Topic: [ANNOUNCE] New alternate cryptocurrency - Geist Geld - page 15. (Read 74190 times)

Hm, the RPC error is quite bizzare - I successfully connect GUIminer "out of the zip" (though its option saving is apparently weird and I have to re-input port every time it restarts). Sent ya some geists

Isepick - I guess your router be playing tricks on ya, especially if strange connection behavior is not limited to Geist's port and also happens for Bitcoin.

probably need to add 'server=1" to the config file ... (and server config settings also if not in there)

As u can see in the picture, the copy option is greyed out. So their is no way to copy & paste the address. I manually wrote the address now.
Besides, after i give exit, GG closes, but windows gives error message that the software closed unexpectedly & want to check solution.
I am using the new version released on 13 sep 2011 GeistGeld_New.zip.

I extracted to a folder & run GeistGeld_portable_cons.exe & It starts downloading blocks & it works fine that way.
getinfo now gave the following result.
version:32400
balance:0.00000
blocks:20523
connection:11
difficulty:104.3888
....


I tried to mine myself using the built in guiminer. But it only opens the roaming\poclbm\poclbm.ini file & uses that config to mine.
& if i try to mine with local host, i get


How to mine solo?
And also if some one can send me one or two GG, so that i too see my balance increase.
Below is the address i manually typed from the picture as copy paste not working.
Please double check before sending GG.

gJ9G2BCrxeSguNsck9Toy9tKtiyFT5Wwhv

Windows 7 Ultimate, 64 bit.

Disabled upnp. Logs showing dd-wrt being flaky. Instead of forwarding to port 7769, the router forwards to some random port instead. Which is really strange since I have uTorrent running fine on another box with its port forwarded the way it should be. Perhaps if I am feeling brave sometime I will update my version of dd-wrt to see if that fixes it.

Looking at my logs further shows it doing the same for all incoming connections on 8333 as well (which I also have forwarded). Its a wonder I can connect to any network and mine at all

Thanks for the input, Theymos.

Well, that's somewhat unnerving. Let's see if such incidents become more common and if anything can be learned from it.


Hm, that's odd.

What's the platform and did you try disabling unpnp in the config ?

Can you check your firewall logs or just get what's running (I think it provides connection info...)

Port 7769, right? Still can't seem to get more than 8 connections.  

Yeah, I'd guess isolation. Unless the network really did produce no other blocks in that timespan.

Well, in a connection sac of some kind, the interval of the blocks you're building with your pals (or with yourself, if you got really isolated totally "alone" and still continued mining somehow) should probably drop, not rise, since you start at higher (native net) difficulty and then adjust lower (which Geld does in less than a minute).

Anyways, very interesting blocks. Maybe we should invite theymos ? I heard he likes blockchains.

P.S.:
Oh, and miners on Geist definitely should open incommings and increase maxconnections=

Having more than 125 can hardly hurt if increased within reason.

It may be a minute between blocks, but because I was building on my own split chain, each block was built on top of the previous one I found. So it looked like this:


And then after a few minutes, they all got orphaned.

I think the biggest problem will be people with 8 connections. Meaning, they don't have ports open, so are really quite poorly connected to the network. These people will get orphaned more often than everyone else. Make your blocks count. Open incoming connections.

Well, that's exactly why I brought up the issue of "how many nodes per a Big Pool" (I PMed Tycho of the deepbit fame, btw)

seems to me that Nodes That Matter in modern bitcoin are, quite frankly, a handful, and the rest could just send them tx "the lean way" and keep the blockchain just for the purpose of feeling themselves like Ye Olde Coin Archiviste.

As for those blocks, there was always in excess two minutes between them (intervals irregular, in one case in excess of 7 minutes), so it does not appear to be exactly a block streak.

That does not disprove the hypothesis that coblee and his peers got sequestered into a "low-diff" fork, but there are other explanations too, like some less than kind person mining on-and off, rushing in at minimum diff, mining till it adjusts, stopping to let it drop (usually missing the norm and going a mite below), then mining again, rinse-repeat.

