Keeping track of my CCed photos

Man gönnt sich ja sonst nichts [TM]Since I’ve been more active again on flickr lately I was interested if any photos of mine have been used in the web. I release my photos under the Attribution-ShareAlike Creative Commons License. This means that people can take my photos, alter them, use them (even commercially) as long as they refer to me and use the same license. I guess some people just don’t care about the license of their photos while others don’t think that anyone would like to use theirs, but I did some research (i.e. googling) and turned up quite a few web pages which are in no way affiliated with me (I didn’t know them before or their authors) but still use my images. How they found those? Not sure, but I guess using a combination of “search-for-cced-images” and the appropriate tags. Here is the list:

Update: Aug 24, 2011

Updates until Feb 6, 2011

Some of these sites are private bloggers who look for a picture to go with their posts. Others however are professional websites. This is a way of showing off I guess, but it also shows that people do care about free photos. Now, there might be more websites with my pictures which I haven’t found, maybe not everyone links back to my photostream. If you do use my photos I’d really appreciate a flickr-comment, it makes me happy to see my motives on other websites ;). (Someone even contacted me once about using one of my photographs in a book about students, but for some reason didn’t do so after all.)

This list is also a reminder for me, and I might make it into a permanent page at some point.

monit and init

MonitMonit is a daemon that looks simple enough and does one thing really good: Making sure services stay up (or: Making sure you know when they don’t). It does some other things as well (like checking up on files and system resources), but I just use it for important processes which could crash (or be killed). Now, I’m not gonna go into how to configure monit since it is quite easy and the config-file can be understood by simply looking at it. You can start monit like any other service via scripts in /etc/init.d and symlinks from the appropriate runlevel-directories. The question that immediatly comes to mind is what would happen if monit itself crashed (Hint: Well, nothing, it would stay down and you wouldn’t even receive an email ;). So, instead of using some home-brewed solution with cronttab that checks monit, I’d suggest just invoking it with init. To do that just append a line like this to your /etc/inittab:mo:2345:respawn:/usr/local/sbin/monit -Ic /etc/monitrcThe -I switch means it runs interactively, which is necessary for processes started by init. To initialize it without rebooting try telinit Q which should already spawn the process. If you kill the monit process now it will be respawned automatically. Voila! Oh, and if you ever want to stop some monitored service (which monit would restart) you can just unmonitor it from the webinterface.

So I’d suggest you give monit a try, even if you think you already have an appropriate solution. Don’t be scared by the looks of the monit-website. And if you think monit doesn’t do everything you need you’re bound to notice it after 15 minutes max, which imho isn’t much time wasted.

github: first impressions

GitHubJust as I wanted to start writing this article I noticed that github is no longer beta. So, you might as well just go there and try it for yourself.

Now, about github. github is a public git-hosting-site. Put short, it’s like repo.or.cz, just with a shiny interface and a few special features. To be honest, I am not the perfect candidate for a service like this, since I don’t have any public repositories, I don’t know any people I’d like to cooperate with and I have my own account on a server where I run my private git repos. Also I often react allergic when I see something that already looks like another one of those Web2.0-mint-colors-big-font websites that offer some kind of paid plans just so that you can have a textbox in your webbrowser which others can edit too…

Since I spent way too much time writing full text in my last article I’m just gonna do some lists now :P

github

The good

  • It uses git and ssh. I know, i know. Let’s just say I’m happy that more and more people (like the Rails-crowd) are coming to their minds and use git.
  • It has a free plan for public repositories which is 100mb (which is a lot for a bare git-repo).
  • For beginners it’s probably a lot easier than repo.or.cz or setting up of an own server.
  • It has private repositories, which repo.or.cz does not offer.
  • It is built like a social network, which means it supports sending mails, pull-request and the like directly inline.

The bad

  • Right now it’s pretty slow.
  • It still feels like it’s missing some features (which I can’t really pin down since I haven’t used it extensively)
  • When giving instructions to connect to their server they didn’t mention the fingerprint of their SSH-key anywhere.

The ugly

  • The theme (which atm can’t be changed) sucks to high heavens. The important things (i.e. the list of files and the commit messages) are barely readable right now, and even less so under worse lighting conditions.
  • It still need some options to view the history and search it with some live-search-box

To sum it up: It’s a step in the right direction. It enables people to have their git-repositories quite fast and without hassle. Maybe the web-ui features aren’t even that important since with git you always get the full repository and can perform all history-related queries (like diffing) on your own machine with the weapon of your choice anyway. I’m really interested in how many people take advantage of the paid plans and if there will be more sites switching to/support git hosting. Having said that, I don’t think I will use it very much for the reasons mentioned before. Unless I start some public project… ;)

git for quick’n’dirty editing

gitYou can use git as a full-fledged, commit-driven VCS, with all its awesome features. I wrote about that (and probably will again) before. But often you just have to track some things over the course of a few minutes/hours and don’t even want a “repository” or a commit-history or branches.

Often you have some code/text in it’s original form and need to change a few lines until something works (again). Sometimes you’d like to keep track of these changes (for example to apply them on a different codebase/host). Before git, I just copied the files (or directory) I wanted to edit to some or destination and later did a diff with the new files. That sucks for obvious reasons.

With git it can be much more convenient: You go into the directory with the vanilla files in it. Just do a git init and git add .. Thats pretty much it, now you just go on to edit whatever you want. To see the changes you made since the first git add . just type git diff (I don’t have to tell you that you can pipe this output into a patch).

If you happen to add new files, I usually commit the original version using git commit -m ‘initial’ (even later on, since you haven’t staged any of your new changes for committing yet), and then do a git add -u and git add <new -files> and then diff with git diff –cached. If there is a better way please let me know!

That, imho, is one of the examples where git beats svn hands-down. The notion of a local repository (or in this case the index) is just great once you understood the concept somewhat. I wouldn’t use svn for little changes since it would mean setting up an svn-repo (in a different dir), checking it out, commit the vanilla version, changing stuff and then diffing it. And then you’d have all those .svn directories cluttering your working tree, whereas with git you can just do a rm -rf .git at the end of the day and forget that you even used git.

Something else I missed in my svn days was committing only some changes (called “hunks” in git) to a file. In git you can use git add –patch or git-gui to do just that and avoid the problem of having multiple unrelated changes in a single commit (which sucks when debbuging). I read a really good article (and comparison with other SCMs) today at http://tomayko.com/writings/the-thing-about-git.

subscribe via RSS