Also - I can guarantee you that even if Bitcoin Core decides to do a 2MB fork due to the current "political pressure" Gavin will immediately then start up with a 4MB (or more) hard-fork proposal.

Perhaps you'd like to take a bet on that?

Seriously - do you really think that doubling the size of the blocks is going to allow everyone to buy coffees for Bitcoin?

Do you think if we made it 20MB that would do it?

Or maybe 64MB?

Or?

Sorry - it won't scale with that approach at all - so caving in to the pressure of mostly stupidity to increase the block size to 2MB might have to be the strategy to keep Gavin at bay but it actually doesn't do anything to solve the fundamental scaling problems at all.

Really - care to quote that?

And even if he did mention something about increasing the size after having reduced it to 1MB why is that relevant now? Satoshi left the project (also presumably if you are going to invoke his name then you should be calling for 64MB blocks as that was the limit originally created by him).

As I stated - if Gavin will announce his retirement then I'll accept 2MB blocks (I don't actually care about the block size limit being 1 or 2 MB but I don't want Gavin allowed to be in charge as virtually no-one wants to work with him).

Because you said so?

And who exactly are you to be in charge of when "time's up" eh?

(another paid shill no doubt)

Then why on earth does Gavin jump from one project to another in order to take back control of the development with a hard-fork?

Who is the "lead dev" other than Gavin that is behind these projects?

(oh - and we can see what happens when Gavin doesn't get what he wants - Hearn quits and next will be Garzik)

Typical post from you (a paid shill and an idiot) - I didn't even mention anything to do with the CEO of Blockstream.

May I suggest you take some lessons in "reading comprehension".

Gavin is the guy who has made this all political in his quest for power.

Why are you so impatient?

The supposed "full blocks" are not full (there are empty blocks being mined every single day) and most of the tx "backlog" are just spam txs.

It is obvious that the recent flood of txs are "attacks" rather than reflecting actual normal user demand.

Personally I find it very amusing that years ago when Gavin appeared on CNN (or CNBC - I forget which) with Amir and Amir was talking about Bitcoin being used for "micro-transactions" Gavin was very quick to interrupt Amir and say "no - Bitcoin is not suitable for that" (in fact I think that he basically spoke over Amir in order to "shut him up").

So when he was "the lead dev" it wasn't suitable for buying coffees but now that he lost his throne (which he abdicated btw) and wants it back suddenly we need to be buying coffees with Bitcoin (and if we can't buy our coffees for Bitcoin by the end of next month or so then the whole project will die).

Stop being sucked in by a narcissist guys (he still calls himself the "Chief Scientist" of Bitcoin for fuck's sake and he barely even contributes any code at all).

If Gavin would agree to retire and have nothing more to do with Bitcoin then maybe one could take the idea of a 2MB hard-fork as something other than a power grab (but until he does that then I am convinced he just wants to regain the control that he thinks he should have).

If Gavin thinks he is so talented then why doesn't he just have the balls to create an "alt" rather than trying to fuck up Bitcoin.

In all likelihood the content of the file was replaced (so is gone forever unless you had kept a backup - and if you didn't keep several backups of a wallet with over 200 BTC then that is actually rather astonishing).

You keep on thinking that "the file was moved somewhere" but it is fairly clear that this is simply not the case. You effectively deleted it yourself by renaming it.

That being said it is possible that the contents of your HDD might still contain the information but you'd likely need to enlist the help of a data recovery service for that.

You should *never* be renaming "wallet.dat" full stop and most especially not while the program is running.

And again - you have been told you should not be renaming important files (you should copy over them instead with care) and most especially you don't do this while Bitcoin Core is *running*.

(if you insist upon doing things that the software wasn't designed to support then you can expect trouble)

Doh - I had meant to type a comma but somehow ended up adding a full-stop (the problem with not actually testing something you post). Thanks for correcting that silly mistake.

I think the key point was renaming stuff "while Bitcoin Core was running" (and that is where the bullet entered the foot).

Clearly the final state of the DB became problematic due to what had happened - but as an end-user you should never rename things like "wallet.dat" only make copies (that is just plain commonsense).

My guess is that some other state file (at the DB level) could be the reason for this problem but it isn't a Bitcoin Core problem really.

I have batch files (for Windows versions of Bitcoin) to work with different wallets - I never "rename" files though (they always use copies and overwrite "wallet.dat" before Bitcoin Core is started or copy "wallet.dat" after Bitcoin Core has shutdown).

The Bitcoin Core code clearly does this - but if BerkleyDB decides to open the file (after your rename) and then not report an error then Bitcoin Core is left unable to do anything (it is relying upon the 3rd party software).

The mistake was renaming rather than copying (and that was a "user error").

It did (i.e. it created the "wallet.XXX.bak" file for just this reason) but then you went and renamed it (it can only do so much to predict what silliness a user might get up to).

Typically when something breaks you should stop and ask for help then (not try and fix it yourself if you've got no idea what you are actually doing).

Okay but I think you are going to cop it when they read this step:


(quite clearly it was *you* that got rid of the "wallet.XXX.bak" file by renaming it - not Bitcoin Core)

Perhaps you should clarify that the issue you are raising is why did the supposedly corrupt "wallet.dat" suddenly seem to not be identified as corrupt (but was replaced with a new one) as this is more to the crux of the issue.

Bitcoin Core itself is not doing what you suggest but it is possible that BerkleyDB is doing something strange in this situation (the __db files are below the level of what is coded in Bitcoin Core so if there is some issue with that then it is an issue with BerkleyDB).

Remember it was *you* (or *him*) that got rid of the file "wallet.XXX.bak" (by renaming it) not Bitcoin Core. What should have been done was to create a copy of this file (rather than rename it) in the very first place.

Also why on earth someone would be silly enough to rename such an important file as "wallet.dat" when Bitcoin Core is running is beyond me (it's as though the OP was trying to destroy his wallet on purpose which is hardly the approach that anyone who valued 200+ BTC would take).

Okay - so basically Bitcoin didn't actually start up properly then (or it started up without a wallet).

In either case it isn't magically deleting anything if you renamed the .bak file yourself.

Are you sure your .bak file is a corrupt wallet?

If there are no traces of the "wallet.XXX.bak" file after being renamed to "wallet.dat" (assuming you deleted the empty "wallet.dat" first) then it is because Bitcoin Core didn't find a problem with it when it restarted (i.e. the file wasn't corrupt).

The code is not ambiguous nor is it going to behave differently on Windows or Linux in regards to this.

If you want to pay me to "spend my time doing things" then you'll need to pay me up front.

You have been told to restore your old wallet and yet you don't do that and then seem to insist that others should waste their time repeating things in order to try and "point out a bug" (if I had just thought I'd lost over 200 BTC I wouldn't give a fuck about whether or not there is a bug).