Roll alpha channels into beta channels

[dankessler]

Per discussion in #378, we've decided to do away with the alpha/beta divide and just keep beta channels.

I've updated the wiki, but before we discard all of the alpha channels, they need to roll into beta.

@heffjos has pointed out that there are several open pull requests that involve alpha channels. To this end, I propose the following approach.

1. Immediately: no more pull requests involving alpha should be initiated. All feature updates go directly into beta now
2. Close pull requests that merge alpha into beta (this will get handled systematically later)
3. Resolve or close and reopen with beta as target any open pull requests currently targeting alpha
4. Merge alpha -> beta per toolbox. I will handle this, and probably not through the pull request machinery to avoid spamming the list. In cases where the update somehow ends up being non trivial, I will kick it to a pull request.
5. Deal with any tricky merges resulting from (4) above
6. Delete all alpha branches. I will handle this

[rcwelsh]

Bottom line: you'll ( @dankessler ) let us know when we can start doing pull requests again, ala http://xkcd.com/1597/

:-)

[dankessler]

Here is the list of open pull requests that involve alpha
(crossing them off as I finish)

• Spm8 batch/update to spm8 r6313 #381
• Conn tool alpha rcw #375 (maybe? could just be robert's local name involves alpha)
• Typo #368 (this is kind of a hotfix, with ConnTool_alpha -> public)
• made a few comments/fixes #367 (this is maybe subsumed into the fix of First level alpha #337 now?)
• First level alpha #337 (this is alpha -> beta, but has lots of great discussion. I'd rather let it resolve first)
• Graphtheory updates 11/21/2014 #366
• minor bugfix for one mc_Error call that was passing an sprintf argument ... #365 (this is maybe a hotfix? not sure if it touches alpha but whatever it fixes needs to get backported)
• Quality checks alpha #361 (this is alpha -> beta, but with some discussion. I'd rather resolve it first)

So in summary, basically all of the pull requests

[dankessler]

Nice reference @rcwelsh

[dankessler]

Here's an update on our progress

(I was able to tick off anything involving closing pull requests because thanks mostly to @mangstad there are no longer any open PRs)

• Immediately: no more pull requests involving alpha should be initiated. All feature updates go directly into beta now
• Close pull requests that merge alpha into beta (this will get handled systematically later)
• Resolve or close and reopen with beta as target any open pull requests currently targeting alpha
• Merge alpha -> beta per toolbox. I will handle this, and probably not through the pull request machinery to avoid spamming the list. In cases where the update somehow ends up being non trivial, I will kick it to a pull request.
• Deal with any tricky merges resulting from above step
• Delete all alpha branches. I will handle this

[dankessler]

Here's a todo-list of toolboxes that have alpha channels that I'll be rolling up

• ConnTool
• ExtractionsScripts
• FirstLevel
• GraphTheory
• ICA
• LSS
• MassUniConn
• MotionRegressors
• MotionSummary
• PPI
• PhysioSummary
• Preprocess
• QualityChecks
• SecondLevel
• cPPI
• core
• jICA
• matlabScripts
• som
• spm8Batch
• svmbatch

[dankessler]

Something weird happened with matlab scripts. I think I accidentally pushed something with my internal prefix system up to universe. I'll loop back to that once I finish the list

Update: it had been absorbed by the appropriate branch, so I deleted it

[dankessler]

One last question. Do we still want to suffix all the toolbox branches with _beta? That's the way it's currently setup. If anybody wants to drop the _beta suffix I'm happy to tweak, but o/w I'm closing this issue.