My vmware referal code blog seems to be number one on Google when searching for "VMWare 5 referral discount code". Someone found this and used it... which mean I get a T-shirt! w00t! :P

The code is: VMRC-MARVAN882

This referral code will give you a 5% discount to a new or upgraded VMWare product! (VMWare 5, ESX, GSX server)

Looking at the following:

1. Firefox 1.0
2. Safari 1.2.4
3. Camino 0.8+
4. Opera 7.54
6. IE 5.2.3

Evaluated Features:

1. Site Compatibility

    5 - Firefox
    4 - Safari
    3 - Opera
    2 - Camino
    1 - IE

    Not any browser works with all sites, and
    every one has it's place. With firefox and
    Safari I can do most things.

2. Tabbed Browsing

  Safari and Camino supports the best middle
  button -> Open in Background TAB options.
  Firefox seems to have broken middle button
  support in OS X, unless you load funky mouse
  drivers (USB Overdrive). Opera can do vertical
  tabs (cool!) and is overall the best here.
  Firefox has a cool configurable extension and
  may even be better than Opera. IE just plain
  sucks.

3. Speed + Footprint

  Camino is the fastest and smallest.
  Firefox is very fast with bigger footprint.
  Safari is slow and large.
  Opera loads the slowest on certain sites.
  IE is the slowest.

4. Integration with OS X

  Safari, Safari, Safari. (obviously)
  It slices and dices...

  Seriously, It supports:

    - System-wide keychain access
    - System proxy settings integration
    - System Spell-checking in input areas (!!!)
    - Good mouse/keyboard handling

  Next best is Camino, followed by Firefox.
  Opera is OK and IE is way last.

5. Mouse features + gestures

  Nothing comes close to Opera mouse navigation.
  With that Said, Firefox can get plugin extensions
  to do most of what Opera touts, but I find this a
  pain to install, and it never quite works like
  Opera. Safri and Camino follow closely and IE
  trails.

6. Application integration and association.

  Safari does not pass .DMG files to an external
  iGetter download manager... this REALY sucks.
  XML viewing is also sporadic...
  anyone shed more light on this can mail me...

My Final Impressions
--------------------

I'll stick with Safari as long as Firefox does
not support proper proxy/keychain/speller
integration. I've gotten the Firefox mouse
handling to be more in line with Firefox on win32
and Camino and Safari on OS X at least.

I'm running both Safari and Firefox at the
moment... I used to run Camino, Safari and IE.

On windows I prefer Firefox above all else...
I hardly ever use IE on any platform. Ever.

Why is it that I don't fear a world where there are only OS X/Apple computers?

Perhaps it's because it's so damn good? And Tiger too seems to be so packed with goodness.

I for one welcome our new Apple overlords.

I've never felt this way with Windows, Linux, OS/2, AmigaOS. Strange... that's funny... oh well.

Having been playing Quake 3 competitively for some time I have found cordless mice to have many problems, some of which can be overcome.

The biggest problem is the fact that most cordless mice (Logitech and MS) had a small 'glitch' in transmitting movement information when a button gets pressed. This becomes quite obvious in quake with the left mouse button bound to jump and when you try and do any movement tricks such as circle jumping and strafe jumping.
This can be solved to a large degree by switching jump to the keyboard (spacebar).
Then there is the problem of losing your connection and battery problems. But by far is the limitation AFAIK that all cordless mice have a sampling rate of only 50Hz (wireless) / 85Hz (bluetooth).

Secondly you need a mouse sampling rate which is at least equal to your monitor refresh (and also your in-game screen update fps) For most games this would be 80fps (For those who think 25-30 or 60 is fine... read up on this at your own leisure) The max I find to be useful is 120 Hz video refresh with at least that in the game FPS and mouse sampling rate. PS2 mice can be tweaked in win2000/XP to be 200 and USB mice are mosly around 125 and cannot be tweaked. (I run everything at around or above 120 if possible).

Then there is also your mouse DPI...
In this field the Boomslang mice range was pretty good, but they were mechanical and the ball sometimes got stuck. For optical the new Boomslang mice are superior, but the Logitech are adequate when it comes to DPI, especially the Dual Optical or MX range, being around 800dpi per sensor for most high-end ones. (The only thing going for MS mice are that they are cheap) The boomslang optical is around 1400 DPI as far as I can remember. (It is also said that the Razer (boomslang) mouse drivers work for the Logitech Dual optical and MX series with only a minor tweak in the INF file, and are superior in most respects.)

An important point to note at this stage is that the Logitech mice are not really 800 DPI, and it's only due to a tweak in the drivers that a 800-like performance is reached.

DPI makes a huge difference whether you are a low or high sens (sensitivity) player. Most people prefer playing with low or no mouse accelleration, to improve accuracy and high sens or with acceleration for movement accuracy. A big problem with MS mice is that a rapid movement to the side will cause the mouse to lose tracking, and may very well point you up and down! This is fatal and irritating. Older Logitech mice tend to just lose tracking and stop moving. This is still acceptable. The newer logitech mice can handle a much greater movement rate without fail. A huge bonus.

This last point is the third major criteria: tracking.
This is measured in deviation/m/s... from what I can tell the new Logitech MX series version X1X seem to do well. The best being the MX 510 in my opinion. Microsoft does not even bother providing these details, and thus points out their lack of focus in this regard.

About the MX1000 Lazer Mouse:
A friend of mine bought one yesterday, and I tried it out.

