There once was an email on this mailing list from David Chisnall where he told someone that they were getting too emotional and should step away from the computer to walk around outside for a bit. Boy, we really need some of that British humor at a time like this!
The website is a huge blocker for attracting new developers. We must first admit we have a problem before we can begin to solve it. On Wed, May 14, 2025 at 06:54:04PM +0200, Riccardo Mottola wrote: > Steven wrote: > > > > Also as part of this exercise, I tried to "survey the current > > landscape", looking at similar projects, other open source efforts. > > This led me to > > > > developer.gnome.org <http://developer.gnome.org> > > developer.apple.com <http://developer.apple.com> > > developer.redhat.com <http://developer.redhat.com> > > developer.ibm.com <http://developer.ibm.com> > > developer.microsoft.com > > develop.kde.org > > gtk.org > > a comment on this would be that I don't see these as similar projects, > but sites of multi-million dollar companies with very diversified > products or large project backed by big companies. > > We are not such, nor we compare directly with "gtk", but somehow a mix > between gtk and gnome. This goes back to a question I asked about what the vision for GNUstep currently is. It's obvious why that wasn't answered now: there isn't any. That led to the beginning of the original thread, where I discussed having a hard time understanding what GNUstep even is. To reiterate my purpose in starting the original discussion: the web presence ("PR" for lack of a better term) makes it difficult and confusing for _new_ developers to understand, and then start using GNUstep. Steven chimed in by creating this thread, describing that he was also having issues getting started. I believe a large part of this is because of the website. An app developer isn't going to go look at the GNUstep code on GitHub, especially given that it's in an almost-foreign language. (I'm not saying that we should abandon Objective-C, but just trying to show that the website is critical for GNUstep given the state of the world.) On Wed, May 14, 2025 at 11:25:40PM +0200, Riccardo Mottola wrote: > The misinformation you write requires a little bit of correction, even > if it will not improve the discussion or change anybodies mind. Trench war. This has to be the best quote of this whole thread! > Gregory Casamento wrote: > > I’m not entirely sure why them being companies is relevant to whether > > this separation makes sense or not. Could you provide an argument that > > makes this clear? > > May it occur to you that the needs of a company with many products and > departments are different from our project? It seems to me that the greatest current need of the GNUstep project is to attract (and retain!) new developers for apps and hopefully libraries too. The current website has failed to assist with this in AD 2025. It may have helped in the past, but the web moves extremely fast and so do the users of it. > Actually, Xfce has core components, Applications, functionalities > divided in "sub projects" not unlike us. Thus checking around is > instructive. I don't want to dwell on this, but this only adds to the confusion. As a newcomer, I do not see GNUstep as a desktop environment. GNUstep has tried to shake off that perception for a long time. Also... even though I use Xfce too, it's only because every other Linux desktop environment is quite poor. Xfce being one of the best DEs only shows how dire the reality is for a desktop Linux/BSD user! On Wed, May 14, 2025 at 11:43:23PM +0200, Riccardo Mottola wrote: > Gregory Casamento wrote: > > > > Reasons to have developer.gnustep.org <http://developer.gnustep.org>: > > > 3) Can provide documentation along with clear examples in a place that > > people can reach simply by typing the URL without having to wonder > > where to find it. > > Personally, If I need apple documentation, e.g., I just slap in the > class name in a search engine and add perhaps "doc" "reference" > "example" and hope something comes out. > Other times, if I need an overview, I go to the website and browse > around. Click on a link or a menu. In that case, it is of no difference > if stuff resides in www.gnustep.org/developer or developer.gnustep.org This is one of the main roadblocks in this conversation. There are two parts to this: 1. This isn't about how you use the web, it's about how new developers use the web. 2. Whether it's a subdomain or not, that shouldn't matter. Maybe gnustep.org/developer just redirects to developer.gnustep.org. Either way, a developer "portal" is what new developers look for (also see my last comment). > Most of the stuff you cite for developers is just GitHub. We don't need > more. People will get code or release form there 99% of the time. I > extra maintain duplicated information on the website or wiki, but that > is convenient for people looking around and for the rare developer using > it. I bet a beer that most other developer either use ready packages > from their distro or just clone the git repo. I'm not sure if this was your point, but I find it difficult to use ready-made distro packages for GNUstep because almost all of the distro packages are woefully outdated. Building from the git repo has become necessary (though I'd prefer not to spend the time compiling this every time I need to test it) because there aren't enough developers and power users to help with the project. > In the case we think subdomains are of use, I would rather gather a > documentation site. E.g. docs.gnustep.org with different kind of manuals > as well as reference documentation. Why not docs.gnustep.org for the cleaned up and updated content from the wiki and developer.gnustep.org for the API reference? Developers can see a link to the docs section on the front page, then if they need the API reference, they can find it under the docs section? Think of the old Mac docsets: all of the developer info in one place and with the ability to keep an offline copy. -- Luke Lollard
