Hi, I want to set up a git mirror for matplotlib, but I 1) have some minor problems and 2) want to know what others think about this. I'm a native git user and I don't know how to use svn properly. So I try everything to avoid svn. Furthermore, I like git so much that I don't want to give it up when developing for matplotlib. As is well-known, there *is* already a git mirror for mpl, using svnmerge.py, i.e., http://github.com/astraw/matplotlib. Anyway, this repo seems to be not actively maintained for monthes. The last comit on branch trunk is from Nov 12, 2010. Yesterday I tried git-svn on the matplotlib sf repo, and it took quite a long time (already ~500 MB), but it stopped with some error message: RA layer request failed: REPORT of '/svnroot/matplotlib/!svn/vcc/default': SSL negotiation failed: Connection reset by peer (https://matplotlib.svn.sourceforge.net) at /usr/local/libexec/git-core/git-svn line 5061 I could imagine the sf server treated me as some DoS user. I attach a part of the log of how I initialised the git repo (without the middle part), and also some tees of $git branch -a and $git log trunk | head -n 150. I realised that I used different options to git svn init (i.e., I used --std-layout) than stated on http://matplotlib.sourceforge.net/devel/coding_guide.html#using-git, but I cannot see why I should use --trunk=trunk/matplotlib --prefix=svn/. From the logs of $git svn fetch, it looks good. If there is no feedback until today eve here in Germany (in ~6 hrs), I'll try again today evening to $git svn fetch some more of the repo. If there is positive feedback, I'll consider writing to the github guys to give me ~2 Gigs of storage for the mirror. If there is negative feedback, I'll consider giving the project up. Note, that it is inevitable for me to publish the mirror when Ben and I want to use it (for some project we have since a few monthes on mpl). I would suggest to always put git changes *on top* of svn changes, so no use of $git merge trunk, but always $git rebase. This is the only way to make usable diffs which can go in into svn on sf as far as I can see. There is a general problem which will pertain all git users: Once the changes went via svn into the sf repo, they reappear on the github mirror, causing conflicts in the branches on github where they originally came from. The solution to this would be clear, but I think there will be no way to get a consensus to switch from svn to git on mpl completely. Friedrich
On Sat, Jan 22, 2011 at 9:09 AM, Friedrich Romstedt <fri...@gm...> wrote: > Hi, > > I want to set up a git mirror for matplotlib, but I 1) have some minor > problems and 2) want to know what others think about this. Late last year, I did some work to convert the svn repository to git. The code to d the conversion is available at https://github.com/darrendale/mpl2git . The resulting git repo is available at https://github.com/darrendale/matplotlib . I have not pursued the project further because there did not seem to be enough interest in migrating from svn/sourceforge to git/github to justify investing more of my time in the project. > I'm a native git user and I don't know how to use svn properly. I read that as "uninterested in learning how to use svn", and such sentiment is probably a fact of life as many (most?) open source projects move to DVC. In my opinion, matplotlib is likely to draw more contributors if it lowers barriers to entry and uses a DVCS that is growing in popularity, like git/github. Darren
On Sat, Jan 22, 2011 at 10:26 AM, Darren Dale <dsd...@gm...> wrote: > On Sat, Jan 22, 2011 at 9:09 AM, Friedrich Romstedt > <fri...@gm...> wrote: > > Hi, > > > > I want to set up a git mirror for matplotlib, but I 1) have some minor > > problems and 2) want to know what others think about this. > > Late last year, I did some work to convert the svn repository to git. > The code to d the conversion is available at > https://github.com/darrendale/mpl2git . The resulting git repo is > available at https://github.com/darrendale/matplotlib . I have not > pursued the project further because there did not seem to be enough > interest in migrating from svn/sourceforge to git/github to justify > investing more of my time in the project. > > > I'm a native git user and I don't know how to use svn properly. > > I read that as "uninterested in learning how to use svn", and such > sentiment is probably a fact of life as many (most?) open source > projects move to DVC. In my opinion, matplotlib is likely to draw more > contributors if it lowers barriers to entry and uses a DVCS that is > growing in popularity, like git/github. > > Darren > > In discussions with Ryan May on the prospect of switching over to git, it sounds like we have a "Bike Shed" problem where we (the main developers) agree that we agree that a bike shed should be built, but we can't agree on the color to paint it... I think the main source of the huge download size is the data that is coming from the basemap toolkit. I do not think that it would be a good thing to have everyone and their mother needing to do a 'git clone' on their computer and find they have to pull down 500+ MB of stuff to get matplotlib. It is because of this that a straight-forward migration from svn to matplotlib won't be possible. What have been the proposed solutions to dealing with basemap's data? Ben Root
