My Google Reader replacement
Published 2013-09-26, 14:48
After Google closed their RSS reader I of course was looking for a web based replacement for it. Wanting more a „river of news“ than Pinterest or „Let us decide what you want to read“, the search actually got quite complicated – many of the new reader replacements popping up didn’t match my requirements, and even if they looked like they did, most of the time they were broken or not really thought through.
A more structured list of my requirements:
- free, cheap or reasonable pricing
- fast
- reliable
- nice UX
- mobile options (own mobile version or via API)
- folders
- folder organization
- „river of complete articles“ view where I can scroll to see all articles, all the articles of a folder or all the articles of a blog
I used ReplaceReader, Russel Beattie’s Readerpocalypse-the-Players-article, this epic list and some Google searches to find the following list of replacements:
Feedly, free
Very feature complete. Sometimes strange UX, quite often doesn’t really save read progress. But has an API that is supported by the biggest iOS readers.
TheOldReader, free
Matches the old Google Reader very closely. Didn’t scale well in the first days after the shutdown, actually seems to closely related to Google Reader to innovate.
Aol Reader, free
Big company, good designers. Makes a great rss reader with a perfect design. Shame they needed sooooo looooong to offer users to reorganize their imported feeds. Small feature, big frustration if it’s missing.
Digg Reader, free
See what I wrote about Aol Reader.
Newsblur, limited free, $2/month, $24/year
Oldie but a goodie. Works great, but I don’t like parts of the interface.
Feedbin, $3/month, $30/year
Ben Ubois created one of the newer Google Reader alternatives, and he did a pretty good job. I really like the general style of the product, but really dislike the user of tags instead of folders and the missing ‚complete articles river‘ view.
Yoleo, free (or $9/year donation)
„the beautiful reader“ is it’s tagline, and Jamie Gilgen did a really good job at that. Unfortunately, she also made some very strange decisions and the river of articles, or „waterfall“ as she calls it, is still missing.
Curata Reader, free
„The Cleanest Reader“. Yup, but also missing lots of the functionality I want.
Feedspot, free, $3/month, $24/year
Again, most of the stuff of Google Reader (and more) works and looks pretty decent.
Commafeed, free
„Bloat-free feed reader“. Support the most important stuff, but is also quite slow.
Hive, free
„Content first, no distractions“
Bloglovin, free
Works, pretty, but also not flexible enough. Very Pinterest/fashion oriented.
kouio, free
Prettyyyyyy. But no river view and not really made for lots of folders.
Others
More great readers, but didn’t really bring anything new to the table beside variations in design and which features were or were not supported:
- Lector, $3/month
- Ridly, free
- InoReader, free
- BazQux Reader, $9/$19/$29/year
- go read, free
- rivered,7-day trial, $2/month, $20/year
Tiny Tiny RSS
Because it actually looks good I also decided to include this self hosted reader.
So which one did I choose? Feedly. It just worked when all the others had scaling problems and it supports all the features I need. It’s not perfect, but now I’m used to it.
Wikipedia Mobile #1
Published 2012-04-08, 02:00
This morning I accidentally installed Wikipedia’s mobile app on my iPad. It felt sluggish, strange and not at all what I expected. As bitching doesn’t help and I wanted to use my professional knowledge in my free time, I started poking around. This is the result.
Starting position
The iOS Wikipedia app has really bad ratings. Users complain about bad performance, bad usability and about the feeling that this is more an Android app but an iOS app. In sum, this creates you a rating of only 2 stars for the current version. Ouch.
- http://itunes.apple.com/de/app/wikipedia-mobile/id324715238?mt=8
- https://play.google.com/store/apps/details?id=org.wikipedia
Android users seem to be happier with their 4.5 star rating. The problems only exists with the iOS version of the (similar) app. Strange. Let’s start with the basics and look into the feature set:
- Read Wikipedia articles
- Fulltext search
- Save and read articles offline
- Share articles
- Articles nearby (articles on topic near my geolocation)
- Supports reading articles in different languages
Doesn’t sound too bad. It’s not only stuff a mobile web site could do, so an app makes sense. So lets start to investigate.
What’s the background of all that?
