How I see pump.io now

(lee este artículo en español)

TL;DR

• Usable, strong potential, needs users and love
• This is only my opinion!

Long version (too long…)

I’ve been using Pump.io since March 2013 for testing, and more intensively since the migration of Identi.ca in July 2013.
I’ve written some posts about Identi.ca, my favorite social network, and on the migration too.
After some months and active development in not only the core software, but a lot of third party services, and after lots of discussions, bugs, shared knowledge and ‘life in the pumpiverse’, how do I see the pump.io network now?

Note: this is my opinion as a, let’s say, involved user of pump.io. I say “involved” because I use it a lot, I’ve tried several clients, opened several bugs and tried to fix others, I do bug triage at the issue tracker and try to help newbies. But I’m not part of the pump.io developer team and I’m not in more contact with the project leader Evan Prodromou than any other user/fan can be. So beware, this is my particular, humble opinion and it does not reflect the official view of the pump.io project.

Usable

The Pump network

The software is usable. It’s installed not only in the E14N servers, other people have deployed it in their machines and they are using it. Some people have installed it in a Raspberry Pi, some in a VPS, some in a personal PC. Some people host just one user (themselves), other have more users (registration open, or by invitation).

E14N, the company offering the pump.io software, offers identi.ca for its previous users, and some public instances where you can register a new account at no cost: fmrl.me , pumprock.net , urmf.net , pumpbuddy.us , hotpump.net , pumpdog.me , pumpit.info , 1realtime.net , microca.st , pumpity.net.

The e14n.com server is also a pump.io server, but does not allow public registration.

The number of instances/users known by the OFirehose (a node where you can register your pump.io server, in order to offer a network-wide public timeline) is in pumplive.com. There are probably more instances in private networks or not registered there.

Jeremy Pope maintains the site pumpstatus.jpope.org where you can see the uptime and checks of many servers.

For what I’ve seen in this time, downtimes, when they are, are mostly due to hardware failures, problems related to configuration or system administration. In many cases the uptime is short due to reboots of the server to update the codebase, in order to get some new features or fix bugs. No longstanding open issues related to performance or operation errors show that they are fixed quickly, as far as I can tell.

From the sysadmin perspective: deploying/maintaining a pump.io server

About installing it, I cannot say it’s hard or easy. I consider myself a power user, sysadmin apprentice (enough knowledge to install a GNU/Linux distro, or a LAMP stack, or download and build source code if there’s not a package for that software in my distro, and the dependencies are listed somewhere).

I tried to install pump.io in my laptop (just for knowing if I’m able to do it, and test some tweaks in the code, not for daily use). In that time, node.js was not packaged for Debian, but I succeeded in downloading it, compiling, installing it, and then repeating the process for pump.io. However, I didn’t setup a database or so, because as I said I just wanted to test the user interface. Now that node.js is packaged for Debian, I suppose installing is becoming easier.

There is documentation about installing pump.io in several platforms in the wiki.

The Pump “ecosystem”

In addition to pump.io, E14N is developing some other pieces of software for certain features or tasks in the Pump network. Among them, we have:

• OpenFarmGame is a social network game to show the possibilities of the pump network
• Spamicity.info is a server that filters spam in the Pump network
• OFireHose provides a public timeline of the Pump network in ActivityStreams format (JSON)
• pump2rss provides a RSS feed of anybody’s public timeline
• pump2status.net (in development) will allow to connect an account in pump.io with an account in StatusNet, acting as a bridge
• ragtag.io (in development) will allow to see all the public objects in the Pump network tagged with a certain hashtag.

Other people have developed libraries, clients and tools for being used in the pump network. Most of them are linked in the pump.io wiki. The ones that I remark:

• PyPump is a Python library for interacting with a pump.io server
• Bashscriptville is a set of bash scripts that allows to post info from certain sources in the pump network, such as “The Word of the Day”, the XKCD comics, Shakespearean insults, the Astronomy Picture of the Day, and others.
• The XR115 conversational bot joined pump.io and it’s learning there.

Needs users

Mainly because of the network effect. Success of a social network is highly dependant of the number of people using it (it also depends on it being flexible, powerful, shiny and so, but we could discuss which is first).
Other reason is that now, the majority of the pump.io users are StatusNet-biased. I like StatusNet, is a great software, but maybe we are stuck in trying to make pump.io be like StatusNet and new people may bring new, better ideas.