On Sat, Jan 22, 2011 at 11:55 AM, Benjamin Root <ben...@ou...> wrote: > On Sat, Jan 22, 2011 at 10:26 AM, Darren Dale <dsd...@gm...> wrote: >> >> On Sat, Jan 22, 2011 at 9:09 AM, Friedrich Romstedt >> <fri...@gm...> wrote: >> > Hi, >> > >> > I want to set up a git mirror for matplotlib, but I 1) have some minor >> > problems and 2) want to know what others think about this. >> >> Late last year, I did some work to convert the svn repository to git. >> The code to d the conversion is available at >> https://github.com/darrendale/mpl2git . The resulting git repo is >> available at https://github.com/darrendale/matplotlib . I have not >> pursued the project further because there did not seem to be enough >> interest in migrating from svn/sourceforge to git/github to justify >> investing more of my time in the project. >> >> > I'm a native git user and I don't know how to use svn properly. >> >> I read that as "uninterested in learning how to use svn", and such >> sentiment is probably a fact of life as many (most?) open source >> projects move to DVC. In my opinion, matplotlib is likely to draw more >> contributors if it lowers barriers to entry and uses a DVCS that is >> growing in popularity, like git/github. >> >> Darren >> > > In discussions with Ryan May on the prospect of switching over to git, it > sounds like we have a "Bike Shed" problem where we (the main developers) > agree that we agree that a bike shed should be built, but we can't agree on > the color to paint it... In my view, the issue has nothing to do with bikeshedding. > I think the main source of the huge download size is the data that is coming > from the basemap toolkit. I do not think that it would be a good thing to > have everyone and their mother needing to do a 'git clone' on their computer > and find they have to pull down 500+ MB of stuff to get matplotlib. It is > because of this that a straight-forward migration from svn to matplotlib > won't be possible. The git repo at https://github.com/darrendale/matplotlib does not include anything from basemap. > What have been the proposed solutions to dealing with basemap's data? Separate repo?
On 01/22/2011 07:06 AM, Darren Dale wrote: > On Sat, Jan 22, 2011 at 11:55 AM, Benjamin Root<ben...@ou...> wrote: >> On Sat, Jan 22, 2011 at 10:26 AM, Darren Dale<dsd...@gm...> wrote: >>> >>> On Sat, Jan 22, 2011 at 9:09 AM, Friedrich Romstedt >>> <fri...@gm...> wrote: >>>> Hi, >>>> >>>> I want to set up a git mirror for matplotlib, but I 1) have some minor >>>> problems and 2) want to know what others think about this. >>> >>> Late last year, I did some work to convert the svn repository to git. >>> The code to d the conversion is available at >>> https://github.com/darrendale/mpl2git . The resulting git repo is >>> available at https://github.com/darrendale/matplotlib . I have not >>> pursued the project further because there did not seem to be enough >>> interest in migrating from svn/sourceforge to git/github to justify >>> investing more of my time in the project. >>> >>>> I'm a native git user and I don't know how to use svn properly. >>> >>> I read that as "uninterested in learning how to use svn", and such >>> sentiment is probably a fact of life as many (most?) open source >>> projects move to DVC. In my opinion, matplotlib is likely to draw more >>> contributors if it lowers barriers to entry and uses a DVCS that is >>> growing in popularity, like git/github. >>> >>> Darren >>> >> >> In discussions with Ryan May on the prospect of switching over to git, it >> sounds like we have a "Bike Shed" problem where we (the main developers) >> agree that we agree that a bike shed should be built, but we can't agree on >> the color to paint it... > > In my view, the issue has nothing to do with bikeshedding. Agreed. We *did* agree to move to git, and the only problem was that most of us lack the expertise and time to be of much help in actually getting it done. I think there was a bit of a hangup over the question of migrating the bug tracking; but I suspect we could agree that this is a secondary question, and that any such migration, if desired, can be done at leisure *after* taking the primary step of switching the code repo. > >> I think the main source of the huge download size is the data that is coming >> from the basemap toolkit. I do not think that it would be a good thing to >> have everyone and their mother needing to do a 'git clone' on their computer >> and find they have to pull down 500+ MB of stuff to get matplotlib. It is >> because of this that a straight-forward migration from svn to matplotlib >> won't be possible. > > The git repo at https://github.com/darrendale/matplotlib does not > include anything from basemap. The key point being that you are converting matplotlib/trunk/matplotlib, not matplotlib/trunk; the latter is what includes py4science, toolkits (including basemap), sample_data, etc. I think this is the right approach. This doesn't all have to be perfect; it just has to be good enough. I suspect your conversion is good enough. If you are reasonably confident, then I think that if you go ahead and set up a github repo as the new official source tree, along with posting whatever notices are needed for anyone accessing the svn tree, the active developers will say "thank you!", and it will be done. The other person who has been most interested in the nuts and bolts of the conversion is Andrew Straw. Although he is evidently very short of mpl time these days, you might want to try to contact him directly to see if he has thought of anything you might have overlooked. I would like to be included in the group with git write access, unless there is a clear decision to restrict this group to a very small core of gatekeepers. Eric > >> What have been the proposed solutions to dealing with basemap's data? > > Separate repo? >
