Rebasing namecoin on bitcoin 0.8

[moa]

Ok, here's the pull request

https://github.com/namecoin/namecoin/pull/30

It is in doublec's repo, branch "add-namecoin-qt".

https://github.com/doublec/namecoin/tre ... amecoin-qt

I strongly suggest that khal now merges this ASAP and devs start using/testing this as much as possible ... and we rebase upon this now combined namecoin-qt/namecoin code so everybody can base their work on it.

Also people need to stop committing to each other's repos and only be making pull requests to the main namecoin/namecoin github (from their own branches) as the commit history is starting look v. dodgy and messy otherwise.

[phelix]

Ok, here's the pull request

https://github.com/namecoin/namecoin/pull/30

It is in doublec's repo, branch "add-namecoin-qt".

https://github.com/doublec/namecoin/tre ... amecoin-qt

I strongly suggest that khal now merges this ASAP and devs start using/testing this as much as possible ... and we rebase upon this now combined namecoin-qt/namecoin code so everybody can base their work on it.

Also people need to stop committing to each other's repos and only be making pull requests to the main namecoin/namecoin github (from their own branches) as the commit history is starting look v. dodgy and messy otherwise.

While of course I think it's great you guys are working on Namecoin as the others I do not understand the purpose of this. Github did not help so please explain.

As of some discussion by pm I had thought we had all agreed on this plan (updated):

Masterplan
x 1.) Release Snailbrain/Thecoder Namecoin-Qt (Q.3.71 currently)
--> People can have a nice GUI and upgraded namecoind, though both may be lacking in some details and are not future proof.
! 2.a) Rebase v0.3.50 to Bitcoin v0.8 (per Khal's strategy)
! 2.b) Discuss necessary changes in fees etc. (make intermediate Qt release only if absolutely necessary (Q.3.xx))
! 2.c) Continue in parallel with all the cool Namecoin based projects to find our first really successful niche.
3.) Integrate Q.3.xx Qt nametab and underlying code into rebased Namecoin*
--> Namecoin-Qt v0.8.5, ready for world domination.

*As this will mean only the integration of some very specific parts IMHO we are pretty free on what we do with Namecoin-Qt (Q.3.xx) now.

IMHO it would be better to do the fork (value size, fees) on the new rebased version only.

edit: namecoind

[biolizard89]

moa said on IRC that the problem is that some devs keep using namecoin-qt as a base rather than the master branch. I'm not 100% convinced that this warrants merging into master now. If moa wants to explain further here he's welcome to.

[moa]

Ok, here's the pull request

https://github.com/namecoin/namecoin/pull/30

It is in doublec's repo, branch "add-namecoin-qt".

https://github.com/doublec/namecoin/tre ... amecoin-qt

I strongly suggest that khal now merges this ASAP and devs start using/testing this as much as possible ... and we rebase upon this now combined namecoin-qt/namecoin code so everybody can base their work on it.

Also people need to stop committing to each other's repos and only be making pull requests to the main namecoin/namecoin github (from their own branches) as the commit history is starting look v. dodgy and messy otherwise.

While of course I think it's great you guys are working on Namecoin as the others I do not understand the purpose of this. Github did not help so please explain.

As of some discussion by pm I had thought we had all agreed on this plan (updated):

Masterplan
x 1.) Release Snailbrain/Thecoder Namecoin-Qt (Q.3.71 currently)
--> People can have a nice GUI and upgraded namecoind, though both may be lacking in some details and are not future proof.
! 2.a) Rebase v0.3.50 to Bitcoin v0.8 (per Khal's strategy)
! 2.b) Discuss necessary changes in fees etc. (make intermediate Qt release only if absolutely necessary (Q.3.xx))
! 2.c) Continue in parallel with all the cool Namecoin based projects to find our first really successful niche.
3.) Integrate Q.3.xx Qt nametab and underlying code into rebased Namecoin*
--> Namecoin-Qt v0.8.5, ready for world domination.