P.S.:
Someone should write a goddaymn visual block explorer that would allow to visualize orphans and such in relation to main chain  In 3-D

Well it's more about the number of nodes than the network hashrate. With a lot of nodes, block information will have to be sent to all the nodes. So there will be a lot of data sent around. Nodes that are unaware of the latest block, will be splitting the chain. So there will be a lot of collision and chains merging. Will be interesting to see if the network can handle it.

As I suspected there is a threshold in which where you add more GH you don't produce any more blocks than with less GH. I'm curious what that number is.

Won't matter if we have 100 TH. After a few quick blocks, difficulty will adjust, and we'll be right back to 15 second blocks.

I did check. I have 8 connections. Since the blocks were consecutive, it makes me thing I was isolated somehow. I will PM you with timestamps and txids.

If you (or anyone else) ever find another "bizzaro streak of blocks", please do a getinfo and see if you have connections, and if so, how many. Because if you have like 6-8 connections and get such a streak, then likely something more that just "connection loss" is afoot.

Having said that, if network issues start causing orphaning of large ammount of blocks, that could be detected when studying the blockchain. According to ArtForz, so far there is no indication that the amount of orphans "on whole" is higher that normal.

P.S.:
coblee, could you PM me the timestamps and txid  on those orphans (should be in listtransactions) ?

Interesting. I wonder how things will play out as we start hitting 100, 200, 300 GH.

I've been mining with 6 ghash and just now I saw that I found 5 blocks in a row. Immediately, I realized something is not right. I can't be that lucky. After a few minutes, I saw that these 5 blocks all got orphaned. I assumed that due to network issues, I was disconnected enough to build up my own block chain, which got wiped out when I reconnected.

Unfortunately, I think these problems will get worse as the network grows.

Don't really know unless you ask a pool operator willing to disclose that information (difficult) but an example which I think is maybe overdone (except if we're talking VPS machines):


10 servers - 10 coin nodes for a big one. Bigger pools probably also use load balancers and/or bigger servers.

On another note I applaud your efforts, and it must be said that they have already brought results.

Cheers.

Well, geist obviously is more network-sensitive and there obviously is a blockrate/decentralization tradeoff at work (I think I said that in post one), and point of Geist experiment is, in part, to see exactly how this tradeoff works out on a "live" network that experiences all the live influences.

Geist, however, is quite ready to trade decentralization for blockrates (up to a point of course ), which I also stated in my first post, and what concerns me most is that network effects that impede Geist decentralization will also impede pool operation, thus also impeding the partial centralization and thus making further increase in Geist hashrate problematic at best.


Well, "thin" clients don't download the blockchain proper anyways, so theoretically nothing keeps people from sending their tx to all the well-known "big miner" geist nodes first, then the "plebes", thus maximizing the likelihood of those ending up in a block quite soon, or some other, smarter optimization of the way the net handles transactions (I'm not quite smart enough to come up with those by myself, admittedly ).

Bitcoin will need many of those optimizations too if it grows "huge enough", and this is why it might be worthwhile for the community to play around with Geist, IMHO.

In the end, yes, it is quite possible that "audaciously fast blocks" will cause either insurmountable scaling issues or problems of sustaining a good enough connection to force your tx in even if some optimizations in transaction transmission by "lean" clients are in place.

The point of Geist is 1) try to find that out in a "live" setting 2) try to seek amendments to overcome those issues 3) see if some of those optimizations are also worth implementing in Bitcoin

If worst comes to worst, there's nothing in the nature of Geist that would prevent gearing down the blockrate while adjusting the subsidy and retarget to leave economics intact (or maybe even give folks some temporal single-issue of pseudo deflation if they shall so demand, though I refuse on general principle to set a hard upper limit on number of Geists in existence. "Coin regeneration" FTW ), however such course of events will be a huge blow to Geist's stance and appeal, even despite all the knowledge gained when assessing the causes which could lead to such an unfortunate and radical solution.


Interesting things will happen when pools will start to get deployed.

P.S.:



I am a bit iffy on the deeper intricacies of the "how does pool form" theory, so,  how many pool nodes connected to the bicoin network (as opposed to miners connected to pool, which as far as I understand have no particular reason to even be aware of the blockchain since they only deal with the pool's share anyways) does on average each of the "big fishes" of pool business represent ?