With Google+ closing today, you might have just signed up to diaspora* looking for a new home. We're so pleased that you've decided to give diaspora* a try. It's fantastic that you've joined us! We have a big existing community of former G+ users, and you could try tags such as #gplusrefugee and #googleplus to connect with people you know from G+.
We hope that you quickly find your way here, and enjoy getting connected with people and content.
Here's a few tips for getting started:
There are some step-by-step tutorials on our project website covering the basics of how diaspora* works, and how to get connected with people. There's a link to these in the side-bar, should you ever want to read them in future, and there's also an in-app Help section you can access through your user menu (at the right-hand end of the header bar).
You can search for people by name or by diaspora* ID (which looks like an email address), if you know it, from the search field in the header bar. You can invite people to join you in diaspora* by sending them a link to our project website, from where they can choose their own pod to register with. Your friends don't need to be on the same pod as you in order to connect and communicate with you.
One good way to start getting connected with people is to start following some #tags of subjects that interest you (maybe #music, #photography, #linux, or #activism ... whatever interests you) so that you will see content on those topics in your stream. You can then start following people whose posts you find interesting, by placing them into one of your aspects.
If you have any questions, post something with the #question or #help tag, and people will try to help you. Make sure to make your post public so anyone who can help is able to see it!
diaspora\* has some very simple community guidelines, which we hope everyone will feel able to follow so that all members can have the best possible experience here.
We're trying to improve diaspora* all the time, fixing bugs, honing performance and adding new features. But how fast we can do this is limited by the time developers have. To improve diaspora* faster, we need more developers! If you've just signed up and can code, or know someone who can and who might be interested in joining our team, get in touch!
I like Diaspora* a lot so far! Glad I found it... It's so nice that here are no ads but I have still got problems to get the right # Tags. Sadly there is a difference to #Twitter and their 500mln users...
This is just my thoughts, and I’m writing this in response to no one in particular, at the same time to all interested in this thread and the future of Diaspora.
I’m one of the new lurkers here and the freshness and freedom of Diaspora feels really good! I’m still trying to figure out where to go with this. Everywhere I go in search of alternative media someone is spewing content I don’t wanna see through (here:) tags I wanna follow. This has been going on for more than two decades now. But being free from commercial channels Diaspora provides a very good version. It could be described as a kind of a slow forum for thoughtful people.
Maybe the most devine (can I even use that word here?) reason for having people come here is to liberate them from being a data-producing product exposed to commercials. But maybe we should not search for the way where we reach and catch as many as possible, but let it grow slowly. Many wanna try something else because of a variety of reasons, not just because they wanna leave now “traditional” communities.
One can almost use the words ecological thinking about our growth. But as all good growing things, you don’t want them to die, and I am openly guessing a worry like that is behind your post @Robin - and a bit this one too. Being an anthropologist, this is a tad interesting to me. If anyone “in charge” would iike to, I can work with them, putting together a simple questionaire for trying to find the intention - or why.
Maybe Diaspora is what it is for now - if so just an alternative version to explore for some. A thing that can help understand what commercial forums are and what the users want with them, not a bad thing in it self. Then there are other ways to go about growth in the long run.
But I too wanna see more interesting porfiles and posts in here. Which returns me to the start of this post - my own lurkingness in here. That’s my “job”. Central in my life now is starting a small business. But I don’t wanna spew certain content in a commercail-free-forum like this. That contributes to my present lurky profile. but more is coming. (This post is one thing.)
The latter is a free-for-all where you will see post from anyone and anywhere with any content.
The former, your personal stream, is determined by Aspects/#hashtags/people sharing with you. Yes, some crap will seep in to that stream, but not as much as you will see on the Public one.
If someone is sharing with you and you don't care to see their posts (not replies, just their original posts), then you can Ignore them by pulling up their profile and click the Ignore icon (circle with line through it). This will prevent their original posts from appearing in your personal stream, but not replies they leave on other people's posts.
