Announcement

Collapse
No announcement yet.

Taking away the number in the urls?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • medved
    commented on 's reply
    But it was there. Other people say it was ok.

  • Wayne Luke
    replied
    Originally posted by medved View Post
    In the first version of vB 5 I was told it was working fine. Later somebody decided to remove it for some unknown reason. Just put it back. Simple.
    It was never working fine in the first version of vB 5.

    Leave a comment:


  • medved
    replied
    Originally posted by Wayne Luke View Post

    I mean in the Address bar of the browser. Showing the entities in the text of the post is a completely different issue.

    We've actually put quite a bit of work in resolving this issue but don't have a publishable solution yet. The primary issue is upgraded databases because we actually don't know what the content in that database will look like.
    You had that already working fine in the first version of vB 5. Just bring it back please. That's all. Simple.

    Leave a comment:


  • medved
    replied
    Originally posted by Mark.B View Post
    It's important to note that "fixing" this involves a major refactor of the existing routes system. It's not a trivial matter and would probably take many months of work to get it in the product. That will have a bearing on whether it is ever changed or not.
    In the first version of vB 5 I was told it was working fine. Later somebody decided to remove it for some unknown reason. Just put it back. Simple.

    Leave a comment:


  • Wayne Luke
    replied
    Originally posted by medved View Post
    OK, so it is a lousy "tool" then. The actual entities shows in any browser. No reason to discuss it any further as I know nobody cares enough to fix it. I mentioned it here long time ago and nobody cares:
    https://www.vbulletin.com/forum/foru...A4%C3%B4%C5%88
    I mean in the Address bar of the browser. Showing the entities in the text of the post is a completely different issue.

    We've actually put quite a bit of work in resolving this issue but don't have a publishable solution yet. The primary issue is upgraded databases because we actually don't know what the content in that database will look like.

    Leave a comment:


  • Mark.B
    replied
    It's important to note that "fixing" this involves a major refactor of the existing routes system. It's not a trivial matter and would probably take many months of work to get it in the product. That will have a bearing on whether it is ever changed or not.

    Leave a comment:


  • medved
    replied
    OK, so it is a lousy "tool" then. The actual entities shows in any browser. No reason to discuss it any further as I know nobody cares enough to fix it. I mentioned it here long time ago and nobody cares:
    https://www.vbulletin.com/forum/foru...A4%C3%B4%C5%88

    Leave a comment:


  • Wayne Luke
    replied
    What browser are you using that shows the actual entities? I see the characters in Chrome. However, this site is not using UTF-8 so there can be problems still.

    Route by Node isn't a solution. It is so people can access nodes and rename them if they have broken URLS. It was never meant to be a solution. Just a tool to help troubleshoot things.

    Leave a comment:


  • medved
    replied
    If that "ROUTE BY NODE" is so good solution, why this website is not using it? Look how this link is displaying:

    vbulletin.com/forum/forum/vbulletin-5-connect/vbulletin-5-connect-feedback/4327340-%C4%8D%C4%BE%C2%9A%C4%8D%C5%A5%C2%9E%C3%BD%C3%A1%C3%AD%C3%A9%C2%9E%C3%BA%C3%A4%C3%B4%C5%88-%C4%8D%C4%BE%C2%9A%C4%8D%C5%A5%C2%9E%C3%BD-%C3%A1%C3%AD%C3%A9%C2%9E%C3%BA%C3%A4%C3%B4%C5%88

    with properly done urls you should see this instead:

    vbulletin.com/forum/forum/vbulletin-5-connect/vbulletin-5-connect-feedback/4327340-clsctzyaiezuaon-clsctzy-aiezuaon

    ROUTE BY NODE is a lousy solution that is not helping at all.

    Leave a comment:


  • Mark.B
    replied
    Originally posted by Lingo View Post
    By the way, I just remembered how WP handles the problem you described: if you have a buy-pear article and then you create another article named always buy-pear then the second one would be name buy-pear-1 , this way you got the number in the url only if necessary.
    Problem with that method is, you need some means to check ALL the existing thread titles when submitting a new thread.

