I have had quite a bit of feedback from my post about moving away from Feedly to a self hosted solution called Tiny Tiny RSS.
Some interesting questions have emerged from people and to be fair it certainly deserves a follow up.
I made my instance of TT-RSS available on the public netw, because I didn’t see the point of installing my VPN software on my eink reader. I also installed the official TT-RSS app which is a 7 day trail before you buy the full version for 4 pounds. I haven’t bought the full version yet because the app doesn’t seem to work well when offline? It would be great if the app understood if the device was offline and automaticilly disabled the update feeds option. It currently doesn’t seem to do this well… I much prefer Greader for this, but ttrss app isn’t far behind.
Simon commented he paid for Feedly because of the IFTTT options, but it seems weird to pay for this because you can easily turn most of TT-RSS into a another feed and IFTTT has a RSS option which you can use to trigger most things. This reminds me of my work along while back about pipelines.
Because of this, I have been thinking about feeding Greader with the RSS from my TTRSS install. The only real disadvantage is nothing would be synced to the server? This is also something I’ve been thinking about with a linux desktop reader like thunderbird because I can’t seem to install a TTRSS reader which works.
I tried a few but each has had problems.
Feed the Moneky looked very promising but when I finally get the appimage loaded, it shows nothing? Feedreader looked great and after finally getting flatpak working, I’m faced with the error that I need to install the api-feedreader plugin in my TTRSS server. How I do this when I’m using docker is a question I have no answer for, except it seems I need to use another docker container?
So generally its going well but hitting a few points which need straighing out. It would be so useful to compile supported applications into a wiki page.
Oh I found this useful when understanding about appimages, snap, flatpak, etc.
I’ve been meaning to switch from Instapaper to Wallabag a long while ago but been so busy. Originally I was going to install it on my own server using Docker as a container then looked into Rkt after a talk with Jack from work. Lofty goals but I did install Rkt and installed the Docker app via Rkt. But thats as far as I got…
Then I tried using framabag.org server but couldn’t get it to connect to any of my clients because its version 1.x it seems. Then I saw a email about app.wallabag.it.
I joined, ported all my instapaper archives over and paid.
Now I have Wallabag on my Eink Android tablet, Nexus 5x, Nexus 7 and Chrome. The only problem I’ve had is getting the Firefox add-on to work with it. Theres some really nice features like the ability to add automatic tags on the fly, custom RSS feeds, 2 factor auth and a kind of Oauth for new clients.
Generally I’m pretty pleased. I would like to see IFTTT support (although RSS helps with this), Gnome shell support and federated server support. I haven’t quite killed my instapaper account but I’m pretty close now. Give it a few days and its goodbye!
I have mixed feelings about Slack. Its good but I worry about the deadend nature of it. A clear sign of this is IFTTT’s recipes involving Slack. Every single recipe has Slack as the output (action), not a single one has Slack at the input (trigger).
This is the notion of the roach motel or walled garden, right?
I have been looking at alternatives; before anybody starts; YES I know IRC, yes I have used IRC in the past and more recently. I get it but I always find it not a great environment, especially for modern work, due to the obscure never-changing syntax and behaviors. I also know people who use slack via IRC and I have tried using Slack with a Jabber/XMPP client but its painful for anything but talking to individual people I found.
I looking at others and found…
But recently someone suggested trying Telegram groups and channels. Then using existing tools like Trello, Google Docs, etc.. for permanence? This may run back to the small pieces loosly joined way of working, which isn’t so attractive. But allows for diversity of uses, clients and services. Dare I say something to think about when thinking work 2.0?
Part of my worry is slack trying to be the end point for everything. Its seductive and easy like Facebook is, but scratch one of the sides and you find the walls are more concrete than expected. Yes there are permalinks, bots and markdown content but it feels very hidden?
I’m now a owner of the Philips Hue lights after my mistake buying cheaper Hue lights from eBay (which turned out to be the American versions) and finally converting all the lighting pendulums to standard B22 bayonets from CFL BLT 4 Pins.
I bought the Zigbee bridge from ebay too and been buying the bulbs one by one. But then I decided to buy a starter kit, as it was more cost effective and I could sell the spare zigbee bridge if not needed. Everything was fine till I couldn’t control the new lights. After a look around the web, it become clear the bulbs were locked to the zigbee bridge which it came with.
I won’t lie I was peed! I took to Twitter to tell Philips what I felt…
Locked internet of things devices, deeply worries me! Philips say they do it for security reasons but frankly thats balls.
After the steam had stopped pouring out my head and I put the bulbs back in the box to send back tomorrow. I looked around and found people talking about a app called Lampstealer. It seemed to factory reset the bulbs so the old bridge could discover it like normal.
The problem… It ran on Windows and OSX only. Of course the hackers got us covered!
When you buy two Philips Hue light start kits, you have the problem that the lights are already paired with the bridge in each starter pack. When you search you will find a lot of people whining about how unfair this is and people talking about the “Lampstealer” OSX app that Philips released to fix it. I tried using the lamp stealer app but it would never find my bridge. I could also not use QuickHue which supposedly supported the lamp stealer function because it was compiled for OSX 10.8 and I still run 10.7.x. And compiling it from source with xcode didn’t work, likely due missing libraries and other mistakes I made since I’m not too familiar with Xcode.
I found out that the solution was really really simple, and requires no OSX, java or advanced rocket science. Place a bulb of the second starter kit into a socket within 30cm of the bridge from the first starterpack. Telnet to port 30000 of the bridge and type:
The light should blink a few times to acknowledge the hostile takeover.
