-
Notifications
You must be signed in to change notification settings - Fork 109
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fix for issue #755 #881
fix for issue #755 #881
Conversation
You might want to work with branches instead of master branch to avoid such issues. If origin is your fork and upstream is the official repo what happens when you do
|
@ilayn, thank you for your suggestion. I'll be happy to test it after one more question: I just did a commit when I saw your message. Is that a problem when executing these commands? |
When you have your changes and also you pick upstream changes thry stack on top of each other in a scrambling fashion confusing git. If you keep your master branch pristine and only use it tobstay up to date also work on features on your local branches theybstay mutually exclusive |
@ilayn, ok, let's see if I have understood you right.
Right? I am still not sure if I should run your suggested commands though ... |
@Mo-Gul |
Also the changes seem to be rather trivial so in any case can be redone without too much effort. |
…mple in `pgfmanual-en-tikz-graphs.tex` work closes issue pgf-tikz#755)
I've rebased it for you. In the future could you please also format your commit messages like
where |
Thank you guys for your input. But I guess this is at my current git knowledge too high for me. @hmenke, I'll do my best. Besides that this pull request is finished from my side. The above mentioned commit relates to the codeexamples and not to the issue handled here. |
It's not you, it's Git. Unfortunately, just like its creator, it commits cardinal sins in usability design. One of the most annoying one is that the function changes its meaning depending on its arguments (matlab anyone?). So here is a simple workflow; if you stay in it, it's all going to be fine. Anytime you want to do something, sync your fork to the official one(I mentioned it above but here again)
This keeps your fork up-to-date such that when you open a PR you don't get annoying merge conflicts because something else got in between. Then anytime you want to do something new, create a branch git checkout -b <insert-branch-name> How you read this ? Switch to my
It works without arguments because the default is origin. It will complain (only the first time) that you don't have this branch on remote. But fortunately it will tell you what to do on the screen. Thanks git
Then you can come to github and open a pull request by pushing buttons. If you stick to this machinery 90% of your problems would go away. For the remaining 10% ask before you do it. |
@ilayn, aah, ok. Currently my workflow to keep in sync with the main repository was something like
and it very often worked fine... I don't know the exact syntax because I am using the TortoiseGit GUI. (And was there a I just invited Henri and you so you have write access to my repository too. So you maybe can fix stuff for me for the remaining 10%. Here the question is, how do I know/recognize that problems will appear upfront? Maybe it's because I did "something" to my repository before Henri rebased it, it seems something is wrong with my newly created branch https://github.com/Mo-Gul/pgf/tree/PimpLuaExamples as well. (It seems so when I try to open a pull request from that branch.) Could you fix that as well (by rebasing or whatever) and state me what then I have to do on my machine as well so everything is fine (again)? Many thanks in advance for your help!! |
I would suggest not to use any GUIs at least until you get some affinity to the whole thing. Or if you really must; use GitHub Desktop instead which is kinda nice. You don't need to invite us because any maintainer can modify the pull-request branches. So that would be an overkill. It's this line that causes a lot of trouble. So better avoid that one, and try to treat the upstream as the canonical source. You are also a maintainer so better not push or merge or anything else that might compromise the
|
@ilayn, "getting some affinity" is the problem. I guess 90% or more the time I use/need Git is when doing stuff for TikZ. And that is only from time-to-time. So also getting used to the GUI isn't really an option at the moment. To remember the 3 stated lines above I have prominently written down on my desktop because usually I have forgotten them the next time I need them. Of course that is not an excuse but an explanation why I sometimes mess up stuff in Git. But I hope that my contributions have a (much) higher value than what I mess up with Git and so it is still beneficial to clean up the mess I leave behind ;) No, I will not push, merge or whatever on the this master. If I should do so, I'll do it by pushing buttons here on GitHub which hopefully will not produce a mess. To see forward: I'll took a note of your suggestion which I'll follow the next time I work on something (so after |
Well as the saying goes, "when in Rome, do as the Romans do" 😃 Don't fear the learning curve, the key is executing commands slowly and reading every bit of warnings. You won't believe how many times I screwed up in public repos elsewhere. But unfortunately under this stupid user interface Git is fantastically powerful. So better put it in your arsenal. I'll check the other issue soon. |
What will work for the other PR will not here (in the same way) because the changes for this PR were done in |
Merged in 197450c Please also have a look at the new commit message to get a feeling for what is sensible and what is not. |
(For whatever reason there seem to some commits as well that were already merged, but I don't know why. Either I did something wrong or could the reason be the
rebase
which was done? Do you have an idea? If not, do you have an advice how to "fix" my fork without deleting and reforking it?)