    On a site with a few thousand threads....not too big an issue.
    On a site with millions of threads...good luck!

    Leave a comment:


  • Wayne Luke
    replied
    Originally posted by Lingo View Post

    So you mean that for chinese, arabic ect. it's better to put a number since some browsers might not read them? please wait, since I have a language forum which would be multilingual, I just opened a topic with chinese characters and as you can see works fine, or, I'm not sure I understood you, should I expect some issues, even though rare? I know this was a problem of the past, with IE, which now is all but dead, but I just checked and now all Browsers including the last versions of IE/Edge read them. If a small percentage might have issues I prefer this that forego the benefts of of a seo url.
    New installs work pretty well. If possible, the install script will even create the database with a UTFMB4 character set so it will eventually support all the Emojis on your phone.

    However, you should try it on a 10 year old database that is using a latin character set but contains UTF-8 characters. That is what many of our customers are dealing with. It is the customers that have been with the company 10 or more years and we have had to use many different workarounds to overcome limitations of technology that have the problems. Those sites have pages of information that number in the 100s of millions.

    Leave a comment:


  • Lingo
    replied
    Of course I did not mean to trick Google, but just have a little boost (albeit a small one) in the SERPs. You also have to take into account the user experience, always better with a friendly url. True keywords in the url are not so important like in the past, but for me, for instance, I got more than 1 billion users to tackle from Baidu, and Baidu i guess still does counts more keys. By the way, I just remembered how WP handles the problem you described: if you have a buy-pear article and then you create another article named always buy-pear then the second one would be name buy-pear-1 , this way you got the number in the url only if necessary.

    Leave a comment:


  • Mark.B
    replied
    If it works you don't have to worry. Different servers can have different problems. It's not all absolute.

    In terms of keywords in URLs, they are grossly overstated. It may have had some merit many years ago, although plenty of sites, mine included, have run for over a decade using showthread.php?t=xxx which was the only format for vBulletin 3, and search engines found us just fine.

    Google have long since stopped indexing purely on keywords, they look at the content. If the keyword references something that it's not in the content, it'll be ignored. If the word is in the content but not in the URL, it'll get indexed. There's this myth that persists that you can 'fool' Google into ranking pages higher using trickery. You can't, at least not for any length of time.

    Leave a comment:


  • Lingo
    replied
    Originally posted by Wayne Luke View Post

    Not all characters in the UTF-8 library translate into routable URLs. The main problem is that browsers required these characters to be encoded. Some decode them for display. Some do not. Sometimes the URL is simply not usable and changing a character can do so. The "Route By Node ID" function is so that users can fix these URLs. However some customers using languages like Arabic, Hebrew, Chinese, Japanese, and Cyrillic find the functionality useful enough to leave it on all the time.
    So you mean that for chinese, arabic ect. it's better to put a number since some browsers might not read them? please wait, since I have a language forum which would be multilingual, I just opened a topic with chinese characters and as you can see works fine, or, I'm not sure I understood you, should I expect some issues, even though rare? I know this was a problem of the past, with IE, which now is all but dead, but I just checked and now all Browsers including the last versions of IE/Edge read them. If a small percentage might have issues I prefer this that forego the benefts of of a seo url.
    Last edited by Lingo; Tue 28 Nov '17, 3:08pm.

    Leave a comment:


  • Wayne Luke
    replied
    Originally posted by Lingo View Post
    This of course I woud never do,I don't know why someone would want urls like these I don't think it's good to lose all the keys.
    Not all characters in the UTF-8 library translate into routable URLs. The main problem is that browsers required these characters to be encoded. Some decode them for display. Some do not. Sometimes the URL is simply not usable and changing a character can do so. The "Route By Node ID" function is so that users can fix these URLs. However some customers using languages like Arabic, Hebrew, Chinese, Japanese, and Cyrillic find the functionality useful enough to leave it on all the time.

    Leave a comment:

widgetinstance 262 (Related Topics) skipped due to lack of content & hide_module_if_empty option.
Working...
X