On Jan 22, 2011 2:28 PM, "Eric Firing" <ef...@ha...> wrote: > > On 01/22/2011 07:06 AM, Darren Dale wrote: > > On Sat, Jan 22, 2011 at 11:55 AM, Benjamin Root<ben...@ou...> wrote: > >> On Sat, Jan 22, 2011 at 10:26 AM, Darren Dale<dsd...@gm...> wrote: > >>> > >>> On Sat, Jan 22, 2011 at 9:09 AM, Friedrich Romstedt > >>> <fri...@gm...> wrote: > >>>> Hi, > >>>> > >>>> I want to set up a git mirror for matplotlib, but I 1) have some minor > >>>> problems and 2) want to know what others think about this. > >>> > >>> Late last year, I did some work to convert the svn repository to git. > >>> The code to d the conversion is available at > >>> https://github.com/darrendale/mpl2git . The resulting git repo is > >>> available at https://github.com/darrendale/matplotlib . I have not > >>> pursued the project further because there did not seem to be enough > >>> interest in migrating from svn/sourceforge to git/github to justify > >>> investing more of my time in the project. > >>> > >>>> I'm a native git user and I don't know how to use svn properly. > >>> > >>> I read that as "uninterested in learning how to use svn", and such > >>> sentiment is probably a fact of life as many (most?) open source > >>> projects move to DVC. In my opinion, matplotlib is likely to draw more > >>> contributors if it lowers barriers to entry and uses a DVCS that is > >>> growing in popularity, like git/github. > >>> > >>> Darren > >>> > >> > >> In discussions with Ryan May on the prospect of switching over to git, it > >> sounds like we have a "Bike Shed" problem where we (the main developers) > >> agree that we agree that a bike shed should be built, but we can't agree on > >> the color to paint it... > > > > In my view, the issue has nothing to do with bikeshedding. > > Agreed. We *did* agree to move to git, and the only problem was that > most of us lack the expertise and time to be of much help in actually > getting it done. I think there was a bit of a hangup over the question > of migrating the bug tracking; but I suspect we could agree that this is > a secondary question, and that any such migration, if desired, can be > done at leisure *after* taking the primary step of switching the code repo. > > > > >> I think the main source of the huge download size is the data that is coming > >> from the basemap toolkit. I do not think that it would be a good thing to > >> have everyone and their mother needing to do a 'git clone' on their computer > >> and find they have to pull down 500+ MB of stuff to get matplotlib. It is > >> because of this that a straight-forward migration from svn to matplotlib > >> won't be possible. > > > > The git repo at https://github.com/darrendale/matplotlib does not > > include anything from basemap. > > The key point being that you are converting matplotlib/trunk/matplotlib, > not matplotlib/trunk; the latter is what includes py4science, toolkits > (including basemap), sample_data, etc. I think this is the right approach. > > This doesn't all have to be perfect; it just has to be good enough. I > suspect your conversion is good enough. If you are reasonably > confident, then I think that if you go ahead and set up a github repo as > the new official source tree, along with posting whatever notices are > needed for anyone accessing the svn tree, the active developers will say > "thank you!", and it will be done. Thanks Eric, that's encouraging. I am reasonably confident, but in the spirit of code review/cya, I would feel better if another dev chimed in that they were also reasonably confident without having to take my word for it. > The other person who has been most interested in the nuts and bolts of > the conversion is Andrew Straw. Although he is evidently very short of > mpl time these days, you might want to try to contact him directly to > see if he has thought of anything you might have overlooked. > > I would like to be included in the group with git write access, unless > there is a clear decision to restrict this group to a very small core of > gatekeepers. > > Eric > > > > >> What have been the proposed solutions to dealing with basemap's data? > > > > Separate repo? > > > > ------------------------------------------------------------------------------ > Special Offer-- Download ArcSight Logger for FREE (a 49ドル USD value)! > Finally, a world-class log management solution at an even better price-free! > Download using promo code Free_Logger_4_Dev2Dev. Offer expires > February 28th, so secure your free ArcSight Logger TODAY! > http://p.sf.net/sfu/arcsight-sfd2d > _______________________________________________ > Matplotlib-devel mailing list > Mat...@li... > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