The cool thing about this mouse is it's AMAZING tracking... It's a bull%^&* argument from Logitech that it's for CAD and businessmen only! What do they ever need to move it at great speeds for?

They say it was not designed for games... I say they designed it NOT for Games. I don't see why they would have put this 'feature' in.

Oh, yes... it's not the lazer powersaving etc... on a B/W IR camera you can see the Lazer only going into powersave more after 15 seconds!

I've got an MX510 and so far it's the best all-round mouse, and even it does not compare to the exquisite tracking (meters/second) of this lazer mouse. Firstly you need all excelleration OFF, then in a game, more it FAST to one side, and then back the exact same point... and voilla... you are there! What executive needs this!!!?

I'm a low sens player, which mean that my movement is flowing and I pack an average of 45% rail accuracy in Quake... this mouse if really good with accuracy, but if they just left the sensor on while picked up, then it would make moving actuallt plausible... since like most low sens players I use a huge mouse pad, and pick up my mouse constantly which makes this mouse unusable... all they needed to do was to keep the sensor on for a bit longer!!!

I then tried playing it as high-sense, and it does become a bit easier to use, since most high sens players hardly ever need to lift the mouse... it becomes bareable in this mode, but still not good enough.

I'm sure this mouse should be 'hackable' with some solder etc.... which would then make it the ultimate... I love the fact that they made it work at 125 unlike ALL the other corless mice.

PS: On another topic, the whole mirror thing. The mouse is obviously outside the visible spectrum, and thus the 'mirror' may not be as mirror-like at those frequencies and may or may not work... hence the Box disclaimer... Just my 2c

Any comments welcome. As a low sens player I may not have encountered most bugs.

Some stats on my Quake (for those who play seriously)... Can complete any Quake 3 level on Nightware (Xaero is a problem but can be beaten) Have a 60% rocket and 40% rail accuracy against medium real players(OSP/RA). Can do rocket launcer platform to rail gun platform jump (no rocket/plasma jumping used) on Q3DM6.
(But I suck at rocket jumping since I switched to space-bar jumping and low sens does not help...)

PS: Settings in quake to try:
  in_mouse 1/-1 (then in_restart)
  com_maxfps 125
  r_displayrefresh (then vid_restart)
  m_filter 0/1
  sensitivity (I use 3-8 depending on OS)
  r_mouseAccel 0/0.5 (depends on 1v1 or FFA)

RFC: marius-audio-rip-rfc0001
Title: Treatise on audio ripping standards.
Version: 1.2

After much changing and experience I think I'm settling on the best way to
organise MP3/OGG files on my HD.

These are the conclusions:

1) There MUST always be ID3 tag information in all files.

    Many programs use these, when filenames just don't cut it.
    Make sure these fields contain the correct information before ripping the
    CD as, even though you may correct the filenames afterwards, the Tag info
    will still be wrong. (And one can always quickly re-create the filename
    from the tag info with handy utils)

    It is preferred to have ID3-V2 tag info, as opposed to ID3-V1.
    (Don't add any legacy ID3-V1 info unless ID3-V2 is not available)

2) Make 100% sure spelling is correct.

    I never knew someone could misspell "Alanis Morissette" in 5 different
    ways!

3) The directory hirarchy: /Artist/Album/...

    This is a good way of grouping files, and helps if you want to be able to
    manipulate by artist or album. Since Winamp can play entire directories
    when it is dragged and dropped on it, this actually has a purpose!
    It is important (more so with Artist names than Album names) to put all
    'The' and 'A' words at the back, eg: "Cure, The". There are usually many
    more Artists, than there are albums, and it helps with the searching.
    Another issue is using abbreviated forms; to keep consistency, all '&'
    should be 'and', as well as any other abbreviations should also be expanded.

4) The filename:
        SongName - Artist.mp3

    There are some variations on this, but they have proved to be inadequate.
    It is useless to encode track numbers into song names since eventually
    people split songs up, in which case the song track number has no
    purpose. There are anyway better ways to order songs. (see next point)
    The reason why SongName comes first, is because it is the most important
    part of the filename. It happens that filenames sometimes get mangled
    (and not just 8.3 file name mangling... 30 char max on iso9660),
    and it is always the last part that gets lost. Also when displaying
    filenames, most programs never have enough space to display the full
    name.

5) Playlist (per album)
              Album - Artist.m3u
      (or Album - Artist.pls, depending on taste - it's all the same)

    This is the best way to keep album order as it is seperate from the
    filename, which can get mangled etc etc.
    The order is important. (see reason in 4)

    The placement of the playlist should be either on the same direcory
    level as the album or the artist directory.
    It's more easily accessible that way.

6) Encoding.

    Encoding should always be at least 128kbps @ 44kHz. Variable bitrate
    encoding is encouraged, since you can effectively get 256 bitrate at
    very little extra size increase. Make sure files did not skip...
    there's nothing as irritating as having a perfect complete collection
    of an artist, except for that one song.... grrr.

    Joint stereo encoding is slower, but could make files much smaller at
    no extra loss.

    NOTE: MP3 files are surpassed by OGG (Ogg Vorbis) w.r.t. Quality,
                but at the moment it is not yet mature and encoding is still
                quite slow. It is, however, a better though-out standard, and
                has a better and cleaner future. AAC is also a better alternative.

Anyway, this is a RFC, so please C.
Domini