help-with-git

[winter sierra]

OK.

I go to try that. I'll come back here when done / or problems

[pallid thicket]

sounds good. You can keep a copy of your edited examples and then later only you'll be able to paste them in to the new working repo.

[winter sierra]

OK, I deleted the fork from your fork. Then I created a fork from Adafruit_CircuitPython_DisplayIO_Layout. Sofar so good. How do I now get the updates of your tab_layout branch

In the just created fork I created a branch called 'hotplug_sensor_examples'

Here is shown the new branch

[pallid thicket]

@winter sierra the why I know how is to create your branch from mine. To do that you'll need to add the foamyguy remote to your repo. so you can run:

git remote add foamyguy https://github.com/FoamyGuy/Adafruit_CircuitPython_DisplayIO_Layout.git

then git fetch foamyguy

[winter sierra]

tnx

[pallid thicket]

Once you've done those then you should be able to use a command like this to make a new branch, sorry I don't know how to do it with an existing branch so you'll end up making another new one with this.

git checkout -b [your_branch_name] foamyguy/tab_layout --no-track

I'll be joining the weekly meeting here in a moment so rill probably respond a little slower for a bit during that. But I'll keep checking in and will be around more afterward.

[winter sierra]

OK, I'll just show you now the result screenshot. No errors

here it is

many thanks for all your help today. I learned quite some new things. Was a bit heavy for me.

yep, your files have been added (much more than these)

[pallid thicket]

You're very welcome, I'm happy to help out. And yep there is a lot to the git stuff that can be daunting. I picked it up fairly late in my development journey and am still picking up new things around it.

[pallid thicket]

Okay, meeting has concluded and I grabbed a snack and am getting back to work now. Let me know where you ended up and if you had any further trouble.

[winter sierra]

@pallid thicket I just booted-up my Ubuntu Linux. Go to try the work there. IN MS Windows / GitHub Desktop app continued to receive the same error msg.

[pallid thicket]

drat. Let me know how it goes on ubuntu. I've got my stream running doing some PR testing. I could walk thgough the terminal commands if you do have trouble and want a chance to see / follow along.

[winter sierra]

OK

[winter sierra]

@pallid thicket I stop for today. Also within Ubuntu Linux a lot of mess. I probably do things wrong. For example: pylint repeats reporting errors of line endings with CRLF instead of LF but I installed and ran dos2unix to cleanup these CRLF line endings but the pylint continues to report those errors even after I did a 'pre-commit clean'.

[pallid thicket]

Okay. I did find a version of github desktop app that works on my computer. I'll try to get more familiar with it and I could potentially offer better help using that. It does seem like a nice interface and similar to a lot of the helpers that I use from pycharm.

[winter sierra]

Ok Tim, thank you so much for trying to help me!

I go to try again today, then I will make a copy of all the output that I get from the pre-commit runs

[pallid thicket]

@winter sierra I have checked out the latest version of your examples from the Dropbox and started working on figuring out the correct commands to run to get set up for a PR.

I'm seeing the pre-commit checks fail for 2 main reasons 1) the new image file BGimage4.bmp needs to have a license file alongside of it with a similar name. (see the ones in examples/bmps/ for example.)

2. the line endings of the files I think are currently set to windows line endings and I believe the checks try to enforce unix-style line endings. It should be easy enough to change them over. I think the git client even has an option that can handle it automatically.

It's not leading to the failures but I also notice the examples still have a large section of comment which came from when I originally wrote the first one. That portion could be removed.

I am working on making some tweaks to the files and then I'll share the results back with you.

Also good news, the PR to add the TabLayout itself to the library has been merged now so there is less of a need to bring in the branch from the foamyguy fork now which simplifies some of the commands.

When you have a chance try running this command:

git config core.autocrlf

and let me know what it outputs. I think this will help tell us how we can get the configuration set up so that it will handle the line endings automagically.