On Sat, Jan 22, 2011 at 3:26 PM, Darren Dale <dsd...@gm...> wrote: > Thanks Eric, that's encouraging. I am reasonably confident, but in the > spirit of code review/cya, I would feel better if another dev chimed in that > they were also reasonably confident without having to take my word for it. I agree with Eric and others in favor of flipping the switch on the trunk/matplotlib conversion to github. Michael and I made some comments on an earlier thread that we were happy with the svn repo, acknowledging that as core developers we were happy using a central svn repo but understanding that for others without easy commit access, a distributed repo might be preferable. So the original decision made many moons ago to switch has never been questioned, it's just that for many of the developers, it has not been a driving need. But we'll gladly stand behind anyone willing to do the work. There are subtleties and difficulties around the mpl svn directories that depend on trunk/matplotlib: sample_data, basemap, sampledoc, etc. I think the solution is the one advocated above: just punt on it. Let's convert the trunk and rely on good practice to keep the others in sync. Ie, when basemap does a release, it can just tag the release notes with "depends on matplotlib release XX or revision YY". Likewise with sampledata, etc. The official repo needs to be https://github.com/matplotlib, which is the matplotlib organization. Darren is already a member, and it looks like he has more bandwidth at this point than Andrew, so Darren if you are ready to "flip the switch" and make an official github repo under this organization, go for it. Once we get the trunk active, we'll worry about the rest, like migrating the release branch. Of course, if Andrew as the original force to move to github, has any comments or concerns, we're certainly receptive to them. But we have a recent release out, the buildbot is broken currently anyhow, and this looks like a perfect time to make the move. JDH
On 23-Jan-11 04:05, John Hunter wrote: > > Darren > if you are ready to "flip the switch" and make an official github repo > under this organization, go for it. Once we get the trunk active, > we'll worry about the rest, like migrating the release branch. Of > course, if Andrew as the original force to move to github, has any > comments or concerns, we're certainly receptive to them. But we have > a recent release out, the buildbot is broken currently anyhow, and > this looks like a perfect time to make the move. +1. And for what it's worth, I keep nagging the IT people at my new employer to set me up the virtual machines for the new buildslaves... -Andrew
On Sun, Jan 23, 2011 at 6:25 AM, Andrew Straw <str...@as...> wrote: > On 23-Jan-11 04:05, John Hunter wrote: >> >> Darren >> if you are ready to "flip the switch" and make an official github repo >> under this organization, go for it. Once we get the trunk active, >> we'll worry about the rest, like migrating the release branch. Of >> course, if Andrew as the original force to move to github, has any >> comments or concerns, we're certainly receptive to them. But we have >> a recent release out, the buildbot is broken currently anyhow, and >> this looks like a perfect time to make the move. > > +1. > > And for what it's worth, I keep nagging the IT people at my new employer to > set me up the virtual machines for the new buildslaves... I need to improve the authorship mapping, so the authors of svn commits will be identified using their git information in the new repository. For the following svn accounts, I need "Real Name <re...@em...o>" information as it will appear when committing to the new git repository (not your old svn info, unless it will be the same). Look in the [user] section of ~/.gitconfig, if you have one. jdh2358 efiring mdboom mdehoon jswhit weathergod leejjoon jouni jrevans ryanmay ianthomas23 cmoad pivanov314 Please send it to me ASAP.
On Sun, Jan 23, 2011 at 9:18 AM, Darren Dale <dsd...@gm...> wrote: > On Sun, Jan 23, 2011 at 6:25 AM, Andrew Straw <str...@as...> wrote: >> On 23-Jan-11 04:05, John Hunter wrote: >>> >>> Darren >>> if you are ready to "flip the switch" and make an official github repo >>> under this organization, go for it. Once we get the trunk active, >>> we'll worry about the rest, like migrating the release branch. Of >>> course, if Andrew as the original force to move to github, has any >>> comments or concerns, we're certainly receptive to them. But we have >>> a recent release out, the buildbot is broken currently anyhow, and >>> this looks like a perfect time to make the move. >> >> +1. >> >> And for what it's worth, I keep nagging the IT people at my new employer to >> set me up the virtual machines for the new buildslaves... > > I need to improve the authorship mapping, so the authors of svn > commits will be identified using their git information in the new > repository. For the following svn accounts, I need "Real Name > <re...@em...o>" information as it will appear when committing to the > new git repository (not your old svn info, unless it will be the > same). Look in the [user] section of ~/.gitconfig, if you have one. > Please send it to me ASAP. I think we are getting close. Here are the remaining commit authors that do not have git commit information. Do we know how to reach some of these folks directly to ask if they want to provide it? mdboom mdehoon jswhit jrevans cmoad heeres mmetz_bn sameerd pkienzle dmkaplan nnemec stevech edin1 kmcivor teoliphant barrett greglielens jvoss2 jaytmiller perrygreenfield jodonoghue