The AppStore description of the app also links to a Twitter account and the source code of the app on Github:
- https://twitter.com/#!/wikimediamobile
- https://github.com/wikimedia/WikipediaMobile
From these two links I basically spent hours reading, clicking, reading again – typical procrastination stuff (Procrastinators beware if you are even near Wikipedia articles…). But I learned a lot 😉 Back to topic: The following links all are relevant to the mobile efforts of Wikipedia:
- http://bit.ly/wikiappbugs (Bugs in Bugzilla instead of Issues of Github)
- http://www.mediawiki.org/wiki/Category:Mobile
- http://www.mediawiki.org/wiki/Wikimedia_Mobile_and_Special_Projects_engineering
- http://www.mediawiki.org/wiki/Mobile_design
- http://www.mediawiki.org/wiki/Wikimedia_Apps
- http://meta.wikimedia.org/wiki/Mobile_Projects/Contribute#Apps
- http://blog.wikimedia.org/c/technology/mobile/
- http://meta.wikimedia.org/wiki/Mobile_Projects/roadmap
- http://meta.wikimedia.org/wiki/Mobile_Projects
- http://meta.wikimedia.org/wiki/Mobile_Projects/Platforms_Support_Status
While reading these pages, I noticed information about two different iOS apps: One is based on Rhodes, one on Phonegap. It turns out the Phonegap version for iOS is quite new and the old one is to be retired and phased out:
- http://www.mediawiki.org/wiki/Wikipedia_iPhone_app
So we have several projects and initiatives in the mobile space of Wikipedia:
- Mobile design
- Mobile website
- MobileFrontend: Extension that produces the mobile website http://www.mediawiki.org/wiki/MobileFrontend
- Zero (text-only) mobile website for Global South (developing countries)
- Mobile apps
Not too bad. Of course the articles in the wiki are quite… organically grown and sometimes hard to differentiate and classify. But I think that’s the case most of the time when volunteers and professionals work together.
The funny stuff: code
Now I have an overview, so let’s get back to the funny stuff: code.
As I have never used git before, I spend some time installing all the tools and learning how to use them. After that I can finally checkout the repository on Github. While I do this, I notice that the README.md isn’t as pretty as all the readmes of the other Github projects I know. So that’s the first thing I decide to fix. 1 hours, lots of reading and 3 minutes of ‚coding‘ and a pull request later, it is done. My commit gets integrated by reedy a few minutes later. I’m in.
But now to the real code. The folder structure of the repository isn’t very clear. There are lot’s of confusing files in the root that don’t really belong there in a multi platform project, several folders that also don’t make sense to me or I can’t identify. Again, this seems to have grown over time. This is definitely something that could and should be fixed.
In /assets/www I find something that looks like the html application. Loooots of Javascript, jQuery and Zepto *confused. Anyway, using –disable-web-security I can open the Phonegap app in Chrome and even get it working! Still, several ‚file not found‘ as non-existant files are requested. But here it becomes quite clear, that the app itself is no bad in itself. It works, the code is more or less structured, you can see what the developer was thinking. Normally that’s a good thing. So as to not fish around without all information, I conclude my little poking session for today.
Conclusion
After getting this short glimpse into Wikipedia Mobile I think that the code isn’t the first problem to fix. First it needs some love in the Documentation and Organization of the code. So that’s where I will start. Let’s see what I can do.
PS: Well, I installed the app on my old iPod 3 (iOS 4.2.1) and tried to use it – completely unuseable. Maybe we should start with the code after all.
iPhone-Apps analysieren
Published 2012-04-06, 22:55
Heute habe ich ein paar iPhone-Apps auseinander genommen und sowohl Funktionsweise als auch Datenverkehr näher angeschaut.
Grundlagen
http://www.uptill3.com/static/iphone_forensics.pdf
Dateien auslesen
http://www.macroplant.com/iexplorer/
(Binary) Plist Dateien anzeigen
http://www.icopybot.com/plist-editor.htm
Cookies.binarycookies
http://it.toolbox.com/blogs/locutus/understanding-the-safari-cookiesbinarycookies-file-format-49980
SQLite Datenbanken anschauen
https://addons.mozilla.org/de/firefox/addon/sqlite-manager/
Web-Traffic nachvollziehen
- http://www.charlesproxy.com/
- iPhone und Rechner in selbes WLAN
- Rechner-IP:8888 als Proxy der WLAN-Verbindung im iPhone einrichten
- Session capturen
SSL-Web-Traffic nachvollziehen
- Proxy -> Proxy Settings -> SSL -> Enable SSL Proxying
- Locations mit den entsprechenden Domaisn füllen
- Help -> Install Charles CA SSL Certificate -> Details -> In Datei kopieren -> Per Mail an iPhone senden und installieren
Vorsicht, besser nicht zu sehr bei der Installation von Apps benutzen, ich war heute 2 Stunden vom AppStore ausgeschlossen 😉
Mit dem Setup kommt man schon recht weit und kann doch so einiges lernen.
