Now that we're getting bigger and sharing more code on this site, I'd like to find something better than putting random files in discussion responses. Has anyone had enough experience with the various code repositories out there to recommend one that would work for our purposes? Most of the ones I've seen require you to use gnarly *nix command-line software tracking tools like QVC and git, which are too complicated for our purposes. Anything simple with a web interface that will just let us upload code and track versions, etc?

Views: 30

Reply to This

Replies to This Discussion

SVN ("subversion") is a popular choice. I don't know how complicated it is to host, but it is quite easy to use.

http://subversion.tigris.org/
Like most of the other pro-quality open source repositories, subversion requires you to learn and use a command line program to upload and otherwise manage code on the website. I fear that's asking too much of most amateurs like us. (I found the whole process pretty mystifying myself). Is there anything more basic (and simpler) than that?

-c
I was likewise less-than-enthusiastic about learning a new tool, but someone took the time to walk me through a basic set of function calls and that greatly simplified the process.
What do you think of a super-simple alternative like Snipplr?
My projects are built around "make" and have a bunch of files and sometimes multiple directories (include, docs, code), so I need something more robust. I think you'll find that's common with projects on 32-bit processors.
Do you know a good resource that explains make, and working with multiple directories and files. I have come up short in my searches for a good tutorial for that sort of thing. It is easy to find "hello world" examples and any kind of syntax help, but a nice tutorial on the basics of modular program design, managing multiple files, and make is something I have failed to find.
I don't think I've actually ever created a makefile from scratch, instead finding similar projects and then performing surgery. I likewise have found tutorials on makefiles less than obvious, but I've still been able to use the tool quite successfully.
Why don't we create a project on sourceforge.net. That way the svn, cvs, bug tracking, etc is handled for us.

Neal.
That is certainly an option, but most developers I work with have chosen to set up their own SVN repositories and manage access locally.
But won't everyone who contributes have to learn how to use the CVS client to contribute? It's pretty intimidating for us 8-bit embedded guys who wouldn't know a makefile if it bit us.
It would be a labor of ... love, but a people could simply take their code save it on a particular page, then daily, weekly, whatever a facilitator could zip them into a single file. Versions would be recorded in the post of each zip/rar file and on a viewable table. If you needed an older version of a bit of code, the table would tell you what zip file that version is located in...

I realize that you "code" pros out there are cringing at the very thoughts but this is simple and allows anyone to contribute and allows for everyone to obtain the code they need.

-Brian
cvs really isn't all that hard to learn, and there are windows interfaces for it. Some people find it easier to click a "login" button than to type "login". Find the mouse (probably on the floor somewhere), move it, see where the mouse pointer on the screen moves, navigate the menus to start up the cvs windows interface, navigate to the login button, blah blah blah ... or you could just type "cvs login" on your already open shell. :-)

Let's see ...

cvs login
cvs checkout
cvs add
cvs update
cvs diff

That takes care of about 99.9% of the cvs commands I use. If you 8 bit embedded guys can figure out why your integers roll over after 255, then you ought to be able to figure out cvs!!!

My main concern is that there are a lot of people working on a lot of diverse projects. It really doesn't make sense to dump all that code into a single repository unless someone wants to somehow unify everything and make it all intercompatible ... but given the amount of hardware and platforms and individual goals there are, it really doesn't make sense.

I think the best suggestion I've seen here would be for people to register their individual projects at sourceforge. SF provides a number of services ... code repository, forums, bug tracking, web site, etc. Individual projects can pick and choose which of these services they wish to use and ignore the rest of them.

Someone suggested running your own cvs/svn repository. That is great if you know how to do that and have a server that is on the net 24/7 with a static ip address. (I do this for the FlightGear repository by the way.) I suspect many people here don't have access to that sort of dedicated server though, and that's were sourceforge can be a huge asset.

At the end of the day though, there isn't one right way to do this. Personally I've had good luck with sourceforge. I've had good luck with cvs (although there are many new contenders, each with their own strengths and weaknesses.) If I was starting a new project from scratch, I'd probably be inclined to look at svn. And I would recommend *not* trying to roll multiple developers projects into a single svn/cvs/git repository.

Regards,

Curt.

RSS

Contests

Season Two of the Trust Time Trial (T3) Contest has now begun. The third round was a reliablilty/aerial photography round for both planes and copters, which is now closed. Stay tuned for the next round, beginning soon.

A list of all T3 contests is here

 

© 2012   Created by Chris Anderson.

Badges  |  Report an Issue  |  Terms of Service