On 1/24/11 6:10 AM, Darren Dale wrote: > On Sun, Jan 23, 2011 at 9:18 AM, Darren Dale<dsd...@gm...> wrote: >> On Sun, Jan 23, 2011 at 6:25 AM, Andrew Straw<str...@as...> wrote: >>> On 23-Jan-11 04:05, John Hunter wrote: >>>> Darren >>>> if you are ready to "flip the switch" and make an official github repo >>>> under this organization, go for it. Once we get the trunk active, >>>> we'll worry about the rest, like migrating the release branch. Of >>>> course, if Andrew as the original force to move to github, has any >>>> comments or concerns, we're certainly receptive to them. But we have >>>> a recent release out, the buildbot is broken currently anyhow, and >>>> this looks like a perfect time to make the move. >>> +1. >>> >>> And for what it's worth, I keep nagging the IT people at my new employer to >>> set me up the virtual machines for the new buildslaves... >> I need to improve the authorship mapping, so the authors of svn >> commits will be identified using their git information in the new >> repository. For the following svn accounts, I need "Real Name >> <re...@em...o>" information as it will appear when committing to the >> new git repository (not your old svn info, unless it will be the >> same). Look in the [user] section of ~/.gitconfig, if you have one. >> Please send it to me ASAP. > I think we are getting close. Here are the remaining commit authors > that do not have git commit information. Do we know how to reach some > of these folks directly to ask if they want to provide it? > > mdboom > mdehoon > jswhit js...@gi... Jeff Whitaker <js...@fa...> Regarding basemap, what do people recommend? Should I create a separate github project for basemap and it's data? -Jeff > jrevans > cmoad > heeres > mmetz_bn > sameerd > pkienzle > dmkaplan > nnemec > stevech > edin1 > kmcivor > teoliphant > barrett > greglielens > jvoss2 > jaytmiller > perrygreenfield > jodonoghue > > ------------------------------------------------------------------------------ > Special Offer-- Download ArcSight Logger for FREE (a 49ドル USD value)! > Finally, a world-class log management solution at an even better price-free! > Download using promo code Free_Logger_4_Dev2Dev. Offer expires > February 28th, so secure your free ArcSight Logger TODAY! > http://p.sf.net/sfu/arcsight-sfd2d > _______________________________________________ > Matplotlib-devel mailing list > Mat...@li... > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
On Mon, Jan 24, 2011 at 7:18 AM, Jeff Whitaker <js...@fa...> wrote: > Regarding basemap, what do people recommend? Should I create a separate > github project for basemap and it's data? I think that would be ideal. Unlink svn, in git you can't have full featured subdirectory checkouts, and since some of the projects, basemap included are too large to include with the core, the best solution is to put them in a separate repository. This complicates keeping versions in sync, but it seems like the best solution. JDH
On Mon, Jan 24, 2011 at 8:48 AM, John Hunter <jd...@gm...> wrote: > On Mon, Jan 24, 2011 at 7:18 AM, Jeff Whitaker <js...@fa...> wrote: > >> Regarding basemap, what do people recommend? Should I create a separate >> github project for basemap and it's data? > > I think that would be ideal. Unlink svn, in git you can't have full > featured subdirectory checkouts, and since some of the projects, > basemap included are too large to include with the core, the best > solution is to put them in a separate repository. This complicates > keeping versions in sync, but it seems like the best solution. That said, I have the conversion routines set up right now to create a separate "toolkits" repository, which currently includes all the contents of trunk/toolkits. I can tailor this further to split the toolkits up and create a repository for each, if thats how people want to do it. Those repos could either be hosted as separate repositories under the matplotlib github organization, or maybe better as a repository under a separate github project. In case of the latter, I can temporarily publish the repository at my own account, long enough for Jeff to clone it (don't fork it using githubs fork button!) and push it to its final destination. Darren
On 1/24/11 7:11 AM, Darren Dale wrote: > On Mon, Jan 24, 2011 at 8:48 AM, John Hunter<jd...@gm...> wrote: >> On Mon, Jan 24, 2011 at 7:18 AM, Jeff Whitaker<js...@fa...> wrote: >> >>> Regarding basemap, what do people recommend? Should I create a separate >>> github project for basemap and it's data? >> I think that would be ideal. Unlink svn, in git you can't have full >> featured subdirectory checkouts, and since some of the projects, >> basemap included are too large to include with the core, the best >> solution is to put them in a separate repository. This complicates >> keeping versions in sync, but it seems like the best solution. > That said, I have the conversion routines set up right now to create a > separate "toolkits" repository, which currently includes all the > contents of trunk/toolkits. I can tailor this further to split the > toolkits up and create a repository for each, if thats how people want > to do it. Those repos could either be hosted as separate repositories > under the matplotlib github organization, or maybe better as a > repository under a separate github project. In case of the latter, I > can temporarily publish the repository at my own account, long enough > for Jeff to clone it (don't fork it using githubs fork button!) and > push it to its final destination. > > Darren Darren: I think having them as separate repos under the matplotlib organization is fine. If you could do that I would be grateful. -Jeff
There is a potential problem converting the entire basemap history to git. In svn commit 4418, trunk/toolkits had basemap and basemap-testing directories. In commit 4419, basemap was renamed basemap-0.9.6.1, so there was only basemap-0.9.6.1 and basemap-testing. In commit 4420, basemap-testing is renamed basemap. The git history only goes back as far as svn4420, it looks like the conversion routines get confused by the temporary absence of the basemap directory. I'm trying to find a workaround, but if I can't... ?
