Definitely a useful addition for anything that git can't really track changes in, binary files for example. When used in conjunction with a history rewrite (only do if you are sure, everyone will have to reclone) it can make things a lot neater.
LFS makes it awesome to track other types of files in git if you just want reference media material in your docs for instance, super great way of handling a 50mb video file you show in a readme or something ⭐
Hi , what if you are trying to replace an existing file in your repo with a file in the lfs...i created a .zip and puashed it to my repo using lfs but now i cant figure out how i replace the existing file in the repo with a file that is in that zip lfs / .gitattributes folder??
what if the file you're trying to upload/track has no extension? I have a file with a name for a game engine that just has a name but no file extension in its details tap in properties. Do we just list the name of the file itself?
I don't see how this changes anything? If those larger files are not being changed, it would get treated like any other unchanged file in your repository. If it gets changed, that changed version is uploaded alongside the previous version that's already on the server. You're still storing large files regardless, why split them into two different servers to create more complexity in the system to track what goes where?
It may help to consider what happens when others try to clone a repo with larger files. Every revision to a large binary file would cause cloning to be pretty time-consuming. LFS helps cloning, pushing, and pulling take less time.
@@GitKrakenHow? The files still need to be downloaded/uploaded. What if the LFS server is down but the repo host isn't? What makes this better than simply having a faster server with more storage?
@@SomeNerd361 To save time with pushes, pulls, and clones. With LFS set up, your pushing MB of data instead of potentially GB of data because LFS has replaced the larger binary files with tiny pointer files. It may be true that incidental binary file changes may not be worth setting up LFS, and it's up to you and collaborators to decide if there's enough pain to make it worth it.
@@GitKrakenThat doesn't answer my question(s). It's not like the files used for LFS magically go from one local machine to another when you clone a repo, they still need to be stored in a server somewhere even if you split the repo into multiple parts (which is what LFS effectively does, I get that), 1 GB of data is still 1GB of data. So again, how does this actually change things for the better instead of creating unneeded complexity? I feel like I'm asking what should be fundamental questions about how the system works, here.
Like, I don't see what's so confusing about what I'm asking, here. The project is the project is the project. It doesn't matter how you break it up, you still need to download THE WHOLE PROJECT. Files still have to be versioned, how is this not making sense to you? A company who's livelihood is ingrained in understanding how this works?
I totally dig your video except for the fact that you chose to not cover installing git LFS on other platforms when the video is ABOUT Git LFS. You missed an opportunity to teach something core to this whole thing and that's a travesty. I can't downvote this because it's a decent video, but I can't upvote something where people go out of there way to teach something core to the video's topic. 😕
Definitely a useful addition for anything that git can't really track changes in, binary files for example. When used in conjunction with a history rewrite (only do if you are sure, everyone will have to reclone) it can make things a lot neater.
Thank you, sir, this video was very helpful!
LFS makes it awesome to track other types of files in git if you just want reference media material in your docs for instance, super great way of handling a 50mb video file you show in a readme or something ⭐
This helped perfectly, thank you!
love this guys, keep up the good work
Do you have to do that for every single file type? I'm using it for unity projects and there many types. Is there a shortcut?
why did you set your shell prompt like that, there is supposed to be a space after the $ lol.
Not taking the time to add cmd for Linux and Windows... :( not much of a tutorial when you exclude 80% of you userbase is it!
one thing I cloud't understand , where to put the original file ?
Hi , what if you are trying to replace an existing file in your repo with a file in the lfs...i created a .zip and puashed it to my repo using lfs but now i cant figure out how i replace the existing file in the repo with a file that is in that zip lfs / .gitattributes folder??
what if the file you're trying to upload/track has no extension? I have a file with a name for a game engine that just has a name but no file extension in its details tap in properties. Do we just list the name of the file itself?
I assume git uses a glob pattern to find files. So adding a file name should work. It's my speculation only
I don't see how this changes anything? If those larger files are not being changed, it would get treated like any other unchanged file in your repository. If it gets changed, that changed version is uploaded alongside the previous version that's already on the server.
You're still storing large files regardless, why split them into two different servers to create more complexity in the system to track what goes where?
It may help to consider what happens when others try to clone a repo with larger files. Every revision to a large binary file would cause cloning to be pretty time-consuming. LFS helps cloning, pushing, and pulling take less time.
@@GitKrakenHow? The files still need to be downloaded/uploaded. What if the LFS server is down but the repo host isn't?
What makes this better than simply having a faster server with more storage?
@@SomeNerd361 To save time with pushes, pulls, and clones. With LFS set up, your pushing MB of data instead of potentially GB of data because LFS has replaced the larger binary files with tiny pointer files. It may be true that incidental binary file changes may not be worth setting up LFS, and it's up to you and collaborators to decide if there's enough pain to make it worth it.
@@GitKrakenThat doesn't answer my question(s). It's not like the files used for LFS magically go from one local machine to another when you clone a repo, they still need to be stored in a server somewhere even if you split the repo into multiple parts (which is what LFS effectively does, I get that), 1 GB of data is still 1GB of data. So again, how does this actually change things for the better instead of creating unneeded complexity?
I feel like I'm asking what should be fundamental questions about how the system works, here.
Like, I don't see what's so confusing about what I'm asking, here. The project is the project is the project.
It doesn't matter how you break it up, you still need to download THE WHOLE PROJECT. Files still have to be versioned, how is this not making sense to you? A company who's livelihood is ingrained in understanding how this works?
I totally dig your video except for the fact that you chose to not cover installing git LFS on other platforms when the video is ABOUT Git LFS.
You missed an opportunity to teach something core to this whole thing and that's a travesty.
I can't downvote this because it's a decent video, but I can't upvote something where people go out of there way to teach something core to the video's topic. 😕