We should really do something about the synch time though. Would it make synch faster if there were more hubs and relays? Do hubs and relays have to have public/static IP addresses?
I tried syncing a full node. It seems that most of the delay is caused by the wallet reading and writing to the local hard drive excessively. I don't know if it is possible to configure sqlite to work with a bigger memory cache or something to cut down on disk access but if so, that should help a lot.
.....
I agree completely -my experiences:
Core i5-321M @ 2.5 GHz
RAM 8 GB
hard disk Seagate MomentusWIn 10 64 Bit
full Sync Time: >5 days (>120 h)Core i7-5500U @ 2.40 GHz
RAM 8 GB
SSDWIn 10 64 Bit
full Sync Time: 1 day (24 h)SSD => factor 5 fasterOn mobile devices without ssd, the synchronization never ends (e.g. SD memory card).
Clear recommendation:
The read / write speed of the device is the bottleneck, devices without SSD (or fast HD) should prefer a light wallet.
This recommendation must be pointed out during the installation in order to protect the users from frustration.
Excellent observation! Thanks! Also, there is probably something that can be done programmatically to improve the situation. Perhapse Byteball should use more RAM for caching and flush to disk less frequently?
I'm not a dev, but database applications on mobile devices may not be a new problem.
There should be experts who can better estimate the possibilities of optimization.
Either we find someone in the community who can support Tony (if he needs support).
Otherwise, we have to hire someone.
At the moment, the community is still poor. After the end of the distribution, I would vote for a big donation call.
(There should then be sufficient stakeholders who are interested in a strong further development.)
Tony and CryptKeeper then (or bevore) need to set the priorities for further development and use of the resources.
The optimization of the light wallet will certainly have priority (in my opinion) because the primary goal is global adoption.
For this scale "
One last thing. The remaining 1% will be given away to the first 100m users who install Byteball wallet, 100 Kbytes to each user...."
the wallet have to run fluently on low budget mobile devices.
Because the target group of this scale:
https://en.wikipedia.org/wiki/List_of_countries_by_percentage_of_population_living_in_povertyhttps://en.wikipedia.org/wiki/Population_growthA key factor for success (large-scale adoption) in the real world is to protect potential users from frustration.
The first contact must be positive -> no hurdles, no techspeak just value for the user..
I think for the light wallets, the network must be correspondingly expanded.
Again, we need to see what the community or donor can do.
A thought about this => give away bundle like this: Raspberry-Model-A1-2GHz-64-bit-quad-core + Ableconn-PIUSB2MSA-Converter-Stackable-Expansion (SSD Support if needed)
or for short :
BmsB (
Byteball
massive
scaling
Box)
price (without discount): ~60 Dollar * 1000 bundles = 60,000 Dollar.
I would say that is not a problem if capitalization reached the level of IOTA.
If it is necessary we push again 1000 afterwards..
If each owner internalizes this fact "
..remember that its value is proportional to the square of the number of active users", things will develop very positively