2011年1月25日 12:19:37 -0500, Darren Dale wrote: > There is a potential problem converting the entire basemap history to > git. In svn commit 4418, trunk/toolkits had basemap and basemap-testing > directories. In commit 4419, basemap was renamed basemap-0.9.6.1, so > there was only basemap-0.9.6.1 and basemap-testing. In commit 4420, > basemap-testing is renamed basemap. The git history only goes back as > far as svn4420, it looks like the conversion routines get confused by > the temporary absence of the basemap directory. > > I'm trying to find a workaround, but if I can't... ? You can maybe do it like this: 1) Write matplotlib.rules so that all of the directories where basemap stuff has been ends in the basemap repository. (I'm assuming this does not error out...) 2) This will create a number of separate heads in the basemap repo that do not share common history. 3) Add graft rules in matplotlib.grafts to stitch the disconnected history graphs together. This happened also with Numpy: part of the old history had a this sort of a rename and so part of the history was not connected to the main graph. So I just stitched the graphs together manually. -- Pauli Virtanen
On Tue, Jan 25, 2011 at 1:31 PM, Pauli Virtanen <pa...@ik...> wrote: > 2011年1月25日 12:19:37 -0500, Darren Dale wrote: >> There is a potential problem converting the entire basemap history to >> git. In svn commit 4418, trunk/toolkits had basemap and basemap-testing >> directories. In commit 4419, basemap was renamed basemap-0.9.6.1, so >> there was only basemap-0.9.6.1 and basemap-testing. In commit 4420, >> basemap-testing is renamed basemap. The git history only goes back as >> far as svn4420, it looks like the conversion routines get confused by >> the temporary absence of the basemap directory. >> >> I'm trying to find a workaround, but if I can't... ? > > You can maybe do it like this: > > 1) Write matplotlib.rules so that all of the directories where basemap > stuff has been ends in the basemap repository. (I'm assuming this does > not error out...) Aha! I thought I had tried that. Thanks. > 2) This will create a number of separate heads in the basemap repo that > do not share common history. > > 3) Add graft rules in matplotlib.grafts to stitch the disconnected > history graphs together. Mercifully, the latest checkout of svn2git seems to take care of that. I've developed a wicked headache. Jeff, the repository is temporarily available at https://github.com/darrendale/basemap . It would be really helpful if you would have a look at the network graph at https://github.com/darrendale/basemap/network to make sure there are no surprises, maybe clone the repository and check that the working directory is identical to your svn checkout. Darren
On 1/25/11 1:06 PM, Darren Dale wrote: > On Tue, Jan 25, 2011 at 1:31 PM, Pauli Virtanen<pa...@ik...> wrote: >> 2011年1月25日 12:19:37 -0500, Darren Dale wrote: >>> There is a potential problem converting the entire basemap history to >>> git. In svn commit 4418, trunk/toolkits had basemap and basemap-testing >>> directories. In commit 4419, basemap was renamed basemap-0.9.6.1, so >>> there was only basemap-0.9.6.1 and basemap-testing. In commit 4420, >>> basemap-testing is renamed basemap. The git history only goes back as >>> far as svn4420, it looks like the conversion routines get confused by >>> the temporary absence of the basemap directory. >>> >>> I'm trying to find a workaround, but if I can't... ? >> You can maybe do it like this: >> >> 1) Write matplotlib.rules so that all of the directories where basemap >> stuff has been ends in the basemap repository. (I'm assuming this does >> not error out...) > Aha! I thought I had tried that. Thanks. > >> 2) This will create a number of separate heads in the basemap repo that >> do not share common history. >> >> 3) Add graft rules in matplotlib.grafts to stitch the disconnected >> history graphs together. > Mercifully, the latest checkout of svn2git seems to take care of that. > I've developed a wicked headache. > > Jeff, the repository is temporarily available at > https://github.com/darrendale/basemap . It would be really helpful if > you would have a look at the network graph at > https://github.com/darrendale/basemap/network to make sure there are > no surprises, maybe clone the repository and check that the working > directory is identical to your svn checkout. > > Darren Darren: https://github.com/darrendale/basemap looks fine, thanks! -Jeff
On 1/22/11 10:26 AM, Darren Dale wrote: > On Sat, Jan 22, 2011 at 9:09 AM, Friedrich Romstedt > <fri...@gm...> wrote: >> Hi, >> >> I want to set up a git mirror for matplotlib, but I 1) have some minor >> problems and 2) want to know what others think about this. > > Late last year, I did some work to convert the svn repository to git. > The code to d the conversion is available at > https://github.com/darrendale/mpl2git . The resulting git repo is > available at https://github.com/darrendale/matplotlib . I have not > pursued the project further because there did not seem to be enough > interest in migrating from svn/sourceforge to git/github to justify > investing more of my time in the project. > >> I'm a native git user and I don't know how to use svn properly. > > I read that as "uninterested in learning how to use svn", and such > sentiment is probably a fact of life as many (most?) open source > projects move to DVC. In my opinion, matplotlib is likely to draw more > contributors if it lowers barriers to entry and uses a DVCS that is > growing in popularity, like git/github. +1. There have been several times I've wanted to fix something in the documentation, and would have via github if it had been available. As it was, I sent an email to the list and someone else did it for me. In my case, the fixes got in, but it took more time and effort for me and presumably also for the people making the fix (compared to just responding to a pull request, for example). I do remember debating whether it was worth the trouble to report the typo when I realized I had to send an email describing exactly where the typo was. Thanks, Jason