I created a branch here: https://github.com/FoamyGuy/Adafruit_CircuitPython_DisplayIO_Layout/tree/advanced_tablayout_example which has your examples in it with some modifications made to them. Mostly cleaning up the big chunk of commented / unused code at the begining. I also tried to fix the line endings and seem to have gotten it resolved on my end but we may still need to update a configuration on yours to make it automatic.

There are still some more errors from pylint but the ones remaining are pylint style rules like lines too long and other things. It'll require some cleanup in the code to get all of them knocked out.

Might be worth starting with 1 example only for the first PR and working through them in one file and then once one is completely passing it'll be easier to re-use the passing code in the other examples.

[winter sierra]

@pallid thicket Thank you for making that branch and doing the editing and testing. Excellent that your PR for the TabLayout branch has been accepted. OK, about the line endings. I'll go to try the git command for autocrlf. Yesterday, the whole day I spent on my struggle to get the pre-commit runs without error. I failed. The REUSE lint is 'hardheaded'. Yes, I tried to fix that license error for BGimage4.bmp. I created a license file for it (which is mandatory) however REUSE kept reporting that the license failed. I made a copy of what you did with icon files and just renamed it to the name of the .bmp file. Then I thought: maybe it is not good to use the word 'image' in the filename. I changed the filename into 'bgimg4.bmp' also for the license filename. REUSE continued with the same error. I don't know what I did wrong.

@pallid thicket then the pylint linter keeps on reporting python style errors that I already repaired. I tried to let the 'system' forget what it reported b4 by issuing 'pre-commit clean', however this seems not to work. The same errors (which aren't present anymore) were reported. How come? I don't understand. It is very frustrating. There was a moment that I thought: I quit this! I don't want to spent my time in useless battles like this! This is fighting windmills.

@pallid thicket I am happy to see that your new branch contains the latest version of my five hotplug sensor examples, as you wrote, modified (certain long comments). I see also that you put my ID in the copyright notice at the top. Thank you, but 90 pct is your work! Can't it be changed in both our nicknames/IDs? That would better represent the reality.

@pallid thicket about a first PR. To do that, do we have to take out four out of the five examples? Is that what you mean?

@pallid thicket In the course of getting rid of the pylint linting errors I modified the functions get_temp() and get_dt() a little. Creating a RetVal variable and changed the algorithm to have only one 'exit', at the end of these two functions. I see that the examples in your new branch have these mods present.

@pallid thicket I tweaked also the file displayio_layout_tab_layout_hotplug_temp_datetime_fm_NTP_and_touch_test.py. Instead of, inside get_dt(), setting the flag to update datetime from an NTP pool server at the moment the hour changes, I moved the flag setting inside the loop in main() and let the flag setting happen at a modulus value of the loop counter. It appeared to me that the call for the datetime re-sync did not happen correctly. Now it does, but more frequent than once an hour.

[winter sierra]

@pallid thicket in WSL2 Ubuntu session the git config core.autocrlf results in ERROR even as 'sudo'.

@pallid thicket in MS Win11Pro pwsh session as admin the git command results in 'true'

[winter sierra]

@pallid thicket latest pre-commit run trial on my new forked repo linked to Adafruit's base and branch 'hotplug_sensor_examples'.

[winter sierra]

@pallid thicket just fetch your updated branches

[winter sierra]

@pallid thicket after issuing 'fgit fetch foamyguy' your latest branches were received. Next I issued: a) 'git add .pre-commit-config.yaml'; b) 'pre-commit run all-files'. The result is in the attached textfile. At least REUSE had a 'Passed'. But tons of pylint linter 'Unexpected line ending format' errors.

[pallid thicket]

@winter sierra I haven't caught up fully on all of your messages and i'm heading to a meeting for work shortly so it'll be a while before I can check in again. A few reponses to some of what you mentioned though:

We can definitely have both of us in the copyright header. So putting both of us on there for these ones sounds good to me.

Yes take out 4 of the 5 examples to start with is what I mean. That way there is only one file to focus on fixing the pylint issues on. Then afterwards the fixes can be propogated to the other examples which are all fairly similar enough that it should be easier once the first one is knocked out I think.

