ya thats fine, if you can easily spot is as spam or and require more fees to process spam by all means do it. shouldnt be hard to get miners to do that.
but nothing is stopping the spammer to make his spam taste better, and with the current blocklimit with blocks often full its not very expensive for the spammer so spam with good tasting spam.
There isn't? I thought making it considerably more expensive to push spam transactions might act as a disincentive towards doing so. I don't understand this idea that "if blocks are often full (which is not true on average), it is not expensive to spam." Blocks being full doesn't make it
cheaper to push transactions; quite the opposite.
cuz the spammer does not need to fill the block completely they are already half full on avg.
if block where on avg 99% full it would be cheap to spam it up so blocks are always 100% full
fee would go up and make the spam attack more expensive but not by much because spammer only really needs to "fill in the gaps" so to speak.
now imagine that blocks are 100% full all the time, how much does it cost to spam the network and create a 365day backlog? its already headed there naturally without spam, so not much!
"All the time" would be a very inaccurate description.
If the spammer does not need to fill 100% of blocks himself, that doesn't make it any cheaper for him on a per-size or per-output basis. The point of addressing spam is itself an end goal -- we don't have to consider it in the context of "OMG, a full block!" The disincentive is there, no hard fork is required, and this may considerably mitigate capacity issues (it may not) while reducing unnecessary bloat.
Sure, transaction volume is increasing on the long term. But I'm not convinced on the proportion of transactions that are "legitimate" (i.e. without spam). The capacity question is a perpetual one -- how do maintain efficient incentives in a constantly evolving economy? This is part of that question.
