fragility

I’ve been using Ubuntu Hardy Heron since I got my new laptop (ended up getting a refund for the CR-21 and got a CR-31) but when I tried to start Firefox 3 it complained that the majority of my extensions weren’t FF3 compatible. I didn’t really had the time to investigate the issue so I installed Firefox 2 (which is still available on Hardy’s repositories) and kept using it. However, I’ve been reading so many blog posts and tweets on how Firefox 3 is blazingly fast that I had to try it. And it is. It’s stupidly fast. I can’t even believe this is really Firefox. As for memory consumption, it’s too early to tell (I’ve only been using it for a few hours now) but so far seems better.

By the way, regarding the new laptop, I’m keeping a wiki page about it which I’ll be updating with more details later.

So, after searching for a while I found out that some of the extensions I use have beta versions which are already compatible with FF3, so I decided to give them a try. Here’s the list of extensions which are not FF3 compatible and how I managed to install them or the alternatives I’m using.

Adblock
Not compatible with FF3 but it seems kind of unmaintained anyway, so I installed AdBlockPlus instead (which seems like what everyone’s been using for some time anyway).

del.icio.us Bookmarks
According to this discussion the extension is ready but it only works over the next delicious version which has not yet been released to the public. I’ll be following the news on the delicious blog closely.

I ended up using a hack which bypasses the check for extension compatibility, but if this turns out to cause any instability I’ll use the bookmarklet.

Firebug
Had to install 1.1 beta from http://getfirebug.com/releases/.

FireFTP
Had to install the latest preview from http://fireftp.mozdev.org/developers.html.

FireGPG
Had to install the latest svn. More info on http://firegpg.tuxfamily.org/?page=install&lang=en.

Gmail Manager
Supposedly there’s a FF3 compatible version on this discussion thread but this is probably just a hacked version of the regular extension and FF3 crashes on start. This is a handy extension but I don’t consider it essential so I don’t really mind not having it around for a while.

Google Gears
Doesn’t seem to be compatible with Firefox 3 yet, but I don’t really use it much these days. Maybe I’ll drop it for now. The hack I talked about before lets the extension load but I prefer to disable it for now to avoid any issues.

Google Toolbar
Doesn’t seem to work either but I only used it to check on the Pagerank so I guess I’ll drop it too.

TinyURL Creator
No compatible version available, but works with the compatibility bypass. There’s also an alternative bookmarklet.

User Agent Switcher and Web Developer
Apparently, the developer of these two extensions refused to release Firefox 3 compatible versions until the Firefox 3 final version is out, but the Firefox developers are asking for add-on creators to update their add-ons to check for any issues so the creator says he’ll be releasing a test version soon. I just hope he hurries because this is has always been one of my favorite extensions and I’m sure lots of webdevelopers rely on this extension as a part of their work. Oh and these two don’t load even with the compatibility bypass.

Can I have the possibility of customizing the top menu (the one with the links to other services) which appears on every Google service ?

kthxbye !

Couldn’t agree more.

Besides, why do Google sharing pages (Reader’s and now this new one) always have such long and complicated URL’s ? Wouldn’t it be easier to have a personalized URL like…Bloglines ? del.icio.us ? Almost everything else ?

I find it amazing how they can come up with such great stuff and just ignore simple things like these.

With yesterday’s launch of Google Desktop for GNU/Linux, I decided to take a look at the desktop search options again. I tried Beagle on the past, but I always found it too slow and ram consuming, and it never seemed to index stuff right. So here’s my new comparison of three desktop search options for GNU/Linux. If you know about any other options, I’d really like to know about it.

Beagle

I hate it. It’s still slow as hell, eats up a lot of ram, takes a long time to index stuff the first time (but it’s not intrusive because it only indexes while the processor is idle) and even when used together with deskbar it’s somehow stupid, because you have to type in the result, select the “search with beagle” option and wait for the beagle application to open and search it for you. I have to admit that I didn’t even let it index everything before I uninstalled it. Some guys I know would say “oh…it’s Mono !”. I’m not really an anti-Mono fanboy (I love Tomboy) but I don’t think Mono is the kind of platform which should be used to develop this kind of app (at least, not the daemon). But then again, what do I know ?

