Closed Bug 426142 Opened 17 years ago Closed 7 years ago

Awesomebar does not learn from HTTP 301 Moved Permanently responses

Categories

(Firefox :: Bookmarks & History, defect)

defect
Not set
minor

Tracking

()

RESOLVED WORKSFORME

People

(Reporter: tom, Unassigned)

References

Details

Attachments

(2 files)

Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9pre) Gecko/2008033004 Minefield/3.0pre Repro: 1) Make a new profile 2) Load www.reddit.com 3) Notice you are at http://reddit.com 4) type "r" into awesomebar 5) Notice suggested domain is http://www.reddit.com Expected Results: It learns the 301'd URL Actual Results: It doesn't
Redirected URL can be completely different from the one you typed in urlbar. For example, if you typed abc.org and you'll be redirected to xyz.com, if Firefox stores xyz.com instead of abc.org, when you type 'a' in urlbar you'll not find nothing. INVALID?
And if I go to http://tinyurl.com/2ndfb I'd want it to not save tinyurl.com, but mozilla.org... VALID? (unlike your disposable e-mail address)
If you notice, at the address bar you can retrieve tinyurl.com/2ndfb AND mozilla.org. And when you go to www.reddit.com, http://www.reddit.com AND http://reddit.com are saved. The history must save what you type and what you use, and prioritize what you use most. Not saving an url because it redirects to another address doesn't seem to me a smart idea, because you can remember that link but not the redirected one. At the most Firefox could lower the priority of redirect URLs.
Keywords: qawanted
Whiteboard: [wontfix?]
Whiteboard: [wontfix?] → [wontfix? see comment 3]
This seems to be an interesting example, because either the bug reporter was mistaken, or Reddit swapped their URLs. http://reddit.com/ currently redirects to http://www.reddit.com/. Either way, though, this is a valid complaint, so I'm confirming it, at least to continue discussion on the matter. See also bug 452267.
Status: UNCONFIRMED → NEW
Ever confirmed: true
I don't think I've mistaken what the reporter said. He said www.reddit.com must not be saved, as it redirected to reddit.com. See also comment 2. The fact now reddit.com redirects to www.reddit.com now doesn't change nothing. I don't think it's a good idea at all to don't save permanent redirects, because an user could be accustomed to use the redirect link. As you said in Bug 452267, sf.net redirects to sourceforge.net. But why Firefox mustn't save sf.net? An user could be accustomed to type 'sf'. If you type 'sf', sourceforge.net is not retrieved. Anyway, is it so much important? How much trouble can cause redirect links in location bar search results?
(In reply to comment #5) > An user could be accustomed to type 'sf'. If you type 'sf', > sourceforge.net is not retrieved. I believe you misread his bug. He is saying that he is going to a site, like lilurl.sf.net, and he doesn't want results for both lilurl.sf.net and lilurl.sourceforge.net in the results, as they are duplicates.
So are you proposing to not display redirect url, if the search string matches both of them? If I understood well, it seems reasonable. But IMO it will requests additional db space and search time for a inessential problem. (Anyway I insist i've not misunderstood noting. As I've written before, see comment 2: > And if I go to http://tinyurl.com/2ndfb I'd want it to not _save_ tinyurl.com, > but mozilla.org... )
(In reply to comment #5) > I don't think I've mistaken what the reporter said. He said www.reddit.com must > not be saved, as it redirected to reddit.com. See also comment 2. The fact now > reddit.com redirects to www.reddit.com now doesn't change nothing. I never said you were mistaken. I never even addressed anything that you said. > I don't think it's a good idea at all to don't save permanent redirects, > because an user could be accustomed to use the redirect link. As you said in > Bug 452267, sf.net redirects to sourceforge.net. But why Firefox mustn't save > sf.net? An user could be accustomed to type 'sf'. If you type 'sf', > sourceforge.net is not retrieved. In these cases, you wouldn't be typing "sf"; you'd be typing whatever the subdomain is. So the difference between "sf" and "sourceforge" wouldn't matter to you when you type, but they would matter to you, when you have to wait, and the server, that has to do the redirect, every time. But the discussion about interpreting 302/303/307 and Location (i.e. temporary) redirects should be had in bug 452267, not this bug, which is for 301 (permanent) redirects. (Incidentally, http://reddit.com/ is a 302 temporary redirect.) (In reply to comment #6) > (In reply to comment #5) > > An user could be accustomed to type 'sf'. If you type 'sf', > > sourceforge.net is not retrieved. > > I believe you misread his bug. He is saying that he is going to a site, like > lilurl.sf.net, and he doesn't want results for both lilurl.sf.net and > lilurl.sourceforge.net in the results, as they are duplicates. > Actually, the full URL doesn't even show up in the results. But again, that discussion is for the other bug.
(In reply to comment #8) > In these cases, you wouldn't be typing "sf"; you'd be typing whatever the > subdomain is. So the difference between "sf" and "sourceforge" wouldn't matter > to you when you type, but they would matter to you, when you have to wait, and > the server, that has to do the redirect, every time. > I can agree with you, but if you don't want that an user re-uses a redirect link, you must prevent Firefox to save it - and this is highly undesirable IMO - since if he type "sf" the only result will be "sf.net", and not "sourceforge.net". IMO the best compromise is what I wrote in comment 7. (incidentally, I got confused indeed in the second part of that comment, sorry) > But the discussion about interpreting 302/303/307 and Location (i.e. > temporary) redirects should be had in bug 452267, not this bug, which is for > 301 (permanent) redirects. > What we are writing is valid also for this bugs, even if the examples are wrong. > I never said you were mistaken > Ok, sorry.
