All of the postings here seem to be correct, however for different reasons.
All USB ports will supply 100mA with or without negotiation.
Since 2007, the USB spec requires that 2 types of "charger" can connect to a USB client. Either a "dumb" dedicated charger or a “smart device” which can chat in order to negotiate the amount of current that it would like. The “smart device” can itself be of two different types, a “Standard device” or a “charge downstream” device. These only really differ for the purposes of this discussion in the maximum amount of current that they can supply [following negotiation], either 500mA or 900mA.
So how does a USB device recognise whether it’s being plugged into a dumb charger or a “smart” host? The charger should have a short between the D+/D- pins which, upon initial connection causes a logic 1 to be seen by the client on D- identifying it as a charger. Whereas a smart host has resistive pull downs, so the D- is seen as a logic low identifying it as a “smart device”.
In the case of a smart device [assuming that the host PC has the correct drivers installed] it will ask for the current it needs, or be quite content with the standard 100mA supplied without negotiation (though strictly speaking this could drop to 2.4mA but I’ve NEVER seen this implemented in practise). Hosts will only allow the device as much current as they can supply. Following this negotiation, an agreement is reached and the device decides how best to use its allocated current: Can it charge itself? Can it power itself? Can it both power and charge itself? Can it do nothing except “talk”? These decisions will also vary depending if the device is “on”, or “in standby”, or “off”.
If the negotiation fails for any reason, such as missing drivers, most real world devices assume that they can have 100mA and proceed with that assumption, though this may slow down charging or stop it altogether depending on the PC USB implementation.
However, this power control is expensive and many real-world USB ports simply supply several hundred milliamps [or more] without any negotiation. These ports will usually be fuse-protected, to prevent a faulty device plugged in taking down a whole motherboard! Un-negotiated large-current capability is nice, as USB fans/lights etc and other essentials [ahem] can function quite easily on MOST computers without any intelligence built in. It really depends on the host manufacturer to decide what to do for the best compromise.
Now regarding chargers. The USB spec requires that a charger can supply at least 1500mA But this is far more current than many devices can use in reality, requiring the chargers to be over-engineered. This over-engineering is very expensive, particularly for OEM suppliers.
So what to do? Simple - do not put the short circuit in between D+ and D-. When the device is plugged in, it does not see the short circuit and so assumes either a “smart” device” or its own proprietary charger! It tries to negotiate for more power, the dumb charger ignores it, and so the device assumes that the proprietary charger is attached and it “knows” that this charger can supply the current that it needs, whatever that is, 100mA, 500mA 750mA or whatever.
If one tries to plug another device into the same charger it may not work as it will not recognise it as either a “real” charger or as a “smart host”, or it may simply assume 100mA capability, but this is probably not enough to power and charge most modern devices. Ironically, plugging the device itself into a proper USB compliant charger should still work.
People have hacked chargers or USB cables to create a short between D+ and D- for these charger types, such that proper clients recognise them as chargers. The danger is that these clients will then assume that the charger is fully USB-spec compliant and can thus supply 1500mA. Trying to draw this much current could overheat the charger or cause other damages. It is also potentially dangerous. There should of course be some over-current protection, but unfortunately this also costs money and cheapo-chargers won’t have very good, if any, overload protection either!
When connecting a device to a portable smart “battery powered” charger, the device can also be fooled in the same way, with the short-circuit-trick, but you are again relying on the chargers regulation and over-current capability. Typically these chargers tend to be very sophisticated and so generally this is not an issue. More generally, some of the better ones (e.g. power traveller products) have automatic dual mode-charging (constant voltage or trickle current) and these can be forced into constant-voltage mode, which is preferred over the “short circuit” approach for bigger items like the PS-Vita.
So, what if a device requires more current than the specified 1500mA? e.g Apple devices or some smartphones. For example the iPad requires 2.1A to both charge and operate simultaneously. This much current would break [or cause them to cut out] the cheapo-chargers and also even the fully USB compliant ones.
Apple use various resistor divider networks on the D+ and D- pins of the charger to create very specific voltage levels, this tells the iPad that the charger is capable of safely supplying the 2.1A required. If the iPad doesn’t see these levels it does not try to draw 2.1A as it believes the charger to be incapable. Various blogs state that Apple are ripping everyone off by forcing them to use Apple chargers. This is not quite true. It is true that this is not in the USB specifications but neither are Apple’s requirements.
In summary: due to conflicting reasons of cost and modern high powered USB devices, manufacturers are finding clever ways to overcome the limitations of the USB power supply and charging specification. Generally most are USB compliant. Swapping over chargers may or may not work and hacking cheap-ones to “appear” to work is not always safe. Fully USB compliant 1500mA chargers [with the D+/D- short-circuit] should be able to charge any USB device requiring less than 1500mA, [Apple anomalies notwithstanding] though these will be more expensive, but safer and will work across more devices. Conversely, if you have a cheap-as-chips charger with a low current capability like “100mA” stamped on it or a "spare" USB port on your PC, it may still work, but try to charge the device with the device turned “off”.
If you are interested in how much current your device, PS-VITA or whatever has "negotiated" then in Windows: just look in Control Panel > Device Manager > Universal Serial Bus Controllers > Generic USB Hub > Power. The Vita is around 2400mAH capacity so if the value here is 100mA it will take 24 hours to charge. With the drivers (or whatever Sony call their software, "Content Manager"?) installed it should recognise the device correctly and this current should be a higher value which will charge faster.