Totally understand about pylint and the other checks being frustrating. It's reporting more errors than I had realized until I sat down last night and got it going on my end. It's 100% okay to walk away from it for a while or stop entirely if it's becoming a drag. Having fun and enjoying what you're working on are very high priority in my mind. If you're okay with it I can also work through the pylint issues in the code and share the result back with you and we can make the PR from there. Beyond just being happy to help out folks wanting to contribute, I also think it could be beneficial to have a video showing the process of working through those pylint errors so I could record that while working on it.

Odd about the autocrlf error in ubunty. I'm not sure what could cause that to error but I'll try looking into a bit when I can.

On windows I believe true should be the correct value for that config, so I'm also not sure how the line endings are getting changed around. What program are you using to edit the files? Mu or another text editor?

I will be back later on and will catch up with the rest of your messages and respond.

[winter sierra]

@pallid thicket thank you for all your responds. I am now working on the one with the touch. Trying to get pylint quiet. I am already at a score of 9.41. To get rid of too long lines (especially in the 'global ...' I renamed page_1_lbl to pge1_lbl and so on. I use several editors. Real editing I prefer in Notepad++ and in VSCode. At least if I have to work in 5 files at a time, because these five examples, as you know, have a lot in common.

[winter sierra]

@pallid thicket Going to take a break now. Created a new subfolder in Dropbox with the whole project (from your latest) with mods I did in the five examples. (Dropbox\Shared_with_FoamyGuy\2022-05-11_18h40utc_version). In there also some files with output from pre-commit results and pylint results

[winter sierra]

@pallid thicket. I have edited all the five hotplug examples until there were only a few pylint warnings and result scrores of 9.6.. I suppressed them by inserting in each file, just above the first import statement a '# pylint: disable-all' line, resulting in a pre-commit without errors. The only final bottleneck is that REUSE, again, doesn't accept the license files I put in the /images subfolder. Consequence: I cannot commit. I put copies of the most recent versions to the shared Dropbox folder. I made an error. At a certain moment I copied the five example scripts to the older Dropbox folder. I also save in the Dropbox folder screenshots of the pre-commit results. ---- It's bedtime here. Good night!

[pallid thicket]

Good night 👋 I'll take a look in the latest version and see if I can get the license check happy.

[pallid thicket]

@winter sierra I checked in your latest versions that that branch that I created. It's passing the checks there now 🎉 https://github.com/FoamyGuy/Adafruit_CircuitPython_DisplayIO_Layout/actions/runs/2310754096

You got a lot of the pylint issues knocked out. Great job! I'll make a pass through them tomorrow to see if I can come up with ways to resolve the remaining ones without the general disable.

Reuse does pass on this version locally for me, as well as in the github action. It's still showing an error for you due to the image license?

[winter sierra]

@pallid thicket Thank you for the flowers! About your last version of that branch where the comment says "some cleanup and try to get unix endings", I want to say this. I found this (I thought I wrote to you) https://ownecho.wordpress.com/2014/06/19/git-a-nightmare-on-line-endings/#:~:text=%24 git rm --cached -r. %23 Remove everything,and this demon should never return to you. (see the label below). I implemented the commands shown there and now my pre-commit at a 'first run' comes up telling me that it failed because of line-endings. Then I issue the same run command a second time and voilá, it received a pass on the line endings. That's was a great relief. It not always happens. Various times it get a 'passed' the first time (sure when I did not add lines).

Yesterday night, in the last half hour b4 going to bed, Reuse was not accepting both the .bmp files I had in /images: BGimage4.bmp and bgimg4.bmp. Both had the .license file edited conform the rules of Reuse. Nevertheless it refused. I don't know why, Somewhere, during the course of the day, yesterday, I was asked to fill in my Git ID and e-mail address (in a Git config file or so) I did. I don't think that interfered with Reuse because the ID is the same as in the .license file.

[winter sierra]

@pallid thicket fetched your latest

@pallid thicket next, using the GitHub Desktop app I tried to do a commit. It failed upon the Reuse error on just one .bmp file ! (And I think nothing is wrong with that file!)