I did it and now I’m sitting pretty with 7 Philips Hues all tied to the zigbee bridge I bought from eBay. Everything is now working correctly and I’m looking forward to playing with the geofencing and ifttt controls. I just need to sort out my lampshades now…
I switched from Pocket/Read it later ages ago when instapaper started supporting Kindle deliver. I use it all the time and at some points, its been the main reason for having my paperwhite kindle. The nice thing is its pretty much automatic, and I tend to just turn the kindle on and it automatically gets the latest instapaper over wifi.
I side stepped the Kindle’s annotation system by having the same document on my phone and being able tweet it directly. Its all pretty great, although I did consider using diigo’s readitlater feature, especially now the mobile app is better.
However I have wanted a solution for saving non youtube videos (I use watch later for youtube at the moment). The amount of times someones posted a video to Vimeo and I have had to bookmark it or something else.
Well it seems somewhere in the updates, they added support for video from youtube and vimeo! Great stuff my pro payment is well spent, not only that the ifttt channel has been updated with a number of improvements, but I really need to start using the highlights/annotation feature more.
Pushover makes it easy to send real-time notifications to your Android and iOS devices.
I love the fact its simple and works with ifttt. Heck I can imagine some scenarios with ifttt and taskrabbit. But I’m left wondering the advantages of a private push service over using something like xmpp/jabber?
Yes its going to be somewhat cheaper than using SMS, more extensible than twitter direct message and less hassle than xmpp? But running yet another service on your phone? Isn’t this what Amazon SNS was about?
I have already talked about IFTTT to death and how I liken it to some ideas and work I had around pipelining.
All the new movement in this area has been in the online space but I found this little app for Linux which operates in a similar fashion to the very old conduit (Conduit) but its focused around system events rather than webservices.
Cuttlefish is a tool which can execute various actions when specific events are triggered. For example, you can change the proxy mode depending on the currently connected wireless network, unlock your computer when a specific Bluetooth or USB device is connected or disconnected and so on.
I can easily see how webservices can be written into the application, although there is no roadmap yet.
Tim Dobson sent me this over twitter for my consideration…
Tim Berners-Lee says demand your data from Google and Facebook…
World wide web inventor says personal data held online could be used to usher in new era of personalised services
Seems people have forgotten the work which took place during the late 00’s as one of the founders of the Data Portability group (which still exists by the way). The group was made up of quite a few people all over the world and we successfully convinced the likes of Yahoo, Plaxo, Myspace, Google, Facebook, etc to take data portability seriously.
The turning point was when Robert Scoble tried to take his contacts out of Facebook and into Plaxo. Interesting to see Tim Berners-Lee finally getting the point.
Although to be fair he goes much further thinking about a standard way to export data.
Right now both Google and Facebook have export features and each one is very different in structure. I personally regularly export my data from them every month along with my wordpress and others. I find Google’s Data Liberation centre the best because it gives you control across the board, but then again Google do have more data from me. But right now its all just for back up purposes.
The next step which Tim hints at is the ability to transform and import the data in a standardised way. To be honest its something we (data portability group) talked and thought about, but we were maybe a little too early. Now seems about right to think about the interchange of data more than ever.
There has always been space for startups to be brokers and transformers of the data. Someone like ifttt.com could make a killing in this space, specially if they start charging for use of their pipes (something I suggested while doing the xml pipeline stuff). Could make a nice sustainable business
Love If this then that aka ifttt.com and I also like the idea of Amazon’s Mechanical Turk but its got its limits. However I’ve been watching the rise of Taskrabbit.
I wonder what it would be like if you combined ifttt with taskrabbit?
Suspect all type of chaos would happen because Taskrabbit is not about automation (can you setup repeat tasks?). I just find the concept very interesting…!
Love the idea of web pipelines, as you already know. But there is no reason why it can’t affect not only the real world but human culture.
Hadn’t really thought hard about how the internet of things and pipe-lining work together. But listening to John Kestner on ITConversations recently, I’ve been re-thinking about some of thoughts around pipe-lining.
Those way back might remember I have a massive thing about pipeling born from my love of using Apache Cocoon and XML. Theres actually a very cool XML syntax called Xproc which I’ve yet to really check out in any detail.
I thought about ways to extend Yahoo! Pipes and declared my love for if this then that. But reaching out into the physical world is something else all together…
“Twine is a wireless module tightly integrated with a cloud-based service. The module has WiFi, on-board temperature and vibration sensors, and an expansion connector for other sensors.” Developer John Kestner describes the device and its development. He reviews how it works, what decisions were made in its design, and how It allows you to connect things to the Internet. He also discusses the community that is working to make the product better.
I saw pachube.com a while ago but never really blogged it or thought hard about it. But you can certainly think about how ifttt and pachube could work together really nicely.
The ability to automate and manipulate the physical world is something very special. I’d certainly like to investigate a little deeper in the near future.
I’ve been playing with a few services recently and I thought it might be worth blogging about…
while moving away from delicious I found diigo did everything I needed a whole lot more.
also posted about if this then in a previous post too but I’m finding nice new ways to use it for example to get read items on my kindle back in a form I can blog about again.
Not quite installed this yet but I’m very tempted because I’ve always wanted a way to manage text messages from more than my phone and it looks like DeskSMS might be the best way to do so. Although a lot of people swear by Mighty text too
I read about Glos.si in the Lifestreaming blog, and decided I’d give it a shot because its been quite some time since I’ve really played with a decent lifestreaming service. The one I host on my own blog is alright but is missing quite a few of the great stuff Glos.si or the old fav’s Sweetcron and Storytlr had. This also reminds me I should kill all the other lifestreaming services I’ve used in the past
This has always interested me but its only now I’ve gotten around to looking at it with some time.