Closed G33kDude closed 9 years ago
¯(ツ)/¯ sounds fine to me
On Sun, Feb 15, 2015 at 12:21 PM, Joe DF notifications@github.com wrote:
I lost count. How about 0.0.17.21 ? 0 : not stable/full yet 0 : dido 17 : as in 17 working algorithms 21 : approx. number of "sub"-commits for the 17 algorithms
— Reply to this email directly or view it on GitHub https://github.com/ahkscript/libcrypt.ahk/issues/13#issuecomment-74426544 .
How about moving code that is not fully working yet ((x)tea, ascii85?) to /src-dev/ and only keep properly working funcs in /src/ that way experiments in /src-dev/ don't require a bump in the version number. :question:
@G33kDude hahha, Alright great :+1: @hi5 hmm good point
"version bumb"
Anyways, branches are cheap. Maybe I should work on the installer in a separate branch as well?
Hmmm... I don't know... What do you think @hi5 ?
I don't mind if it is developed in the main branch - I haven't tried it yet but evilC has an installer script which may be of use https://github.com/evilC/AHK-Library-Setup
Although that looks well thought out, the "put a placeholder #Include script in ProgFiles" is something I'd rather avoid for two reasons. The first is that it'd require administrator (which can be alleviated by putting it into the MyDocs folder), and the second is that it'd require the person to keep their copy of the library in the same spot all the time. This isn't an issue for people who clone repos into a central directory and leave it there, but that's not the average AHK user, I think.
That being said, it made me realize one of the advantages of having a template. That is, when there's a template you can just #Include said template and use the changes live as they happen (without first requiring it to be built again). I think that make.ahk might want to also output a template to the build directory, for this purpose.
We should consider opening another issue for further discussion.
I lost count. :P How about
0.0.17.21
? 0 : not stable/full yet 0 : dido 17 : as in 17 working algorithms 21 : approx. number of "sub"-commits for the 17 algorithms