Kurze Analyse von Forge of Empires
Published 2012-03-29, 15:30
InnoGames hat mit Forge of Empires ein neues Browser-Aufbaustrategie-Spiel veröffentlicht. Logischerweise habe ich mir das ein wenig genauer angeschaut:
Technik
- Flash-Client!
- Game-Seite sauberer Quellcode
- Nur inline CSS
- Combined JS
- Enthalten
- http://code.google.com/p/swfobject/
- http://swffit.millermedeiros.com/
- http://jquery.com/
- z.B. swfobject oder jQuery stehen unter MIT Lizenz:
[…] The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. […]
Die fehle allerdings überall.
- Enthalten
- Inline JS für eigenen Code
- Google Analytics
- Preloader lädt eine lang.mo, mit POEdit „dekompilierbar“ zu Quell-po-Datei
- geladene Datei aus Verzeichnis en_DK
- 1258 lokalisierte Strings
- Unterschiede Zwischen Vorlage und Übersetzung, obwohl beide EN
- Server unter /game/amf
- Selbe Domain wie Hauptseite, also auch Analytics-Cookies bei jedem AMF-Request
- Animationen und andere Assets als PNG(-Sprites)
- Manche Assets mit Reset-Cache-Query, andere ohne (siehe Liste vorne dran)
- Verschiedene SWF-Files je nach Zeitalter, vermutlich Interface, auf jeden Fall Musik
- Soundeffekte als MP3
- Musik aus verschiedenen Fetzen (dynamisch?) zusammengesetzt
- Server:
- Server: nginx/1.0.10
- X-Powered-By: PHP/5.3.10-1~dotdeb.1
- Tutorial-Content wird als XML-File ausgeliefert
- http://en1.forgeofempires.com/data/tutorial.xml
- Code ist nicht valide wegen kaputtem Comment 😉
- Tolle Payment-Einbindung
- Von Flash erzeugter Screenshot als Hintergrundbild
- Vordergrund Iframe der Payment-Seite
Allgemein
- Nette, ruhige Hintergrundmusik
- Per Default eine Menge Freunde
- Farmville meets Die Siedler II
- Kasten in Kasten in Kasten
- Dauer-PVP
- Support
- Plündern
- Highscorenachbarn statt Freundesleiste
- Keine Freundesliste
- Tutorial sehr restriktiv
- Event History bei Login ist ein super Feature
- Monetarisierung
- Produktion
- Instant abernten
- Instant fertig
- Doppelte Produktion
- Militär
- Personenslots bei Militärgebäuden
- Negotiation (kampfloser Gewinn) ohne verfügbare Gegenstände
- Reduce Training time
- Units heilen
- Instant scouting
- Gebiet
- Verzeitige Expansion
- Forschung
- Forschungspunkte
- Produktion
Performance-Optimierung am Beispiel: betamode.de
Published 2012-03-22, 20:07
Ich habe mich ein wenig mit Performance Optimierung von Webseiten beschäftigt. In den nächsten Tagen und Wochen werde ich einige Beiträge dazu veröffentlichen. Den Anfang macht diese Beschreibung des Optimierungsvorgangs von betamode.de:
Beispielseiten für Optimierung
- http://betamode.de
- http://betamode.de/2011/07/31/server-data-monitoring-graphing/
Ausgangssituation
- http://www.webpagetest.org/result/120318_NC_3MJ57/
- http://www.webpagetest.org/result/120318_GB_3MJ59/
Erste Beobachtungen und Auswertung
- urchin.js – uralter Google Analytics Code?