But from where can come users to pump.io? I see three kind of users:

• People that don’t use any social network: if you’re new to social networks, please use one based on free software. It can be Pump.io, Friendica, Diaspora*… whatever. I know Pump.io, that’s why I recommend it. I came to pump.io like an accident, since I was using Identi.ca for microblogging, and knowing that the migration from StatusNet to pump.io was going to happen, I began to be curious about the platform. It’s new, so many things are still on development, but it’s new, flexible and modern, learning from other platform’s failures. Why not? And the most important thing: there’s cool people here! Not overwhelming activity, no adds, no spam (but if you want it, you can go to the firehose 🙂
• People using other social networks: Pump.io can be another one, or a substitute for Facebook or Google+, for example. If you are a company or all your friends are in the other social networks, and “you cannot not be there”, maybe you’ll want to wait until we have bridges for crossposting everywhere (anyway, look at the ones that we have already, maybe they’re enough for your needs, or you can use Friendica for crossposting everywhere with minimal effort). If you have a certain interest in the libre software world (maybe you contribute in a project, or are a company based on libre software, or just like hacking), it’s nice that you have a presence in a libre software social network like this one.
• People looking for social networks for private use: in my opinion, pump.io looks perfect for setting up a family private network or an institutional one inside the intranet. Remember, it is thought to be scalable up and down, and federated from the beginning. You can configure a common web interface but also give freedom to users for using the client that they like. We’re working on internationalization and localization, but if you cannot wait, you just translate the strings in the code and the utml templates and go. It’s free software, you can do it!  E14N is working on shipping plug-and-play pump.io boxes. Own your data, enjoy your communication.

Needs love

I’m just a pump.io user, I lurk in the IRC channel, in the issue tracker (well I do some bug-triaging), and follow Evan and the pump.io hackers. I’m by no means official voice of the pump.io needs or roadmap, but here you are my personal ideas on how pump.io could be improved and how can we help.

Roadmap

I’m sure that Evan has a clear roadmap for new pump.io features and bug-fixing. We can smell part of it in the issue tracker, since some issues have been tagged with a milestone, and others have been closed with the message that those things should be improved by third parties. But a clear roadmap could help users to know what to expect, and contributors to help more effectively.

Maybe it’s just a matter of classifying the opened issues by milestone, and write a wiki page. Mmm I’ll ask Evan about it.

On the user side, I think it’s useful to review the wiki from time to time to know the more important features that are already implemented (in pump.io itself or by third parties), and to search the issue tracker prior to posting a supossedly new problem. Or maybe ask in the IRC channel.

E14N ‘status’ page, contact page, and sysadmin helpers

I will not say how Evan has to rule his machines or his business. So please take this paragraph as just one humble opinion that can be moved to trash if it’s not useful.
E14N servers (specially identi.ca, but not only) are the public image of the pump.io software and network. And many people may not know to distinguish between a bug in the software than a puntual problem in the network or the system. Many people don’t know where to report problems or requests that are not tied to the software development, or they know that it’s by mail to admin [at] e14n.com but they don’t know if somebody already reported the problem.
For now, pumpstatus.jpope.org is being a great help and some people talk in the IRC channel when having issues, before sending an email. But having a “status” webpage where to look if we’re experiencing problems could help. And maybe giving writing rights on that page to some persons in the role of “sysadmin helpers” would help too.

Features that give control to the user

Pump.io is libre software, that means that the user can control what the program do. But not everybody knows how to hack on node.js in order to tweak the program to do what they want, or are not able to setup their own instance. Some features that in my opinion, would attract users and keep the existing users happy and involved:

• Activate email recovery and notification in the rest of E14N servers different than Identi.ca
• Ability to backup an account, delete an account, change the email address, configure email notifications.
• Search posts and users.
• Internationalization / localization of the web interface to different languages.
• Display contexts in the web interface.
• A public timeline (like the Firehose), accesible by a web browser, with no spam (or maybe a “hot topics” view, with the public posts that are shared or liked at any time). A window where a possible new user can look at.

Contributors!

Enrico Zini said once: “Don’t ask yourself what you can do for Debian. Ask yourself what Debian can do for you. Then work until you make it do it”. I think this thought goes well with Pump.io too. There are many pieces of code already written that you can copy or hack, and there are lots of projects (not only the core, remember the clients, libraries, side projects…) where your help is valuable.

Hey, and don’t say “I don’t know JavaScript nor node.js”. I don’t know either… but code is code, you can abstract the syntax and learn and understand (at least more or less) how it works. Other pieces of the pump.io ecosystem are written in other languages. Or you can help in the issue tracker trying to reproduce errores or finding duplicates, or read/write documentation or FAQs, suggest ideas, testing features/clients/platforms…

And if you are a programmer, please have a look at the issue tracker and pull requests from time to time. There are probably some proposals to fix some bugs that can be improved or supported in order to be merged quicker, or rejected so the sender can rethink his/her proposal or just go to help in other part of the code or the system.

Promotion

Finally, I think it would be nice to have a video presentation about pump.io (let’s say 5 minutes, maybe less) that can be translated to different languages and shared in everyplace to spread the word about this social network. Maybe it’s already done (if yes, please send me a link!). Maybe some users can try to make it. But I think that I’m not the suitable person for this task. Look at this post, it’s too looong! Write something for just 5 minutes is a torture for me 🙂