@pallid thicket 2nd commit trial, almost immediately after the one shown above

[winter sierra]

@pallid thicket Notepad++ shows which mode of line-endings a file has. Conclusion: Reuse marks this file as not conform their rules. I don't grab it.

[fickle torrent]

I was going through the github tutorial and got stuck after I made a 'commit'. It wouldn't let me approve it because I was the author of it, and I couldn't figure out how to take the next step in the tutorial. it seemed like it wanted me to do something that it also wasn't letting me do (approve the commit). can anyone help me get past this hump?

[fierce field]

you made a pull request, or a plain commit?

[fickle torrent]

not 100% surer - i made a pull request for sure

[fierce field]

can you point to it?

[fickle torrent]

i guess it's a commit 'request'? it was a tutorial and then it went crazy

https://github.com/landru428/github-slideshow/pull/3

can you see this link?

this is where i got stuck

[fickle torrent]

I guess that wasn't helpful

[queen mango]

@fickle torrent I can't see the link. Is it possible that repo is private? Does it load on your end

[fickle torrent]

thanks for taking a look - the repository says it's public, if you tried to access it after midnight last night, it was deleted. I'm going to just start fresh and see if i can get a little farther the second time around.

[queen mango]

ah I see

What tutorial are you using?

[fickle torrent]

the one provided by git

i got stuck after following the instructions to create a pull request

after that, i couldn't approve the request because I was the creator, and there were not other relevant instructions.

[queen mango]

Hmm odd

Yeah that sounds like an issue with the perms in the github repository you made

If you get to that point again I'd be happy to help you figure that out

[fickle torrent]

thanks for taking a look!

[coral sable]

wow this is very fancy here. I need some help setting up SSH with github on a raspberry pi. I followed the instructions on https://jdblischak.github.io/2014-09-18-chicago/novice/git/05-sshkeys.html to generate an SSH key and saved it on github as instructed. I find that I still need to enter my username and password to use git from my terminal i.e. I was expecting I won't have to enter my username and password (for github) to push/pull/commit.

I see you were in the same boat as me several months ago. I can't follow the conversation with @waxen hill very well. Can someone please handhold me while I set up this SSH stuff on my pi :(?

[fierce field]

I vastly prefer https to ssh for this, though you may want ssh. See https://docs.github.com/en/get-started/getting-started-with-git/caching-your-github-credentials-in-git for caching github credentials on Linux (i.e. RPi), using the relatively new gh command-line tool, which is available for RPi: https://github.com/cli/cli#installation, https://github.com/cli/cli/blob/trunk/docs/install_linux.md

[full steeple]

I'm afraid I don't have any more experience than you see there. IT policies prevent me from SSHing anywhere!

[cosmic saddle]

Are you sure it's the GitHub password and not the password to decrypt your private key?

[coral sable]

Yes, it's the github password. I don't have a passphrase for decrypting the private key

Unsafe, I'm sure, but I need to learn to do this first before I lock myself out with yet another passphrase

[coral sable]

So I got this far https://docs.github.com/en/authentication/connecting-to-github-with-ssh/testing-your-ssh-connection I get ```> Hi username! You've successfully authenticated, but GitHub does not

provide shell access.but I still have to enter username and password when I do something likegit fetch origin main```

Ok, wonderful! I had to update the remote URL to the SSH URL under the "Clone" button on the github website. https://docs.github.com/en/get-started/getting-started-with-git/managing-remote-repositories#switching-remote-urls-from-https-to-ssh

I'm so happy that it's all set up. I use VSCode on my desktop PC to SSH into the RPi and now I've set up the ssh key for github so everything is smooth going.

I wish all of these instructions were listed together though.

[cosmic saddle]

Glad to hear it's figured out! My next thought was going to be adding -v to SSH commands as sometimes they provide more info

[coral sable]

One day, when this ssh stuff is not working for me anymore, i.e. my anxiety level about finding a better solution reaches its threshold, I'm going to move to this (https instead of ssh) better way of doing things. One day.