Tracker

Extremely fast indexing stuff on the first time, but in order to be that fast it uses your processor a lot, which can become intrusive, so you’d better have it index your stuff over night on the first time. Has a low ram consumption, it’s fast, and has a great integration with deskbar (gives you live results within deskbar, just like spotlight on the mac).
By the way, thanks a lot to Filipe Carvalho for writing about it. I had heard about Tracker before, but completely forgot its name, and I didn’t try it at the time because the project was still very young.

Google Desktop

Looks good and it’s easy to setup. Takes a long time on the first index (but like Beagle it’s not intrusive). The keyboard shortcut (hitting Ctrl twice) is not configurable and can become quite annoying, specially if you’re a keyboard shortcut freak like me. I use Ctrl+Alt+Directional Keys for desktop change, and sometimes Google Desktop just comes up unexpectedly. Also, the fact that it only shows a few results and in order to see more you need it to open a browser page, is not very practical.

Considering these three options, my favorite is undoubtedly Tracker. It’s fast, light and has a perfect integration with my Gnome desktop.

Update:
Looks like there’s a way to integrate live Beagle searches into deskbar as well (thanks to those who pointed that out on the comments). That’s not the biggest hassle I have with Beagle anyway, so I’ll still keep Tracker. Other thing I noticed and forgot to mention: Tracker eats less disk space than Google Desktop.

Gears is beta indeed. Usually, most of the “beta” stuff that comes out of Google is relatively stable, and most people trust it really easily, but be careful with Gears in this early phase.

As I’ve mentioned yesterday, I decided to try Google Reader again. I had also installed Google Gears some days ago, when it appeared. I haven’t really been using the offline feature on Reader, but last night something happened: I couldn’t see anything on the Reader page except for the logo, and the top right link bar. I thought “hey, this is beta after all” and I went to sleep.

This morning I tried Konqueror. It worked. I asked other people who were also using Firefox. It worked for them. And then I noticed: Firebug gave me a JavaScript error. I tried disabling Firebug, same thing. I banged my head against the head for a while until it hit me: Gears ! It had to be Gears. And it was Gears.

Not that I’m pissed by it. I don’t use Gears anyway, just installed it for the curiosity. And yes, I know, both products are beta and I shouldn’t be complaining. I’m not. I’m just warning other potential users of this combination.

By the way, I’m loving Google Reader. One huge advantage I’m noticing over Bloglines: no repeated items. Bloglines has a nasty habit of showing you old items as new, even when they have not been updated.

Reader has really come a long way since I first tried it. Not that I’m a Google fanboy (I’m probably starting to act like one sometimes) but I was expecting a lot more when it first showed up. I guess it just needed some time to mature.

Update: Nice to see the guys at Google are paying attention.

I decided to give another try to Google Reader. I use Bloglines, but lately I’ve heard so much about Google Reader and how much it improved that I just let myself get taken by the hype.

One thing i really like: on Bloglines, when you are viewing a folder (therefore viewing more than one feed on the same list) different feeds get separated by a title with the name and description of the feed. If you use keyboard shortcuts to go through feeds, most of the times you can’t even see those separators because it just scrolls down to the next item, and once in a while you get lost and have to scroll up a bit to so you know what you’re reading. On Google Reader you get a “from Feed Name” on each item, and the item you’re currently reading is highlighted.

The fact that things like embedded Youtube videos appear on the reader is really nice. Saves you a click and reloading that same content (and some more from the rest of the page layout).

The way tags are used in Google Reader is somehow confusing. Because tags and folders are the same thing. Therefore, it’s confusing if you have a feed on more than one folder.

So I’ll just try it for some days, weigh the pros and cons compared to Bloglines, and maybe I’ll report back on this with some more thoughts.

When you google for “define: something”, links for Wikipedia *most of the times* appear at the end of the definitions list. Coincidence ? I don’t think so. Is this their way of saying “trust Wikipedia only as a last resource” ?

Update:
s/always/most of the times/
Someone just proved me wrong :)