yeah, nice to see. http://www.foresightlinux.com/
I did not realise that it was out. I just wondered why conary updateall took long than expected. Download and test it!
I like to talk about how FLOSS projects can or should develop. The first thing is to accept some facts:
As rPath does not have a 2.4 kernel it won’t work. Because I would need an i686. I think it should be possible to habe a 2.4.x kernel in rBuilder and than remix rPath? As the whole idea of rBuilder is what one could call “Remix Linux” Let’s see.
Tomorrow I am going to change our house intranet server. Until now it ran on Fedora Core 4. As readers of this blog know I am switching away from Fedora for many reasons . The current setup is:
In the future the setup will be
This way I am able to run the stable, underlying rPath system as long as it has power – and work with the other appliance(s) as I like. Maybe a second network card would be a good idea? Generally I like the idea of seperating the intranet and the extranet traffic. So that the server exchanges internet data through one card (eth0) and data for client computers on the other (eth1). This also would allow to clearly control with a netfilter what should come in and what not.
Well I COULD also just set up an IP alias eth0:0 and work with that. hmmm. But my guess (as a non netfilter expert) that it is more clean to do this on two cards. Anybody likes to comment on this one?
A bit of a disappointment as there were no preparations for possible presentation and no talks (besides my own). Also I missed a bit of a discussion culture – I would have expected that a group of people try to keep the discussion in a circle so that everybody can listen and add some points. – If discussion is split it splist even more and people get distracted, which again distracts more people. So it is needed that everybody keeps up some basic discipline and care about how good everybody can follow. Maybe the problem is that people are used to have some kind of leadership that enforces discipline – and if there is none they are not willing to follow some basic rules. As an anarchist I don’t like leaderships and i think that as grown ups we should be able to communicate with each other on a free basis. Nobody is forced to come to this event (again), neither am I. I hope the webmonday in Kiel will develop its own culture so that is fun and also informative to go there.
I am very much a friend of software like The Rosetta Translation Portal. Still many translation happen with email communication which takes long – although maybe in the communication process people can learn more. But thats not my point. I think, I wish that more software would try to
combine common results. As: Common translations – so if somebody translates a menu entry in KDE or GNOME you have the option to just use this in Firefox or the other way round. So immediate worldwide collaboration. Chatting should be integrated (take Jabber!) The same could be true for bugs. There are bugs that have the same roots in many distributions and also in GNOME or KDEs bugzillas. But still: No direct data exchange is there. Why should a user have different bugzilla accounts? Couldn’t we have the one developing ID (with openid?).
So this is something I would like to describe as Web 3.0 – its not that I liked the Web2.0 hype – but as we have this term I like to tell with the term Web 3.0 that there is much more than just nice graphics and AJAX. And if you see the web as a real WEB it is sad to see that not much of interwebed websites do exist. Web 2.0 has some nice features – but it does not solve collaboration problems. Most Web2.0 is about interactivity – but what we really need is automated collboration. And again I think that the old principles of Unix could be an example of how things could work.
I now just talk about that in the light of help for free software projects, because I see how much time is wasted in doing things that sure have been done and said many times before. And that is because web solutions are seperated. Web2.0 tried to solve these things by making web sites stronger and more easyto use. This is good – and we should not walk back – but I feel that their is a huge overweight in making better usability while less work was spent into making work redundant.
So we now have dozens of independent social bookmark systems or wiki solutions. But they are not talking to each other. So we have to have hundreds of accounts – and if we switch a provider – we have to start from scratch. if we are lucky one provider was able to write an importer for the data of the old provider – but this is seldom and mostly will only be done if the data is in a market dominant tool. This is what Salesforce.com did for Excel spreadsheets. Even on the desktop – OpenOffice.org did not care about Gnumeric spread sheet standard – both prefer to use Excel as exchange. So that one could wonder what FLOSS applications would do without Microsoft?