[fierce field]

take it slow, as you want; the gh stuff is pretty recent, and makes it a lot easier to save the credentials. Before that I had to compile a helper program that came with git: kind of ridiculous.

[honest shuttle]

Hello... I'm pretty new to GitHub but I do have some code on it... Which brings me to my question. I used the Pyportal User Interface code as my platform and changed and added some stuff. Is there a way to link my code to the original git page? Thanks!

[fierce field]

you can fork the repo where the code came from, and make your own changes in that repo. If there are changes "upstream", you can then merge them in.

[honest shuttle]

So, basically there would be a duplicate of the code... The one on my page and the fork? Just trying to wrap my head around it.

[fierce field]

I'm not sure I understand what you mean about linking your code to the original git page. The PyPortal has its own copy of the library files. Or are you talking about fetching webpages at run time.

Are you talking about this file? https://github.com/aioue/Adafruit_Learning_System_Guides/blob/main/PyPortal_User_Interface/code.py

[honest shuttle]

That's the page I'm talking about... I just want people to have an easier time finding the changes I made so they may benefit from what I've learned and maybe they have a similar project.

[fierce field]

You could copy the code and create your own repo, or you could fork ours, but that's pretty big, and 99% of it is irrelevant. If you think your changes are generally useful, feel free to submit a pull request, to change the original code, or to add a new example.

[honest shuttle]

Thanks @fierce field !

[acoustic atlas]

I'm new in using GitHub I want to know why I can't upload large files like 400MB or more?

[dusk jungle]

Mainly because it's a free service and that's a terrible use for source-control tools.

[acoustic atlas]

oh, ok so is it monthly membership?

@dusk jungle

[dusk jungle]

I presume that one of the paid plans would have higher upload limits, but I haven't checked.

(But again, storing large binary files in git version control is an atypical use case.)

[acoustic atlas]

@dusk jungle are you a freelancer?

[dusk jungle]

Indeed I am... part-time embedded development. My main contract at the moment is for firmware to run on satellite thrusters.

[acoustic atlas]

Ok, would you mind if you give me some advice, I just opened a freelancer account to see how everything works sometimes I find some simple jobs that I could do but not get excepted, so can I do some free work for getting ratings or I do what?

I mean maybe if they rate my work I will be noticed next time 😂

[dusk jungle]

I don't have much experience with freelancing platforms that way, I'm afraid, as I've generally gotten my contracts through networking instead. I have hired people myself on platforms, though, and I tended to pay much closer attention to their portfolio of projects rather than their ratings in making a choice. But I don't know how common that trend might be, just something to consider.

[waxen hill]

What’s the reason you want to upload binary data?

In many cases the “releases” feature is the right solution for binary blobs

[acoustic atlas]

I want to keep my unity projects there

[waxen hill]

I’m not an expert with unity

What exact binary data are those big files?

Git will be fine for your source code

[acoustic atlas]

Not only codes it also some games I made so I can send them to my friend

Anyway I sent it to him by email

[waxen hill]

You want to put the build description under revision control, but not the build artifacts

[acoustic atlas]

What is that? I don't know

[waxen hill]

How do you compile your games?

[acoustic atlas]

I'm using c# and unity to make games

I guess the project will be sized about 48MB if not building it and take all the files but still says GitHub takes 28MB maximum storage

[quick tusk]

This sounds like a good fit for GitHub "Releases", which are just git tags that have extra github-specific attachments, such as binaries. You for also look at Git LFS, a corporate-backed open source extension to git that allows big files to be stored outside the git tree and referred to from within the repo.

[acoustic atlas]

Thanks, I will try this later!🎉

[golden shale]

Is there a way to merge specific lines of code from feature branch to master?

[cloud hound]

you can merge specific commits, but not lines

[golden shale]

That works too

[cloud hound]

the command is called "cherry-pick" and you specify the hash of the commit to merge

[onyx night]

if you do want specific lines from a commit you can rebase-edit that commit, reset to head~, then commit specific hunks into distinct commits, and continue the rebase
and then cherry-pick (or rebase!) from that

[still iron]

Made my first ever pull request today. 🎉

[sturdy talon]

Congrats!

