SourceForge?

[Sonnington]

Recently it's come to light that SourceForge is embedding adware into some of the software. A lot of different software devs have jumped ship due to the news and bad publicity the site has been receiving. Are there any plans to move Exult to a different hosting platform like GitHub?

[Dominus]

No
1. we used sf for 1 ½ decades, it seems strange and also kind of as if we owe them
2. I personally dislike github strongly. It makes some good things but following changes of a repository is challenging and not well done. The ticket system has all the shine but less the functionality I'd hope for
3. most importantly and dwarving all other reasons, we have almost no active members anymore. No one to move the code, no one to setup the site elsewhere, no one to setup the forum elsewhere. Calls for help have gone unanswered. It seems getting people involved in open source is being harder these days since the ones most open are getting older and have too many other responsibilities (jobs, families...)

1. and 2. don't matter much. 3. is the culprit. I asked on irc hoping to get answers from at least two other members, but apart from Kirben agreeing that SF sideloading crapware is really bad, no answer.

[galneon]

Hugely embarrassing long-term hosting failure: yet another good reason to leave SourceForge.

[Dominus]

Yes, unfortunately... 3.

[alagner (not logged in)]

Dom,

one more question regarding github/bitbucket:
would it be ok if I forked Exult and put it to my own, freely downloadable repository on, say, github?

If not, is there any other way I can properly and legally fork Exult without violating any copyrights?

[Dominus]

Sure, you can easily fork it and legally so. There is already a clone and the iOS fork on github.
You can save yourself the svn2git by just cloning/forking those

[alagner (not logged in)]

Forked, https://github.com/alagner/Exult/tree/master/exult.

BTW - is shared_ptr brach in development more or less or did it stop completely (the 2nd one judging from commit history)? I'm thinking of migrating the code to the new cpp standard, so it might be easier to maitain

[Dominus]

That branch never got anywhere. It was meant as a playground for Marzo to fix the big disappearing objects bug.
But that one was fixed by Jeff directly eventually...

[Frob]

Honestly, I think that sourceforge might be a reason that no one is interested anymore. A move to GitHub could spur involvement.

[Dominus]

Honestly, I think you are wrong...

[Colourless]

Yeah, don't see why that would affect anything

[DeplicatorForgotHisPassword]

I don't disagree with the reasons to stay with SourceForge, but to play devil's advocate...

I'm about to finish my CS degree and the only revision control I've use in school and in internships has been git. Every professor and speaker that talks to us says get a github account to showcase your work. It would seem new developers are only getting experience in git, so the notion moving to GitHub would bring in new people is not ridiculous.

I would jump into development no matter where it's hosted but school and family and other projects have to come first right now. Maybe someday

[Dominus]

I suspect that people that are put off by the version control are likely to not make any other effort and abandon it on the first problem or disagreement.
If it's just about how to get the code, there ARE mirrors on github and they've seen near zero axtion in the last two or three years (except for one bug report).
So no, I highly doubt to git or not to git makes a difference.

[Arr]

"I suspect that people that are put off by the version control are likely to not make any other effort and abandon it on the first problem or disagreement"

But most contributions on github come from sporadic contributors. Is not that people would abandon it, but that most will be able to constribute a little.

[Dominus]

ok, how about this: there ARE inofficial mirrors on github, at least two and the iOS fork (which apart from the iOS specific stuff is still very much working for other systems).
Whenever there is a significant contribution happening on these we will probably rethink our stance on moving to github. So far there has been ZERO interest.
What I'm aiming at is that I suspect this is a very theoretical discussion as the many people contributing on github is IMO not gonna happen.

[Arr]

"there ARE inofficial mirrors on github"

It is unlikely someone would put the time to contribute in inofficial mirrors. Who is the guy behind them? will that person review a pr ? Which mirror should i contribute?

Another thing that strikes me, i searched over one of them (https://github.com/alagner/Exult) the term "movement" (i wanted to know how the game handles entity collision) and got this https://github.com/alagner/Exult/search ... q=movement:
- exult/tags/Release0_93Alpha4/server/objserial.h
- exult/tags/Release0_94Alpha5/server/objserial.h
- exult/tags/Release0_98RC1/server/objserial.h
- ...

wut.

[Dominus]

As unlikely as...

What is your issue with the mirror by Alagner? That's just the tags we have in our SVN. When you mirror our SVN on GIT, the tags are being kept.

[Arr]

It makes searching stuff over github harder.

But well, you made your mind that migrating to github is not worth the effort. I think that, if you changed your mind, the worst thing that could happen is nothing. On the other hand, you could open a door for new constributions.

But is up to you, new york.

[Dominus]

Well, that's just what happens with a faithful conversion from SVN to GIT.

I actually haven't decided that moving to GIT isn't worthit, I'm just convinced that it won't change anything.
I *am* working on the iOS port on Github. It's been there for almost a year. It got 12 people watching, no forks, no contribution other than Litchie's and my code...

[Arr]

Have you thought on writing about the development of exult? Programmers like that kind of story and they easily reach the reddit frontpage, which in exchange can make someone interested in the project. Exult is an unique project, i for one would be interested in your thoughts on the challenge on making it and surviving over the years.

[Dominus]

End of January SF/Slashdot got a new owner. In their first announcement they did away with the dev share (adware crap) program.
Let's see what happens next

http://sourceforge.net/blog/sourceforge ... ture-plans

[Corpsecrank]

For quite a while firefox was blocking sourceforge pages from loading and warning people that it could be malicious. Recently I see that this has stopped probably no most assuredly because of the ownership change and doing away with the issues.

I do feel like a lot of damage has been done to the SF reputation though and that is sad since a lot of great projects like this one and pentagram are hosted on it.

I think moving the project to git is maybe the way to go for more reasons than interest alone. I feel like git is the platform of choice while SF is becoming a relic of the past. Because of that I think moving to git is a choice that makes sense if the project should survive long term.

Beyond the debate of preferred platform I think interest can only be achieved through exposure. Ultima games in general seem to be slowly slipping away these days and there is less and less happening in the already very small community every year. The best we could hope for is getting what few people are interested in ultima these days to know this project is here wherever here ends up being.

[Dominus]

Git = version control system
SF = platform, offering many different things, including the version control systems SVN and Git
You can't compare Git and SF.

[Dominus]

so... we switched to Git, the code is on github https://github.com/exult. Everything else remains on SF

I'm looking forward to the many people that will add code, now that we are using Git...