On 10/2/06, Brion Vibber <brion(a)pobox.com> wrote:
Cortado (which is used in Michael's work) is
claimed to work on the old MS VM.
If you have access to a computer with this limitation, can you please test and
confirm?
Last I checked it worked without sound in javasound less VMs. I don't
personally have access to anything with that VM
2) As a
contributor I would be upset that users are being given a poor
quality version of my contribution rather than being asked to install
improved software.
Well, that's pretty silly.
I don't think so.
I've been aggravated, and I think rightfully so, when someone
complained about an image I spent hours getting right being low
quality when it really was just an artifact of our image handling.
This happens all the time due JPEG artifacts, which are often visible
even on the image page depending on the users's display... or the
image appearing being blurry because straight filtered downsampling
can be unflattering.
It is perfectly reasonable to adapt to the users environment to an
extent, but it can be carried too far: For example, we could 'help'
uses with browsers without image support by converting images to text
using [[AAlib]].
I think t we'd be helping users more by spending our efforts on:
1) Having a really good set of instructions for one click installation
of support software.
2) Encouraging browser vendors to ship support with their browsers.
(The reference Theora and Vorbis codecs are BSD licensed, so there
shouldn't be too much of a license-political issue there).
... instead of dishing out a hacked up player that will have a much
lower audio quality on their system than the flash based sites.
Some food for thought:
Latest stats (from the last days logs) from the toolserver java audio
player (linked from enwiki and other places):
27,828 unique IPs loaded the PHP page
23,327 unique IPs loaded the JAR 16% have no java enabled.
17,789 unique IPs loaded an Ogg 19% have Java but it failed
Overall it worked for about 64% of the IPs.
As far as I know, lack of javasound API (i.e. old MS JVM) is the only
significant cause of loaded but failed right now.
These numbers are a bit different from when I first introduced the
player.. Perhaps client side caching is influencing the results. ...
It also could be because folks without Java have either installed it,
or they have learned not to try the player.
Hm. The above numbers to make a greater argument for extending support
to uses with the old JVM.