[still iron]

I wasn't sure if I should paste Dave Astels copyright in there or mine. I mean he wrote the library I only wrote an example script.

[sturdy talon]

If you wrote the example, it's yours. The library will have Astels' name on it, but the script can have yours. Two separate pieces of software.

[still iron]

ok thank you. i think i created a 2nd commit to update the header but couldn't create a pull request on it, perhaps because the 1st pull request hasn't been approved and file doesn't actually exist yet?

[sturdy talon]

So, you want to continue working on the same branch. If you push to the same branch on GitHub that is active in a PR, it updates the existing PR with your new commit. That is how you should be doing it.

The file exists in the PR until it's merged. So you need to work within the PR until such point it is part of the main branch on the repository, in which case, you would start fresh and make a new PR for changes.

[still iron]

I used tortoisesvn like a decade ago. This is a bit different. Will need time to figure out how it works. I think I'm up and running though. 🙂

[sturdy talon]

It's often harder to learn something new when you have a wee bit of ancient knowledge to contend with, versus learning it entirely fresh.

[sturdy talon]

Folks are around to help here. So post questions as you have them!

[jaunty pulsar]

Can confirm

[lone cradle]

I have two CP PRs, mine for the Wemos C3 Board, and todbot's PR to add wifi.radio.tx_power. Can I rebase my PR to be on top of his PR so I can make a combined test build? Rebase might not be the right word...

[wide iris]

only for testing?

[lone cradle]

yeah, then I can say his PR is good and mine is as well.

My PR will require his for WIFI to work on that board.

[wide iris]

If it is a temporary use case you could create a new local branch from one and merge the other into it for testing -- then delete the combined branch when done.

For temporary work tend to make a new branch to make sure I can't accidentally push it.

[lone cradle]

I just thought of a kicker... his branch would be in his fork...

I could gh pr checkout PR# and then copy in my changes and not commit them...

[wide iris]

you could add his fork as a additional remote repo

[lone cradle]

I don't think I could push to it anyway if I did commit.

Thanks for the ideas @wide iris, I have paths to try now

Even better... it is already merged !

[still iron]

they tried to push my first pr today but it failed because i didn't have a .license with images and didn't use pre-commit. when attempting to run pre-commit from my github directory i get errors (pre-commit.log)

[still iron]

i'm using git for windows program. is manually running pre-commit not compatible with it?

[still iron]

looks like it's failing to install black.

[thorn star]

Based on your log, you're trying to install pre-commit in a non-git repository folder.

If you use pre-commit install you have to be in the repository folder so it can hook into git appropriately

[still iron]

Defaulting to user installation because normal site-packages is not writeable
Requirement already satisfied: pre-commit in c:\users\devon\appdata\roaming\python\python310\site-packages (2.19.0)
Requirement already satisfied: identify>=1.0.0 in c:\users\devon\appdata\roaming\python\python310\site-packages (from pre-commit) (2.5.1)
Requirement already satisfied: virtualenv>=20.0.8 in c:\users\devon\appdata\roaming\python\python310\site-packages (from pre-commit) (20.15.1)
Requirement already satisfied: nodeenv>=0.11.1 in c:\users\devon\appdata\roaming\python\python310\site-packages (from pre-commit) (1.7.0)
Requirement already satisfied: toml in c:\users\devon\appdata\roaming\python\python310\site-packages (from pre-commit) (0.10.2)
Requirement already satisfied: pyyaml>=5.1 in c:\users\devon\appdata\roaming\python\python310\site-packages (from pre-commit) (6.0)
Requirement already satisfied: cfgv>=2.0.0 in c:\users\devon\appdata\roaming\python\python310\site-packages (from pre-commit) (3.3.1)
Requirement already satisfied: setuptools in c:\program files\python310\lib\site-packages (from nodeenv>=0.11.1->pre-commit) (58.1.0)
Requirement already satisfied: filelock<4,>=3.2 in c:\users\devon\appdata\roaming\python\python310\site-packages (from virtualenv>=20.0.8->pre-commit) (3.7.1)
Requirement already satisfied: six<2,>=1.9.0 in c:\program files\python310\lib\site-packages (from virtualenv>=20.0.8->pre-commit) (1.16.0)
Requirement already satisfied: platformdirs<3,>=2 in c:\users\devon\appdata\roaming\python\python310\site-packages (from virtualenv>=20.0.8->pre-commit) (2.5.2)
Requirement already satisfied: distlib<1,>=0.3.1 in c:\users\devon\appdata\roaming\python\python310\site-packages (from virtualenv>=20.0.8->pre-commit) (0.3.4)

