just call it 2 phase

Even if the dc was in pulses, the radio would work. When plugged in the right way, the current would still flow through the rectifier tube and be filtered by the same filter used for ac operation.
Bill

The arcing commutator would generate so much hash that all you would get would be a loud buzz. Any time the brush loses contact with the armature, it arcs.

I had a portable radio with a hand crank generator. The generator is nothing more than an ordinary Mabuchi DC motor as you'd find in motorized toys. Spin the motor and you get something resembling DC out which is sufficient to charge a small NiCd cell as well as power the radio even if the cell is completely dead. A generator might not provide absolutely clean pure DC but I'm pretty sure they could have made it work. Cars used generators up until at least the 1960s, they made radios work in those too.
This would all be easy enough to test. Take a mains voltage permanent magnet motor out of something like a string trimmer (weed eater) and spin it with another motor to get around 110V out of it, then connect that to a transformerless radio, lots of those around still. Or heck, just scope the output of a DC toy motor being spun by another motor.

Thjis is why in my previous post I mentioned that a modern generator would have various bells and whistles that may not haver been present in early Edison generators. A generator is likely to have a rotating brush assembly so that the brushes could be placed where the arcing is minimized. Interpoles will also help commutation. The brush material itself is somewhat resistive to damp out noise a bit.
Remember that in the days of six volt car batteries charged by a dc generator, radios worked just fine.
Bill

Radios in cars has absolutely nothing to do with this discussion.
The battery in cars is a very low impedance and would hide any "Pulsing" in the output of the generator.
John G.

They have plenty to do with this discussion. The battery is low impedance and does tend to "hide" pulsing just like the B+ filter cap in the radio. As I mentioned, the hand crank radio would operate fine, even when the battery was old and completely worn out to the point that it would no longer take a charge. Anyone is welcome to try this, I see lots of speculation but no hands on experiments here aside from my own from a long time ago. Speaking of old car radios, most of them used a mechanical vibrator to generate the B+ for the plates, talk about hash, the crudest brushed generator would have a cleaner output than one of those still, with shielding and noise filtering, they worked. Additionally, if homes were fed by pulsed DC mains, they could be used to maintain the charge on a bank of batteries similar to the old 32V rural power systems in which batteries were charged by windmills and/or gasoline generators.

Generators are nothing.
In the earlier days of car radios, they used miniature vacuum tubes. Tubes didn't work so well with 6V on the plate, so the radios had things called "vibrators" (no not THAT kind), something that rapidly interrupted or reversed current flow to a step-up transformer, the output of which was rectified to a B+ voltage that tubes could use. Standard filtering of the leads allowed them to work.
I once had a police 2-way radio that used a bunch of 12AT7 class tubes.

If the brushes in Edison's DC generators didn't contact two segments at a time at the current they produced, the arching would destroy the commutator in a few minutes. Just like the similar design of a universal motor, it is the biggest problem with commutated brush type motors or generators.
To intentionally pulse the DC, the gaps between segments would have to be wider than the brush so they would operate in a break before make mode.
I worked on hundreds of them, in the '60s. Have you ever studied the vibrator supply design? There are buffer capacitors across the primary winding of the transformer to reduce the loss of contact material. The vibrators ran at 115 CPS (Hz) to reduce the hum. It was a tradeoff between the cost to filter the DC, the life of the vibrator, and the RF interference caused by the vibrator. The vibrators were in thick steel housings, with a thick molded foam insert to reduce the mechanical noise. The socket had a spring loaded clamp that held the vibrator case tightly to ground it to the chassis to reduce EMI/RFI noise. Even after all that, the life of a vibrator could be as little as a month, if the radio was in a taxi or someone who drove all day with the radio on. The foam deteriorated from the heat and contaminated the contacts, so the shelf life wasn't good, either.
The first solid state replacement vibrators came out in the '60s, built with Germanium power transistors in T)-3 cases, inside a standard vibrator can. If the buffer capacitor was bad, or failed, they blew immediately.

The DC had some ripple, but it wasn't pulsating. I have looked at the output of a generator on more that one pre '63 car with a scope. One of the first signs of a failing generator was noise in the radio.
BTDT, 40 years ago.