Keywords: qawanted
(In reply to comment #9) > What we are writing is valid also for this bugs, even if the examples are > wrong. I agree, the scope of this bug could be easily expanded, however I don't really like learning from temporary redirects because those can always change. At the very least, if a 301 redirect is hit, the redirect-or could be listed as second, and whichever url it redirects to (redirect-ee) would be given priority (and listed in the results dropdown right above the redirect-or).
(In reply to comment #1) > Redirected URL can be completely different from the one you typed in urlbar. > For example, if you typed abc.org and you'll be redirected to xyz.com, if > Firefox stores xyz.com instead of abc.org, when you type 'a' in urlbar you'll > not find nothing. > > INVALID? correct, we want inclusion not exclusion here, such that the search is successful whether you remember the source or the target of a redirect. the penalty of the occasional duplicate is small price to pay for actually finding what you're searching for. (In reply to comment #3) > The history must save what you type and what you use, and prioritize what you > use most. Not saving an url because it redirects to another address doesn't > seem to me a smart idea, because you can remember that link but not the > redirected one. At the most Firefox could lower the priority of redirect URLs. gmail is a good example of this: gmail.com redirects to mail.google.com/mail. if you didn't save the source url and then typed "gmail" into the awesomebar, you'd get no match... -> WONTFIX
Status: NEW → RESOLVED
Closed: 16 years ago
Resolution: --- → WONTFIX
Whiteboard: [wontfix? see comment 3]
(In reply to comment #11) > gmail is a good example of this: gmail.com redirects to mail.google.com/mail. > > if you didn't save the source url and then typed "gmail" into the awesomebar, > you'd get no match... I don't appreciate flat out lies. I've never been to gmail.com, and this screenshot shows what happens when I type "gmail" into the awesomebar.
Also, since I didn't notice the first time (and do regret commenting before noticing): Yes, Gmail itself is bookmarked. However, the login page for gmail you're redirected to when going to gmail.com (which I just did for the first time) contains "Gmail" in the title element. The example, going to http://reddit.com and getting sent to http://www.reddit.com, would also work just fine as the user would be typing 'reddit'.
Yes I was mistaken, for gmail it would match on title for the target url of the redirect. However, this is not always the case.
I too wish something could be done about this. See also my bug 654711. I'm attaching a screenshot of what my address bar dropdown looks like (the first two entries) when I type "redd". A couple of notes: 1. The first time I went to reddit prior to taking the screenshot, I typed reddit.com/r/gaming (without www), not www.reddit.com/r/gaming (with www). Reddit, of course, redirects me automatically to www.reddit.com/r/gaming. 2. Since I actually typed the www-less entry into the awesomebar, then even if I browse reddit quite a bit and click a bunch of links (leading to the reddit domain) on the page, the www-less entry stays high on the list (after typing "redd" in the bar), since I actually typed it into the bar, whereas most other links on reddit I simply click (and it seems that the awesomebar places these clicked links lower on the dropdown list than the ones that were explicitly typed into the bar, which are placed higher automatically). So even if I click a certain link many times on reddit that leads to a page still on the reddit.com domain, the www-less "reddit.com/r/gaming" entry without the favicon and title still stays at the #2 position in the dropdown when I type "redd" in the future. 3) This doesn't happen only with reddit, it happens with all sites that automatically redirect from site.com to www.site.com (and a lot of sites do this). It just looks kind of ugly and seems pointless to always see essentially duplicate entries right above/below each other, especially since one of them doesn't have any favicon or title. One other thing: I believe that Firefox (which I love) is the only browser I've experienced this on. Chrome, IE, and Opera all seem to hide from view in the address bar drop down those www-less URLs that automatically redirect to the www "subdomain".
Dietrich, we should be able to still match on the original URL but return the one it redirects too. In particular, HTTP requires that for permanent redirects we store that information and not revisit the old link.
Status: RESOLVED → REOPENED
Resolution: WONTFIX → ---
Status: REOPENED → NEW
Component: Location Bar → Bookmarks & History
Bug 737836 which landed last year changed the redirect "frecency" values so that the target of the redirect gets a higher ranking than the originally entered source. That seems to be what is being asked for here, so I'm going to close this as works for me - likely by that bug, but possibly others over the years. Note there's still a few issues with redirects, e.g. bug 487813 and bug 468710, but I think the main part of this has been taken care of now.
Status: NEW → RESOLVED
Closed: 16 years ago7 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: