my "keep" script

[Skaperen]

i have, indeed, had what i thought was a problem, tried to solve, came up with a solution, and as a result, had a worse problem, then abandoned that "solution" to reduce back to the original problem.

---

now, what would be useful would be a tool that keeps files i edit in a local copy and in a version control repository in a way where i won't need to do anything more, along with a grep tool that can grep the entire history of that file.  even better if such a tool can support GIT, HG, and SVN.

[wavic]

My point of view on this is that the time never comes back. Concerning the codding, when I don't have a progress on a task for a half an hour, I start searching the internet. I remember how I wrote a function to count the appearance of each letter in a text.  Don't remember how long it took me but remember that I have learned a lot about the defaultdict. It was at the very beginning of my codding experience. Many weeks after that I have discovered collections.Counter. I am not telling that I have lost my time. But all of our knowledge is on top of the efforts of the people before us. We have to appreciate these efforts and need to value our time more. Doing that we will have more time to do something else which can be used by someone else in turn. The purpose of this forum is the same as well SO and many others.

[Larz60+]

Sounds a lot like what my IDE will do. It can handle git, CVS, mecurial, subversion, CVS and more.
I had used it with git for quite some time, but just integrated GitHub into the mix. It was incredibally
easy to set up. The version control has a one click commit, and a few more will push all changes to GitHub as well.

Don't know if this will fit your requirements or not.

[metulburr]

now, what would be useful would be a tool that keeps files i edit in a local copy and in a version control repository in a way where i won't need to do anything more, along with a grep tool that can grep the entire history of that file.

I dont even integrate my IDE with git. I just write code in any IDE usually either Geany or Vim. and then when i leave, i push the code to the repo via terminal commands.

git add .
git commit -m "changes made were XYZ"
git push origin master

Then when you feel comfortable with that you cna learn branches , etc.

But yes there are IDE's that integrate this into the IDE so you push a button for pushing to the repo etc. But what fun is that?

[snippsat]

now, what would be useful would be a tool that keeps files i edit in a local copy and in a version control repository in a way where i won't need to do anything more, along with a grep tool that can grep the entire history of that file.  even better if such a tool can support GIT, HG, and SVN.

This is exactly what GitHub,Bitbucket does,that is powered bye Git.

The way you share your keep.py now is a uncommon way(you have or own way for most thing),
and i guess very few or no one will use it.

There is a standard to share stuff in Python,
is usually GitHub,BitBucket Repo in combination with PyPi for medium to large project, 
smaller project can just be PyPi alone or just share in Forum like this for small scripts. 

A good example(TextBlob) how this can be done for a medium size package:
TextBlob doc, TextBlob GitHub Repo and TextBlob PyPi Repo that also has wheel.

[metulburr]

TextBlob doc, TextBlob GitHub Repo and TextBlob PyPi Repo

This is how people share code.

People expect there to be a readthedocs to understand how to use the code, prerequisites, installation, introductions, API usage, and examples. This is meant for when the github readme is too small and the docs are too complicated for that.

Github or Bitbucket for housing the code and git and hg for pushing and pulling the code to the repo. And the readme is also there pointing to readthedocs, or a version of it. People can fork it and make their own version and even upload it back to yours if you allow it. Multiple people can work on the same project and create branches of the code, merging them in when complete, etc. They can keep track of problems with "issues". It really unified how everyone disperse code to each other. Or you can forget all that and just upload to your own repo. It is safe from harm and not local, while also being local to any computer you pull it to. You dont have to keep them local and you dont have to worry about housing your own code, or things like hard drive failures.

Pypi is basically index of all python packages. pip is the method of installing from pypi. pip would be to pypi what git and hg are to github and bitbucket in some way. Sometimes you find packages in github that are not in pypi, but usually if they get enough support, they eventually put it in pypi. A tool of python should be in pypi.

[Skaperen]

i do a lot of things the old way because the old way gives me more control over things.  the old way to distribute a file was running you own server.  even before that it was FTP.  even before that it was ... you don't want to know.

as to where to host a free file, in theory i could use github.  but i am concerned about their motives since their signup form is trying to scrape passwords to other sites.  i have not seen that happen to download stuff, yet, but if it ever does then i'll quit downloading from their, too.

also, for any hosting site, it must be possible to script an upload, so all i need to do is run that script for a new file.  if they are offering standard interface repository, then a script around that is likely good enough.  but i must be able to do everything, after signing up, using the standard revision control tools (configure with password, etc).

one limitation for me: nothing run by google or microsoft.

[buran]

but i am concerned about their motives since their signup form is trying to scrape passwords to other sites.

Do you care to elaborate further on this? And do you really think something like this will go unnoticed by hundreds of thousands software developers using GitHub?

[metulburr]

one limitation for me: nothing run by google or microsoft.

I guess i can understand this if you have nefarious code. Some of my code i would not like some corporations to see i would not put in any cloud.

But for all else, yes its public domain.

[Skaperen]

their signup form grabbed a password to another site and put it in the new password field, perhaps hoping i would just go ahead and use it.  smart people would highlight it and type in another by default.  this might be another firefox javascript bug.  i have no idea how their code got it.  i'm not good enough with javascript to read it and see what does.  i could not see which password it was since it was displayed with all asterisks.  but it had the length of 12 i use on web sites.  the user name field had "skaperen in it.  i tried it again on another user (i have a couple dozen /etc/passwd users on my laptop that are for me) that had no passwords saved and it did not put one in.

---

i don't write nefarious code.  i just don't want these people tracking me.  one of the many google recruiters that has emailed me made reference to a few C programs i wrote.  sure, what you have done can help show your qualifications.  but he could not say where he found my code.  if they scrape them from other sites where i put them, well so what.  but i don't want to give them extra clues of thin.gs like where i login from and how often i do that.  lately, i have been using reddit.  i have no idea what they are doing, but theirs do seem to be focusing a bit on my interests.