If homes were still fed with DC, the generators would have to be a couple miles from your home. Also, it doesn't have to pulse to charge a battery.

That wasn't my point. I was just saying that *if* for some reason AC had not succeeded and we still had DC generators powering our homes, radios could still be made to work directly from the generator, even generators build with mid 1800s technology.
I didn't mean to imply that batteries need a pulsed voltage to charge, but simply that hypothetically if design constraints of the generator caused the output to be pulsed, it could still be used to charge batteries and power devices.
In both cases I was simply countering arguments from someone else that this wouldn't work.
Funny thing is with modern technology, DC power distribution would actually have some advantages, though in a residential situation the disadvantages of dealing with high voltage service and large DC-DC or AC converters would far outweigh the advantages, but it could be done. They use it in some long distance transmission lines afterall.

It is actually becoming the preference.
We could make a system for residential DC service with the batteries out at streetside. That is a chemical and environmental mess though.
Until battery technology takes a few steps forward or until solar cell technology gets so good that we care less about battery bulk, we will be in an AC fed, AC consuming world. Make for easy conversion, but a magnetically noisy environment.
DC makes for instant 'kill-n-cook' situations though. You really do fry.

Sure, but do you want a 500 KV DC line directly to your home? ;-)

Edison's machines were true DC machines. Some earlier machines were "Gramme Ring" machines which could have given you this impression (falsely). As far as I can see from pictures, his machines were drum armature 2 pole machines with a conventional commutator (which is a synchronous switch). Note that a commutator, properly used, switches the current only in the individual coils under the brushes, shorting the coils (2 in his case) on the neutral axis-at a time when the individual coil voltage changes polarity. During the time that the brush contact moves from one side of the coil to the other, shorting the coil in the interim, the voltage in that coil would be nearly 0. The total voltage would have little ripple as the rest of the coils would be producing normal voltage so the total voltage certainly did not go to 0 or near 0. Brush width and material was a problem until it was discovered (by Brush or Thompson- I can't recall which) that carbon was ideal. Modern machines are designed to compensate for armature reaction which shifts the neutral axis and causes arcing. In Edison's day this compensation was done manually. In operation, except for considerable refinements in design, Edison's machines were essentially the same as modern DC machines.

By the time the "mains powered" radios came along, the supply was AC. In addition, while an arcing commutator produces hash, it produces more important problems in large DC machines and the design of the commutation system as well as proper maintainance of the commutator surface eliminates the arcing. Small motors such as used in drills do arc because a) nobody maintains the commutator until the arcing becomes a problem, and b) compensation for armature reaction is not provided.

Sure DC transmission has its advantages- at high voltages for long distances or for underwater or underground cables or for asynchronous connections between systems. All point to point rather than grid systems where specific technical requirements are needed or the distance is long enough that the savings in transmission costs exceed the costs of the end equipment. For mid level and distribution, it has serious disadvantages(no current zeros).- AC offers great advantages in switching and flexibility of networks and the optimization of voltage levels for the task at hand with relatively cheap and efficient transformers.

Some areas still had DC power well into the '60s and even early '70s. A friend of mine in the Army told me what his dad had to go through to get AC for the new elevator in his building near Chicago, around 1970. The elevator company refused to repair the old DC model, and only installed new AC powered elevators. People in those areas used the transformerless radios on the DC power lines, and wouldn't notice the switch to AC, as long as the usual two electrolytics following the rectifier were still good.
Add to that, the crappy speed controls in variable speed power tools, and you get even more noise, because they are too cheap to add a small inductor to filter the noise before it enters the power line and is radiated to nearby electronics.

I read somewhere recently that the last DC service to a building in New York I believe it was, was shut off sometime in the 1990s. I had no idea it lasted so long.
Somewhere there is a website with pictures of some of the gigantic rotary converters used to convert 25Hz AC to DC to drive subway trains. Some of these were still operational at least up to a few years ago. I've seen relatively recent pictures of polyphase mercury arc rectifiers still in operation doing the same job.

Your date is a little off. The last plug was pulled November 14, 2007. ;-)
(Geez, Google is worse than even I thought!)