It is what it is... until improvements in true blocking are instituted here.
You are correct about the #hashtags being insufficient in themselves. The problem is that spammer, click-baiters, all-round-troll-assholes will often tag their posts and replies with false #science, #mathematics, #news, etc. tags. Unfortunately, the system will add that disguised crap into your stream. :(
@Christoph S of course, and it does not have the automatic data grabbing from google translate. But IMO it is too much of a hassle to go to the deeple website and have a message translated. A Firefox addon is so much more convenient. You don't have to leave the page you want to translate. IMO this is second best (after having an option to translate in diaspora itself)
@Rob Bosch that's pretty cool. I didn't know there was OS stuff out there to play with. The basic command line app they provide even does auto language detection, which is pretty sexy.
Couldn't get an idea on the amount of horsepower required to do the actual translation, but assuming a manual translate button for a post/comment, the engine has to detect language (this might be doable with dictionary words) to determine what model to use, and then do a syntactical translation from detected language to target language. I could possibly do it for one or three people on my personal server, but even my small pod, I wouldn't want to try it on my cloud server.
The models are a bit trickier. Moses uses the same machine learning techniques Google and others use - to provide a good syntactical translation, you need to translate sentences, not words, so you have to feed the machine equivalent sentences in both the original language and the target language, and the more the better. This is done with books. For example, for a minuscule dataset, if you want to translate FR->EN, you might feed Moses The Count of Monte Cristo (an exceptionally large book) in both languages and tell it to sort things out. From that, Moses creates a model for the translation engine. Repeat for FR->ES, ES->EN, etc.
Straightforward, but where the problem comes in is that you need a huge dataset, or funny things can happen [Google for example had a 'hilarity did ensue' moment a few years back because in some cases, translating some African languages to English started spitting out Bible verses - because some more obscure languages don't have a lot of literature that is available in both source and target languages, but nearly every language has the Bible.]
So big datasets = big horsepower. The Moses documentation suggests a small dataset of 300K sentences can take 1-2 days to build a model on a gaming PC. So that gets you a minimal FR->EN. You have to build a separate model for EN->FR, etc. So ideally, you'd have a cluster available (to be fair, some of this may be able to be done by loading the learning tasks on top of an @home engine, but the main server would still have to be pretty beefy).
I'd be happy to "donate" surplus cpu cycles of my VPS (6 xeon E5 cores + 16GB RAM) to help with the learning part... it would be awesome if there could be a seti@home-like option to help training the translation engine...
@Rob Bosch first, we would need thousands of person-hours worth of native language speakers to scan, OCR, and proof-read tens of thousands (or more!) pages of text for each target language to build a dataset Then we need a host server/cluster to store tens of gigabytes of data per model) Then we need a server/cluster (same one as before, probably) to distribute the tasks and combine the results Then we gotta assume we have that cluster, and it can host translation services. * Diaspora* pods can host a couple thousand people on two cores and 8GB RAM. I seriously doubt a translation API on such a server could sever for more than like 20-50, including the stream performance hit, (And since most pods are hosted by individuals providing the service for free, asking them to radically upgrade their servers isn't particularly fair) But, removing the hardware issues, and assuming a decent set of volunteers to do the data & software efforts, how long do you think it would take? Two years? More? Post editing, groups, and some other features require a fraction of the effort. Being able to Federate with ActivityPub (like Mastodon) services would require less effort. And any of those add functionality that doesn't duplicate what can be implemented by users today via a browser plugin
There are a lot of things that are interesting, sustainable, good for distributed social networks, and some of them may even be even a good idea for diaspora*. Running our own translation service for the entire network is neither of those things.
@email@example.com may not work if your server doesn't "pull" those contents, and you'll end up nothing for the hashtags. Small servers need people to "follow' others to pull in the foreign (federated) content, and hashtags will only work afterwards.