You should not use 0.0.0.0 in this way. Even if it has worked. Reason: You're bound to bump into troubles.
Using 0.0.0.0 for HTTP communication is a violation of a standard, which officially classifies 0.0.0.0 as an "unusable" address. The term "unusable address" in IPv4 refers to the first address and the last address of any network or subnet. 0.0.0.0 is the first address of every IPv4 subnet that it is a part of.
Having the first address be treated special is discussed (using various ways to specify the notation) by at least RFC 894, RFC 919, RFC 1122, and RFC 3021. The basic reason that it was declared to be "unusable" is that it was treated as a broadcast address. Since then, the more common IPv4 standard is that the last address is a broadcast address, but still, the first address continues to be widely recognized as "unusable". I'm sorry, @harrymc , that I must disagree with comment based on jargon. Even if the address can technically be used, and therefore fits the definition of the English word "usable", there are some technical terms in the networking world, of a "usable address" and an "unusable address". Based on this industry standard jargon, 0.0.0.0 is classified as "unusable".
Perhaps there is no technical reason why the first address must be unusable. This is not the result of a restriction forced by some limit of how many numbers can fit within some group. However, it is a standard by consensus. Rightly or wrongly, that is the reality. People are trained to avoid using this address. As a result, many pieces of software may check to make sure that this standard rule is not being violated.
If you spend enough time getting advanced in networking, you find that a lot of modern networking was designed around the concept of agreeableness. Therefore, consensus is important.
If you wish to try to violate this rule, you're welcome to try. There's lots of "open source" software that you can adjust to your whim. However, if you knowingly violate common standards (or refuse to adapt after learning that you were unknowingly violating common standards), you should not be surprised if you end up with certain troubles. For instance: if the authors of a major web browsing software does not treat an "unusable" address like a normal address.
Besides being unusable, 0.0.0.0 is filled with zeros, and makes it seem very similar in concept to "::" which is officially the "unspecified" address in IPv6. Therefore, treating 0.0.0.0 as the "unspecified" address in IPv4 seems rather sensible, and is consistent with how the netstat command (especially with "netstat -na") does things (in both Unix and Microsoft Windows).
In that sense, DHCP's usage of 0.0.0.0 makes a lot of sense; before an address is assigned, the computer is using the "unspecified" address. Actually, what I suspect the computer is doing is relying on the MAC address of the frame (Layer 2: Ethernet/Wi-Fi) instead of the IPv4 address (Layer 3) so that the right computer is getting the right address.
Some people may think of 0.0.0.0 as referring to the concept of "any" address, perhaps largely since "netsat -na" seemed to do that. If you have some idea of what you want 0.0.0.0 to mean, know that there are some other ideas out there, which are rather widespread. Conflicts and other such troubles are expected for trying a different approach.
Speaking of netstat, you may find that "netstat -nr" shows a routing table entry for "127.0.0.0/8" or "127.0.0.1/8" (IPv4 subnet mask 255.255.255.0), and other network addresses that you have assigned to your network card, but not "0.0.0.0/32" (IPv4 subnet mask 255.255.255.255). Although, you may see that "0.0.0.0/0" (IPv4 subnet mask 0.0.0.0) is set to send traffic to your IPv4 default gateway. That's standard subnetting and routing at work. If you haven't been trained in such techniques, you're best off to not be trying to invent a new set of personal expectations for how 0.0.0.0 ought to work.
Finally, to address Google Omnibox: I have no idea about how that works. However, any usage that relies on 0.0.0.0 is so unstandard that I hereby predict that the software developer (Google, in this case) might make changes at any time. And if they do make changes at any time, they are likely to tick off only a statistically insignificant amount of people. So, your best bet is to not rely on any special way that software currently treats the 0.0.0.0 address. It may be that Google is searching for the text because Google is treating the input (what you typed in the address bar) as invalid text-- something that Google isn't recognizing as a valid address. That's rather what I would expect many web browsers to do.
I do understand the pain you might feel about "it worked before; why did it break?" That's understandable. It may be that the previous way of doing things broke. That's something that you risk whenever doing something that isn't standard. As I mentioned before, since no working behavior is clearly expected by any universal standard, "custom" ("non-standard") behavior can be expected to break at any time. Your best bet, to avoid this, is to use standard techniques that have clearly defined behavior.
If you want an easy-to-type network address, you could use http://10.1.1.1 or a private DNS name (like http://l.l.l.l (a.k.a. http://L.L.L.L)). There are some arguments why a private DNS root is undesirable, but it will likely work with less problems than trying to rely on http://0.0.0.0. Or you could use http://l (for the local web server), and use your hosts file, which you may find is less work than setting up custom DNS, at least for a quick short-term small-scale solution. Those solutions will probably work better than continuing to bang your head against the wall in pursuit of trying to force the world to treat http://0.0.0.0 the way that you want it to.
If you want a solution for having things work with web browsers, follow the advice from heavyd's comment to try http://127.0.0.1 or an IP address that your NIC uses. You could try http://localhost (but other machines, when trying to connect to the web server, should use an IP address other than http://127.0.0.1 or http://localhost or http://[::] )
If you keep insisting on using 0.0.0.0 then you are likely to encounter these issues:
-
It may or may not work, today
-
People will get frustrated with trying to help you since you're violating standards that many other people accept
-
Even if you do get things working today, it may break in the future if you do something like update some piece of software. The update may decide to start treating 0.0.0.0 as a non-normal address.
If you are willing to work around any problems, including adapting source code, then nobody here is trying to stop you. Just don't expect other people to help, because they may see the whole endeavor as a waste of time (compared to the better alternative, which is to just go along with the standards). In the long term, there are other approaches that will require less effort than trying to impose a certain behavior onto the 0.0.0.0 address.
0.0.0.0is not a valid IP address to navigate to. Chrome knows this, so it only allows you to do what it knows is valid for the given input (ie search).http://0.0.0.0, its not a valid IP. Perhaps you mean to usehttp://127.0.0.1orhttp://<YOUR_IP_HERE>?