[tahoe-dev] GSoC Share Rebalancing and Repair Proposal

[Kevan Carstensen]

On 4/23/13 1:18 PM, Zooko O'Whielacronx wrote:

> I'm not sure, but I have the _feeling_ that the #1382 that Kevan
> Carstensen started may be sort of a critical basis for successful work
> on related tickets. See his github branch for the code he's written:
> https://github.com/isnotajoke/tahoe-lafs/commits/ticket1382
>
> Kevan: do you think Mark could profitably work on other repair and
> rebalancing tickets while leaving your #1382 branch alone? Or do the
> two of you, Kevan and Mark, think it might be a good idea to have Mark
> take over Kevan's branch and finish it?

The #1382 branch is pretty old. I haven't tried to port the changes to 
the current codebase & don't know to which extent they've bitrotten. As 
a purely practical matter, I wouldn't advise putting too much time into 
making that branch suitable for further development; it doesn't 
represent much work on my part, and could prove to be a distraction 
and/or more time consuming than simply starting from scratch with the 
current Tahoe-LAFS code.

I think it makes sense to fix #1382 before tackling the stuff Mark 
proposes. Done right, #1382 should ease implementation of those changes 
& help make the result more maintainable going forward. OTOH, #1382 
could easily be an entire GSoC project in itself, and I'm skeptical that 
it could fit in with the other stuff in the proposal (though I'd be 
happy to be proven wrong on that). I think the stuff that Mark proposes 
could be done without fixing #1382 (though with the potential risk of 
having a lot of his work undone if/when #1382 is fixed), and would be a 
valuable contribution to the project.

- Kevan