*As this will mean only the integration of some very specific parts IMHO we are pretty free on what we do with Namecoin-Qt (Q.3.xx) now.

IMHO it would be better to do the fork (value size, fees) on the new rebased version only.

edit: namecoind

Several reasons;
That was the Masterplan but it would have been the 'better' way only if steps 1-3 were done in a timely manner ... as that hasn't happened we now have situation where namecoin-qt 'unofficial' might become defacto standard client, bringing along danger of unforseen network forks, etc. Since at this point the code is the protocol it is important that everybody be running with similar code.

The git management of namecoin-qt is not the best and the longer that goes on the harder it will be to clean it up later for merging back into namecoin/namecoin. Best do it now and let others experience with git handle the merge requests, code management issues, i.e. code development on namecoin-qt can be done on main namecoin head, which is a good thing.

The rebase onto bitcoin 0.8 can take place in a parallel branch which it is already doing anyway. The trade off will be that it will probably make that bitcoin rebase messier ... but something was going to suffer either way because the approach taken previously.

Basically, it is just another way to get the same point we all agree on. At this point in time, it is safer to go this route now.

[snailbrain]

Ok, here's the pull request

https://github.com/namecoin/namecoin/pull/30

It is in doublec's repo, branch "add-namecoin-qt".

https://github.com/doublec/namecoin/tre ... amecoin-qt

I strongly suggest that khal now merges this ASAP and devs start using/testing this as much as possible ... and we rebase upon this now combined namecoin-qt/namecoin code so everybody can base their work on it.

Also people need to stop committing to each other's repos and only be making pull requests to the main namecoin/namecoin github (from their own branches) as the commit history is starting look v. dodgy and messy otherwise.

While of course I think it's great you guys are working on Namecoin as the others I do not understand the purpose of this. Github did not help so please explain.

As of some discussion by pm I had thought we had all agreed on this plan (updated):

Masterplan
x 1.) Release Snailbrain/Thecoder Namecoin-Qt (Q.3.71 currently)
--> People can have a nice GUI and upgraded namecoind, though both may be lacking in some details and are not future proof.
! 2.a) Rebase v0.3.50 to Bitcoin v0.8 (per Khal's strategy)
! 2.b) Discuss necessary changes in fees etc. (make intermediate Qt release only if absolutely necessary (Q.3.xx))
! 2.c) Continue in parallel with all the cool Namecoin based projects to find our first really successful niche.
3.) Integrate Q.3.xx Qt nametab and underlying code into rebased Namecoin*
--> Namecoin-Qt v0.8.5, ready for world domination.

*As this will mean only the integration of some very specific parts IMHO we are pretty free on what we do with Namecoin-Qt (Q.3.xx) now.

IMHO it would be better to do the fork (value size, fees) on the new rebased version only.

edit: namecoind

Several reasons;
That was the Masterplan but it would have been the 'better' way only if steps 1-3 were done in a timely manner ... as that hasn't happened we now have situation where namecoin-qt 'unofficial' might become defacto standard client, bringing along danger of unforseen network forks, etc. Since at this point the code is the protocol it is important that everybody be running with similar code.

The git management of namecoin-qt is not the best and the longer that goes on the harder it will be to clean it up later for merging back into namecoin/namecoin. Best do it now and let others experience with git handle the merge requests, code management issues, i.e. code development on namecoin-qt can be done on main namecoin head, which is a good thing.

The rebase onto bitcoin 0.8 can take place in a parallel branch which it is already doing anyway. The trade off will be that it will probably make that bitcoin rebase messier ... but something was going to suffer either way because the approach taken previously.

Basically, it is just another way to get the same point we all agree on. At this point in time, it is safer to go this route now.

it's not merging back into the main branch so this is irrelevant ?

what is wrong with the git management? also, who's going to look after the main branch, assuming we can get anything pushed into it?
I can edit all qt links to say unofficial and also link to namecoinq repo (official one on namecoin.info)..

I thought everything was going fine tbh.. and there is practically nothing left to do on the qt atm until the rebase is done (except maybe gui improvements (extra json field things))..

