction: I was wrong about "latest": even if I add "latest" to the search string, I'm still presented with the wrong link, so the situation is actually worse than what I described. 62155 2 jefro 2016-05-12 17:12:18 +0000 Michael - can we tell google not to only index the "latest" directory? thanks 74412 3 jefro 2017-06-20 21:20:04 +0000 fixed in the new website! going forward we will hide the older versions of documentation behind robots.txt and in an archive, so only the current set is indexed.