So after a few weeks of using Foresight what should I say? I like the community which is generally helpful and friendly. I like that it is possible to do things and people actually beeing happy to get help. Would I recommend Foresight for simple users. Well the conary commands are very long and thing can get complicated because you do have many options and sometimes might want to get things from different repositories. So the command line tools are not really for simple users if it is getting complicated – on Fedora you often only install a repository info RPM and then make ‘yum install packagename’. With conary you do not need to install a package info or edit any files – you can type in sources and then you have it. But as long as you are happy with using an uptodate GNOME desktop you should be able to use Foresight also if you are not a developer. To install a a package from command line you simply type ‘conary update packagename’. There is a graphical tool that works with a web service, so you are accessing it via a web browser. This works most of the time but still has some minor issues (stops), but if you restart an update it has worked every time. We are expecting Foresight 1.0 to be release soon. There are some issues that still need to be resolved and if they are i think it will be a very nice distro and sure many more people will love to try this out. I think Foresigh really adds something to the distribution market because they have some new approaches for common problems. And I like the distributed development idea. BTW: Conary is CPL (Common Public License) whichs is incompatibe with the GPL but only because it forces somebody to give rights for patents to all those who use the software and you can license parts of the software as proprietary software. This licences was coined by IBM.
I like to share this idea with you: I think that it would be interesting if we use IPv6 not only for pure network purposes but also for identifiying ressources. And with ressources I also mean ideas, people, products, stars, chemical elements, words, languages, revisions, spam mails, viruses (computer and human)…
So this would mean that every product would have a product ID that really is an IPv6 adress and also its class and its revision are IP adresses. So this would enable everybody to uniquely identify a product and its heritage. Right now instead of IPv6 adresses websites use an IPv4 adress plus multiple idenficators.
These adresses could also be used to identfy actions, face expressions, statuses of roads, weather conditions, exit statuses of programs, meals, recipes, cities,…. – And this could also include many different suppliers of these adresses. So Wikipedia may have uniqe adresses for every entry – and the USA have another entry – and then you have two identifiers for one city – but you can combine them – make relations to them. I could also imagine that as the possibilities are just endless that we could again face a limit in the far future of these adresses. I know – all insects may get an identifier – but as this system would include potentially endless elements to get an ipv6 adress and also to give combinations an Ip adress we could produce VIRTUAL instances. So a combination of Berlin and Beer would not only be two IP adresses but this combination would get an IP adress of its own. And also every element in the world could potentially have endless adresses itself.
Why would this be so powerful? Because we would only have one system that would allow multiple suppliers and endless meanings and interpretations -but still you will always know who owns the ip adress. So maybe if you type a random IPv6 adress in on a website it can tell you what it means. The beauty in it is that you will always know what an identifier is unique – and therefore will not ever need to think if it might be used somewhere else (well a t least if you typed it right).
I dont think that this would be the end of human readable identifiers – they are handy – so people will still use them – but at least in computers behind every identifier there will be a truely unique idea. This will enable science to make a HUGE jump because tools get more reliable and errors will decrease drastically. This can also produce new errors and allow misues, because knowledge is power and whoever knows an internal ID might be able to do really bad stuff with it. But still i think we have to do IT!
I am sure somebody else has had this idea before, because it is just too obvious, but I have not yet found it. if you like please add related ideas with your comments.
They just GPled it : http://secondlife.com/developers/opensource/getit
So thats what most Linux users in SL will not like to see, but I am VERY happy.
Indeed, I am getting tired of too many developer blog entries that either just shed light on private stuff that nobody really cares – or they just talk about what they do without giving users something to understand. Not that I think that this should not happen – this can help other developers a lot. But if you look at planet.gnome.org you ask yourself if this post should not either be kept private or be put on developer.gnome.org. The big question is why we really should put these blogs near main WGO (www.gnome.org). my guess is that only 10 percent really are posts that have the user in mind. And OTOH – as I already said on CVIsm – I more and more get the feeling that many developers live in their own world and have no idea what users want. Well maybe it is enough for them to do their thing – but when you try to build a common desktop like GNOME or KDE tries to – you better listen to the user! Ok, there is no ONE user but many – and developers are also users. But generally they act very differently from normal users. Even “advanced” users (in a MS Windows sense) will not be able to do something rather simple as to compile a program.
So I hope some developers may start from the beginning from time to time. Think how something may look if you would not know – what you know. This is what I had to learn from my customers. Then things that look simple can get really difficult – and also sometimes – for a user it is more easy to go a long way that he is able to find herself – as if we have a solution that really requires somebody to be some what of a developer to even start with it.
Usability studies are nice, but what really should change is the attitude and also the process of finding solutions. I think free software often is much easier as the Windows world. We had automatic updates long before – and still many things can be handled with Windows at all. But some tasks are just too complicated for simple users to accomplish.