C:\Users\Devon\Documents\GitHub\Adafruit_CircuitPython_BitmapSaver>```

did it on adafruit_circuitpython_bitmapsaver, seems to work ok? still fails to commit though.

[thorn star]

That's because that's the git repository so it installs it as a git commit hook. Basically it will run pre-commit when you commit changes and reject/change them if they fail. If you'd like to just run pre-commit to check and modify the code, you can run pre-commit run --all-files from the command line in the repository.

[thorn star]

Also, it does look like you pip installed pre-commit, so now you can either use pre-commit install to install pre-commit as a hook as you commit changes, or just run pre-commit run --all-files to simply run it on the existing code in the repository. I personally do the latter until it passes all the checks. That will also solve the same issues for the GitHub Actions CI.

[still iron]

An error has occurred: FatalError: git failed. Is it installed, and are you in a Git repository directory?
Check the log at C:\Users\Devon\.cache\pre-commit\pre-commit.log```

same error if i do pre-commit install from that directory

[thorn star]

Strange, what happens if you type git log?

[thorn star]

Actually my guess based on your filepath is that you're using the GirHub desktop client right? You may need to also install Git for Windows, depending on what that git log says. I can't remember how it handles installing git and whether the rest of the system can access it.

[still iron]

yes i'm using github on desktop that's what i meant, but it's not working right with pre-commit and black or something.

'git' is not recognized as an internal or external command,
operable program or batch file.```

[still iron]

making progress. path to git.exe wasn't specified and it was buried deep: C:\Users\Devon\AppData\Local\GitHubDesktop\app-3.0.3\resources\app\git\cmd now git log works and i hope a lot of other stuff too.

[still iron]

aha! after adding the correct github desktop git.exe path in environment variables (it was incorrect) i ran pre-commit run --all-files again, edited 1 file to remove a single white space, made a commit, and it just went through. made me work for that green check mark sheesh.

[thorn star]

🥳

[lone cradle]

This always happens and I'm not sure I'm not doing something wrong... I use OhMyZsh and the Git plugin, but I don't think is the issue.

On my fork of CP, when I want to make a new branch I use
gcb new-branch-name upstream/main
Then I do a pull gl
And there are always updates... which seems right, I never update my own main...
And then status gst

gst
On branch new-branch-name
Your branch is up to date with 'upstream/main'.

Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git restore <file>..." to discard changes in working directory)
  (commit or discard the untracked or modified content in submodules)
    modified:   extmod/ulab (new commits)
    modified:   frozen/circuitpython_picoed (new commits)
    modified:   frozen/pew-pewpew-lcd (new commits)
    modified:   ports/broadcom/firmware (modified content)

Untracked files:
  (use "git add <file>..." to include in what will be committed)
    .python-version

no changes added to commit (use "git add" and/or "git commit -a")

that .python-version makes sense to me, that's my stuff. But why do I have those other unstaged changes... am I doing something wrong? I usually just ignore them when I make my own commits to the branch prior to any PR.

Thanks!

[lone cradle]

make fetch-submodules cleaned all that up but the ports/broadcom/firmware line

[small hemlock]

are those all submodules? from the names they all sound like some sort of external dependencies. those can be tricky to make work seamlessly, my guess is that this repo is setup a bit weird because of that

if my assumption is true, ignoring them shouldn't hurt anything short term except make your output noisier, which might make you make a mistake

[lone cradle]

Yeah, I think they’re all submodules.

