[sebsauvage]

Looks nice, but again...:

  * another closed source service ?
  * Are there import/export options ?
  * There is not even a RSS feed ?

Sorry for looking like parasiting a topic, but I need to rant: I'm sick of close-source online bookmarking services, especially when I can't import/export all my data. So I'm moving away from them. I want to have control. This my data.

I developed my own bookmarking application, Shaarli (opensource, bookmarklet, html import/export, private links, tag cloud, picture wall, "daily" digest and many other features). http://sebsauvage.net/wiki/doku.php?id=php:shaarli

My own instance is: http://sebsauvage.net/links/

I don't say hosted services are bad, but I'm definitely done with them.

[iamthedata]

Gratz, Shaarli is pretty cool. Diigo is quite decent, but now I rather run Shaarli on localhost. Archival feature would be great, even if it just shells out to httrack. If used on a local machine, the disk space is not a great concern.

With 0.0.40 beta, I was having the problem about "Wrong login/password" and "Wrong token". It would be great if you could apply the fix described by this user:

lgaggini, 2012/11/10 18:59, http://sebsauvage.net/wiki/doku.php?id=php:shaarli:discussio...

> I host last beta of shaarli (0.0.40) on my local network and I access it by custom name defined in clients hosts file and relative virtual host configuration in my nginx web server. My browser is Chromium (22.0.1229.94 (161065)). After the first configuration, every time I try to login I receive the Wrong login/pass error. After some investigation and debugging I discover a problem with cookie management (only in Chromium, in Firefox all work flawlessly). With the default code and Chromium, the cookie is correctly written by the browser, but it's not retrieved so all checks performed on sessions value (mainly tokenOK() and isLoggedIn()) fail. To fix I have to replace, in the main code index.php, the session cookie path parameter, from dirname($_SERVER[“SCRIPT_NAME”]).'/' to dirname($_SERVER[“SCRIPT_NAME”]) (short story, i remove the trailing /). I hope to help someone else with similar problem. Regards and keep up the good work.

[sebsauvage]

Thanks :-)

Oh, and thanks for the bug report regarding the cookie path. I did not pinpoint this behaviour before. (And I currently lack time to update Shaarli.)

For the archival feature, you use the export option (in tools menu), then import it later to restore. Or you can backup all files located in the "data" subdirectory.

Shaarli grew too fast and I did not expect to have so many people interested in it (This is a typical scratch-an-itch project). I plan a complete rewrite, with multi-users support and many other features (I have a heavy to-do/feature request list :-)

[StavrosK]

How many services are there where you can't import/export your data? I run one (http://historio.us/) and import/export has been available from day 1, it's pretty much the most requested feature after actual storage of bookmarks.

[sebsauvage]

I used to be a delicious user. Their export was broken for several months, and they did not show signs of wanting to fix it. I had to write a Python script to backup my data by parsing their HTML pages.

Regarding Diigo, their Firefox extension sends every visisted URL to their server. Thanks for the respect of my privacy.

Enough is enough.

[booruguru]

RSS functionality and exporting will be available in a little while.

I wanted users to be able to import Delicious links upon sign-up but my developer ran into some trouble with the Delicious API limits. So I'm looking into alternatives.

(Perhaps I'll need to get users to export their Delicious links and upload them to Klipbook. -- If anybody have any suggestions or alternatives, I'd love to hear them.)

[sebsauvage]

Good to know you're working on it.

In my experience, Netscape bookmarks html export worked better. But even then I experienced a lot of broken exports from online bookmarking services. This is a pain. Good luck with that.