2011年1月22日 Eric Firing <ef...@ha...>: > This doesn't all have to be perfect; it just has to be good enough. I > suspect your conversion is good enough. If you are reasonably > confident, then I think that if you go ahead and set up a github repo as > the new official source tree, along with posting whatever notices are > needed for anyone accessing the svn tree, the active developers will say > "thank you!", and it will be done. > > The other person who has been most interested in the nuts and bolts of > the conversion is Andrew Straw. Although he is evidently very short of > mpl time these days, you might want to try to contact him directly to > see if he has thought of anything you might have overlooked. > > I would like to be included in the group with git write access, unless > there is a clear decision to restrict this group to a very small core of > gatekeepers. To support Darren's conversion work, here are the links to some numpy discussions which might be, I think, of interest here: [Numpy-discussion] curious about how people would feel about moving to github: http://thread.gmane.org/gmane.comp.python.numeric.general/38133 [Numpy-discussion] Technicalities of the SVN -> GIT transition: http://thread.gmane.org/gmane.comp.python.numeric.general/38263 [Numpy-discussion] First shot at svn->git conversion: http://thread.gmane.org/gmane.comp.python.numeric.general/39443 [Numpy-discussion] update on the transition to git/github: http://thread.gmane.org/gmane.comp.python.numeric.general/39102 [Numpy-discussion] Github migration?: http://thread.gmane.org/gmane.comp.python.numeric.general/40020 [Numpy-discussion] Can we freeze the subversion repository and move to github this week?: http://mail.scipy.org/pipermail/numpy-discussion/2010-September/thread.html#52744 (this thread is apparently lost on gmane) [Numpy-discussion] Numpy SVN frozen; move to Git: http://mail.scipy.org/pipermail/numpy-discussion/2010-September/thread.html#52811 (lost on Gmane too) [Numpy-discussion] Commit rights on github: http://thread.gmane.org/gmane.comp.python.numeric.general/40812 [Numpy-discussion] Development workflow: http://thread.gmane.org/gmane.comp.python.numeric.general/40816 [Numpy-discussion] Another merge at github: http://thread.gmane.org/gmane.comp.python.numeric.general/40967 [Numpy-discussion] whitespace in git repo: http://thread.gmane.org/gmane.comp.python.numeric.general/41013 Okay, I think that's it. I remember there were some problems getting the complete svn history correctly transferred (it should be in above), but I cannot tell specifics; we would have to ask on the numpy list for this. Also, there was lots of discussion how to design the commit rights (somewhere included above); since on github "pull requests" can be filed, getting changes in with a committer different from the author is simple. Cf. http://github.com/numpy/ (list of members of the github organization). It is possible to register "organization" accounts on GitHub, with several members. Friedrich
On Jan 22, 2011 3:38 PM, "Friedrich Romstedt" <fri...@gm...> wrote: > > 2011年1月22日 Eric Firing <ef...@ha...>: > > This doesn't all have to be perfect; it just has to be good enough. I > > suspect your conversion is good enough. If you are reasonably > > confident, then I think that if you go ahead and set up a github repo as > > the new official source tree, along with posting whatever notices are > > needed for anyone accessing the svn tree, the active developers will say > > "thank you!", and it will be done. > > > > The other person who has been most interested in the nuts and bolts of > > the conversion is Andrew Straw. Although he is evidently very short of > > mpl time these days, you might want to try to contact him directly to > > see if he has thought of anything you might have overlooked. > > > > I would like to be included in the group with git write access, unless > > there is a clear decision to restrict this group to a very small core of > > gatekeepers. > > To support Darren's conversion work, here are the links to some numpy > discussions which might be, I think, of interest here: > > [Numpy-discussion] curious about how people would feel about moving to github: > http://thread.gmane.org/gmane.comp.python.numeric.general/38133 > > [Numpy-discussion] Technicalities of the SVN -> GIT transition: > http://thread.gmane.org/gmane.comp.python.numeric.general/38263 > > [Numpy-discussion] First shot at svn->git conversion: > http://thread.gmane.org/gmane.comp.python.numeric.general/39443 > > [Numpy-discussion] update on the transition to