[phelix]

While of course I think it's great you guys are working on Namecoin as the others I do not understand the purpose of this.

Several reasons;
That was the Masterplan but it would have been the 'better' way only if steps 1-3 were done in a timely manner ... as that hasn't happened we now have situation where namecoin-qt 'unofficial' might become defacto standard client, bringing along danger of unforseen network forks, etc. Since at this point the code is the protocol it is important that everybody be running with similar code.

There certainly is a risk of an accidental fork. Another solution were if everybody would switch to the Q version and then to the rebased version once it is ready.

The git management of namecoin-qt is not the best and the longer that goes on the harder it will be to clean it up later for merging back into namecoin/namecoin. Best do it now and let others experience with git handle the merge requests, code management issues, i.e. code development on namecoin-qt can be done on main namecoin head, which is a good thing.

But with the masterplan above it would not matter as only a small and relatively separate part of code (GUI Nametab and handling) would have to be added to the rebased Namecoin.

The rebase onto bitcoin 0.8 can take place in a parallel branch which it is already doing anyway. The trade off will be that it will probably make that bitcoin rebase messier ... but something was going to suffer either way because the approach taken previously.

Basically, it is just another way to get the same point we all agree on. At this point in time, it is safer to go this route now.

I have a feeling this would open a third box of pandora... why change the plan now?

Also Khal or whoever else will work on the rebase will have to decide on the starting version. Khal is already at 0.4 and I doubt he would throw that away, why would somebody else?

I thought everything was going fine tbh.. and there is practically nothing left to do on the qt atm until the rebase is done? except maybe gui improvements (extra json field things)..

That's what I thought, too.

[biolizard89]

I don't have much to add to this discussion which wasn't already said, but at the moment I think I agree with phelix. End users should be using the Qt version until the rebase is ready; GUI dev should go in Qt and will be added to master when the rebase is ready; non-GUI dev should be deferred until the rebase is ready.

[moa]

I'm trying to do you guys in the namecoin-qt branch a favour here before you get too far off the rails ... but by your responses it sounds like you probably can't even see that. E.g. explain to yourselves why commit 7279639c5ab291490ca6856ba204ab67dfcc019b is not in your namecoin-qt branch? https://github.com/namecoin-qt/namecoin ... ter?page=2 and it is in the main namecoin-qt branch https://github.com/namecoin/namecoin/co ... amecoin-qt you are not even rebasing on simple merge commits tidily .... I'm not sure if you even know what you don't know.

In the end, just do whatever you like, it's incredibly frustrating, exasperating and frankly I can't be bothered with the hassle much longer.

[biolizard89]

@moa If you had explained what was going on prior to making the pull request I think there would have been a lot less potential for drama. As it happened, snailbrain/phelix/I were completely confused as to what the issue was, and I still don't think it's been particularly carefully explained what changed between now and last month when we all decided to keep namecoin-qt separate. I don't want to cause drama; in fact I'm fine with this thread stalling for a day or two so that everyone can have time to ponder.

We're all on the same side here, let's all take it easy.

[snailbrain]

I'm trying to do you guys in the namecoin-qt branch a favour here before you get too far off the rails ... but by your responses it sounds like you probably can't even see that. E.g. explain to yourselves why commit 7279639c5ab291490ca6856ba204ab67dfcc019b is not in your namecoin-qt branch? https://github.com/namecoin-qt/namecoin ... ter?page=2 and it is in the main namecoin-qt branch https://github.com/namecoin/namecoin/co ... amecoin-qt you are not even rebasing on simple merge commits tidily .... I'm not sure if you even know what you don't know.

In the end, just do whatever you like, it's incredibly frustrating, exasperating and frankly I can't be bothered with the hassle much longer.

edited this post a 1000s times..

no point arguing, maybe you are right.. anyway, we are done with the qt we have already spent more time on it than we anticipated, i don't want to be responsible in anyway for failure in this project. We hope it gave the project a good kick start, enjoy.