- Mehrere kleine GIFs, sollten sich einfach spriten lassen
- First Byte Time ist sehr hoch
- Kein Caching der statischen Ressourcen
- Kein CDN
Optimierung
Anwendbare Regeln aus High Performance Website Sites:
Rule 1 – Make Fewer HTTP Requests
Folgende Bilder können zusammengefasst werden:
- http://betamode.de/wp-content/themes/betamode/images/suche.gif
- http://betamode.de/wp-content/themes/betamode/images/clock.gif
- http://betamode.de/wp-content/themes/betamode/images/comments.gif
- http://betamode.de/wp-content/themes/betamode/images/xml.gif
Schritte:
- Bilder von <img> auf background-images umbauen
- Bilder als Sprite zusammenfassen: http://betamode.de/wp-content/themes/betamode/images/sprite.gif
- Sprite als background-image einbauen
- Da sich bg.gif nicht mitspriten lässt wird es als data-base64 direkt ins CSS eingebunden
Ergebnis
- http://www.webpagetest.org/result/120318_NM_3MJST/
- http://www.webpagetest.org/result/120318_C2_3MJSV/
Rule 3 – Add an Expires Header
Mit W3 Total Cache lässt sich sich das wunderbar erschlagen.
Ergebnis
- http://www.webpagetest.org/result/120318_0K_3MJY7/
- http://www.webpagetest.org/result/120318_4C_3MJZ0/
Rule 6 – Put Scripts at the Bottom
Das Analytics-Script ist zwar schon ganz unten, aber noch das alte synchrone. Also mal das neue besorgen und einbauen.
Ergebnis
- http://www.webpagetest.org/result/120318_B3_3MJZY/
- http://www.webpagetest.org/result/120318_TS_3MJZZ/
Rule 10 – Minify JavaScript (+ HTML + CSS)
Javascript gibt es keines, aber für Minify von HTML und CSS kann einfach W3 Total Cache konfiguriert werden, fertig.
Ergebnis
- http://www.webpagetest.org/result/120318_C4_3MK1C/
- http://www.webpagetest.org/result/120318_6G_3MK1D/
Dann zum Abschluss noch ein bisschen serverseitiges Caching um DB-Anfragen und so weiter zu minimieren mit W3 Total Cache.
End-Ergebnis
- http://www.webpagetest.org/result/120318_57_3MK1Z/
- http://www.webpagetest.org/result/120318_XM_3MK22/
Tadaa, schnell.
Wireshark: Copy content of compressed requests
Published 2012-03-07, 17:40
Copying the content of a request compressed by using gzip is not easy in Wireshark:
- Click the request you want the content from
- In the Middle panel, click „Packet Details“
- Collapse all branches
- Right click on last Branch (Could be „Line-based text data: text/html“ or „Media Type“)
- Click „Copy“
- Click „Bytes (Printable Text Only)“
Now you have the content in your clipboard and paste it where you want.
Anwendungen und Tools zum Monitoring und Graphing von Serverdaten (Load, Memory, Traffic)
Published 2011-07-31, 18:19
SAAS:
- http://www.serverdensity.com/
- http://newrelic.com/
- http://www.appfirst.com/
- https://www.bijk.com/
Self-hosted:
- http://munin-monitoring.org/
- http://www.cacti.net/
- http://www.eluna.org/eluna_graph_system.html
- http://www.monitorix.org/
- http://www.prbproject.org/
- http://serverstats.berlios.de/
Ergänzungen willkommen.
Fix ugly fonts in Firefox 5
Published 2011-06-27, 09:40
- about:config
gfx.direct2d.disabled
auf true setzen
Secure XAMPP: only local connections
Published 2011-03-25, 12:23
1) Open file xampp\apache\conf\httpd.conf and replace
Listen 80
with
Listen 127.0.0.1:80
2) Open file xampp\apache\conf\extra\httpd-ssl.conf and replace
Listen 443
with
Listen 127.0.0.1:443
3) Open file xampp\mysql\bin\my.cnf|.ini and put in
bind-address=localhost
directly after [mysqld] (NOT [mysql]!)
Dual-boot a hackintosh with Windows 7
Published 2010-07-07, 10:26
http://thebackpackr.com/hackintoshing-with-snow-leopard/
http://www.chip.de/news/Microsoft-Tool-Windows-7-vom-USB-Stick-installieren_38622482.html