[small hemlock]

submodules are a bit weird. every time I've tried to use them I've abandoned them eventually.

but generally when you're operating in the parent project you can see statuses of your submodules, but things up update won't (by default) do anything to them

so them being out of date kinda makes sense

you can do something like git submodule foreach git status to see more details about the differences in each

[lone cradle]

Thanks for your insight @small hemlock … seems like my strategy of just focusing on my changes is the right way to keep going.

And knowing what to focus on is super helpful.

[night reef]

@lone cradle are you on mac? I think that the broadcom firmware has filenames that need to be on a case sensitive file system 🙄

[lone cradle]

I am on a mac...

@night reef That's a good hint... tx!

[silent fog]

Is there a way for git to ^not^ convert the line endings for a single file or directory in a repo?

The situation I have is that there's a csv file in a repo that requires Windows style line endings (long story). However, by making git use Windows style line endings for the whole repo, it results in entire-repo changing commits.

The only solution I found is to use .gitattributes and to define the file or directory as a binary. But this feels kludgy and prevents using git diff. Is there a better option?

[still iron]

Added a PR for a 16MB cypress flash chip in nvm.toml. It's needed for a custom board/build. Any idea how long that kind of review might take?

[fierce field]

Is it ready 🙂 I saw several commits.

it's missing an ending newline

[still iron]

is that required? i intentionally removed it because i thought having a blank line was a bad thing.

yes i guessed at the total_size hex the first time and then found some 16MB examples so i'm more confident about the last commit.

after finding the 16mb examples it filled in other blanks i had for capacity and type. looked up many examples. confidence is high it's ready to go.

[fierce field]

Have you tested it already? That's when we would normally approve it.

I think you see a blank line if you are using the GitHub web editor. There would not be a blank line if you submitted it from your own computer.

[still iron]

thought you said first the chip has to be supported. when i try to build it fails because the definition for chip isn't there.

i use github windows desktop app.

[fierce field]

but you can test it locally on your own build. Just edit it locally in the circuitpython tree and try that build.

When you've tested it, then you can submit the nvm.toml PR and then the circuitpython PR

[still iron]

something is wrong then. i already added it to my local nvm.toml and the build still fails

~/circuitpython/ports/nrf$ make BOARD=feather_bluefruit_sense_16N
Use make V=1, make V=2 or set BUILD_VERBOSE similarly in your environment to increase build verbosity.
<command-line>: error: 'S25FL128L' undeclared here (not in a function); did you mean 'S25FL216K'?
<command-line>: note: in definition of macro 'EXTERNAL_FLASH_DEVICES'
make: *** [../../py/mkrules.mk:62: build-feather_bluefruit_sense_16N/supervisor/shared/external_flash/external_flash.o] Error 1```

[fierce field]

hmm. yes, I just meant that nvm.toml is a separate repo and needs its own PR when you're ready to submit everything

[still iron]

yes i have it as a separate repo.

but also within the circuitpython repo

[fierce field]

did you do a clean build?

[still iron]

running linux as windows subsystem with visual studio code for building circuit python, then i have github desktop on windows for individual repo contributions, basically running 2 gits.

unsure how to do that, was following the build circuit python learn guide

[fierce field]

so you made the change in nvm.toml in WSL?

or are you referring to the same actual repo clone in Windows and Linux?

make BOARD=... clean

then make BOARD=...

in WSL, remember to do make -j<n>, where n is the number of cores available, to speed up the builds

[still iron]

how long does it usually take to build?

looks stuck after qstr updated

ah that did it!

i have a uf2 🙂

my first custom firmware, gonna take a screenshot and hang it on my wall

now to break out my soldering iron and pray to the computer gods

[fierce field]

circle back in #circuitpython-dev for more 🙂

[still iron]

whoops, forgot where i was.

[fierce field]

me too