On Apr 21, 2008, at 11:47 AM, John Cremona wrote: > Sorry if this is a stupid question: if you are going to make a new > complete repository with a patched version of mercurial, does that > mean that native mercurial installations will not work with Sage from > now on, only Sage's "own" version? > > John
No, it's just a changeing the internal names (the hex string called nodeid) of the patches. Mercurial keeps track of what the "parent" of a given changset is, but sometime in the last two years of so they changed how they compute this hash which makes it difficult to export and re-import an entire repository as a string of patches. My hacked mercurial essentially allows one to rename all the old changesets according to the new method without changing the content, history, or timestamps of a repository. - Robert > > 2008/4/21 Robert Bradshaw <[EMAIL PROTECTED]>: >> >> On Apr 21, 2008, at 11:34 AM, mabshoff wrote: >> >>> On Apr 21, 8:24 pm, Robert Bradshaw <[EMAIL PROTECTED]> >>> wrote: >>>> On Apr 21, 2008, at 11:14 AM, William Stein wrote: >>> >>> Hi, >>> >>>>>> Yes, I could. This would mean that no pre-3.0 bundles would >>>>>> apply to >>>>>> post-3.0 (short of re-basing the bundles--and the big one >>>>>> (coercion) >>>>>> I could rebase myself). Patches should be just fine, and most >>>>>> things >>>>>> aren't big enough to warrant bundles. >>> >>> The number of bundles in trac is rather small and most of those >>> bundles either have review issues or shouldn't be bundles in the >>> first >>> place [as you stated above], so applying them to a pre-3.0 tree, >>> extracting the patch and so on should be doable. >> >> Sure. The other concern is people with as-yet unsubmitted code on >> their own computers. One will no longer be able to pull/push. (Does >> the current upgrade try and do that?) >> >> Maybe I could schedule doing it sometime when you're sleeping (does >> that ever happen? :-) 'cause it can't be done in parallel to merging >> very well. >> >> >>>>>> Does anyone know if mercurial >>>>>> 1.0 changed how hashing is done (yet again) or is it finally >>>>>> stable? >>>>>> If so I think this would be a good thing to do. >>>> >>>>> Well this is definitely the right *time* to do it. >>>> >>>> I'll do that then. Probably best to do right before the release, to >>>> not disrupt the development cycle (and as the actual code won't >>>> change (check with a diff) we won't need to be concerned about >>>> breaking Sage). Perhaps the other packages should be changed as >>>> well. >>> >>> The main ones, i.e. extcode and scripts, too and I guess it would be >>> nice to get all the hg repos in the spkgs fixed, too. >> >> Certainly. >> >> >>> Does this >>> require that we upgrade to hg 1.0 or is it fine with the release we >>> ship? Upgrading to 1.0 should be quick and I think I will get it >>> done >>> during the 3.0.1 cycle. >> >> It requires a hacked version of hg I have on my computer, and not >> the >> kind of patch that would ever get merged upstream (without cleanup). >> I just asked the Mercurial guy who answered my original question if >> the hashes changed (again) in 1.0, but I got the impression last >> time >> that they have been sable for a while (just not as long as Sage has >> been around). >> >> - Robert >> >> >> >> >>> >> > > --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