git/github: > http://thread.gmane.org/gmane.comp.python.numeric.general/39102 > > [Numpy-discussion] Github migration?: > http://thread.gmane.org/gmane.comp.python.numeric.general/40020 > > [Numpy-discussion] Can we freeze the subversion repository and move to > github this week?: > http://mail.scipy.org/pipermail/numpy-discussion/2010-September/thread.html#52744 > (this thread is apparently lost on gmane) > > [Numpy-discussion] Numpy SVN frozen; move to Git: > http://mail.scipy.org/pipermail/numpy-discussion/2010-September/thread.html#52811 > (lost on Gmane too) > > [Numpy-discussion] Commit rights on github: > http://thread.gmane.org/gmane.comp.python.numeric.general/40812 > > [Numpy-discussion] Development workflow: > http://thread.gmane.org/gmane.comp.python.numeric.general/40816 > > [Numpy-discussion] Another merge at github: > http://thread.gmane.org/gmane.comp.python.numeric.general/40967 > > [Numpy-discussion] whitespace in git repo: > http://thread.gmane.org/gmane.comp.python.numeric.general/41013 > > Okay, I think that's it. > > I remember there were some problems getting the complete svn history > correctly transferred (it should be in above), but I cannot tell > specifics; we would have to ask on the numpy list for this. I used Pauli's work on numpy as a template to convert the mpl tell, and I am fairly confident that the history was correctly transferred. svn2git seems to have improved since numpy made the leap, and appears to have handled the merge of the transforms branch correctly, which may have been a considerable stress-test for the conversion routines. That said, I would be more comfortable saying "lets do this" if I heard from more of the debate that the drop looks ok. > Also, there was lots of discussion how to design the commit rights > (somewhere included above); since on github "pull requests" can be > filed, getting changes in with a committer different from the author > is simple. Cf. http://github.com/numpy/ (list of members of the > github organization). > > It is possible to register "organization" accounts on GitHub, with > several members. > > Friedrich > > ------------------------------------------------------------------------------ > Special Offer-- Download ArcSight Logger for FREE (a 49ドル USD value)! > Finally, a world-class log management solution at an even better price-free! > Download using promo code Free_Logger_4_Dev2Dev. Offer expires > February 28th, so secure your free ArcSight Logger TODAY! > http://p.sf.net/sfu/arcsight-sfd2d > _______________________________________________ > Matplotlib-devel mailing list > Mat...@li... > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
On Sat, Jan 22, 2011 at 4:23 PM, Darren Dale <dsd...@gm...> wrote: > That said, I would be more comfortable saying "lets do this" if I heard from > more of the debate that the drop looks ok. In case anyone was wondering, that was the result of the android spell checker, not alzheimers.
Hi, >> >> What have been the proposed solutions to dealing with basemap's data? >> > >> > Separate repo? I just fished up some previous discussions: http://comments.gmane.org/gmane.comp.python.matplotlib.devel/8275 http://comments.gmane.org/gmane.comp.python.matplotlib.devel/8461 Do I remember correctly that a plan was needed to keep track of the relationship of matplotlib-proper, the sample data and basemap, when they are not part of the same repository? See y'all, Matthew
On Jan 22, 2011 5:28 PM, "Matthew Brett" <mat...@gm...> wrote: > > Hi, > > >> >> What have been the proposed solutions to dealing with basemap's data? > >> > > >> > Separate repo? > > I just fished up some previous discussions: > > http://comments.gmane.org/gmane.comp.python.matplotlib.devel/8275 > http://comments.gmane.org/gmane.comp.python.matplotlib.devel/8461 > > Do I remember correctly that a plan was needed to keep track of the > relationship of matplotlib-proper, the sample data and basemap, when > they are not part of the same repository? I think this could be done with a separate repo, using modules to pull in the various components and keep the relationship in sync. I'm using modules in one of my own projects, and have been satisfied with the way it works. But we should really hear from the basemap devs, I don't know if they are on board with a switch to git.
On Jan 22, 2011 9:06 PM, "Darren Dale" <dsd...@gm...> wrote: > > > On Jan 22, 2011 5:28 PM, "Matthew Brett" <mat...@gm...> wrote: > > > > Hi, > > > > >> >> What have been the proposed solutions to dealing with basemap's data? > > >> > > > >> > Separate repo? > > > > I just fished up some previous discussions: > > > > http://comments.gmane.org/gmane.comp.python.matplotlib.devel/8275 > > http://comments.gmane.org/gmane.comp.python.matplotlib.devel/8461 > > > > Do I remember correctly that a plan was needed to keep track of the > > relationship of matplotlib-proper, the sample data and basemap, when > > they are not part of the same repository? > > I think this could be done with a separate repo, using modules to pull in the various components and keep the relationship in sync. I'm using modules in one of my own projects, and have been satisfied with the way it works. But we should really hear from the basemap devs, I don't know if they are on board with a switch to git. Damn spellchecker. That should read submodules, not modules.