"ARGH! And now we're going to build another half-step child of Mozilla? Like the world needs _THAT_?"
Think of it as Mozilla forking, but the fork being blessed by mozilla.org. It was clear in the runup to 1.0 that there were tensions within the project; running forks internally keeps the developers happy and interested, and produces interesting things that can be adapted to the main trunk.
(Or, as is their goal, to replace the main trunk.)
"Umm why download nightly builds of a usable, stable application?"
Well, the version number in this case is accurate: this is an 0.2 and will act like one from time to time. You can actually expect noticeable changes from day to day.
Beating on nightlys gives immediate feedback on the effects of changes made that day - catch serious bugs early. Being a tester is a way to contribute greatly to a project as Joe User. And if there's a bug that's really been annoying you, you can get the fix straight away instead of having to wait until the next full release.
I think Phoenix is doing it this way because that's how Mozilla does it - and it works very well for Mozilla - and therefore because they can (being in the Mozilla build system).
"-Miracleman is an old character published by some of the big companies. In the 80s two writers took a swing at him-- Alan Moore and Neil Gaiman-- and created some very, very good work out of it, stories whose reputations have only grown in time. It was work for hire, though, and so reprint rights remained with DC Comics."
Not with DC. It's much more complicated than that. Check the book Kimota! - covers the entire saga.
I assume you've read Kimota! (a book of interviews with the creators and editors of Marvelman/Miracleman) on the subject?
The most likely guess is that when Dez Skinn revived Marvelman/Miracleman in 1981, the rights had lapsed into the public domain. However, after having published the character for a while, he had then created interest in it. And, name and costume aside, the Alan Moore Marvelman/Miracleman was quite different to the Mick Anglo Marvelman.
Kimota! gives the impression that Gaiman would like the series to continue. Moore wouldn't mind seeing it in print again either. Gaiman could FINISH THE DAMN STORY...
(I just bought Kimota! last Saturday, and was despairing of ever getting hold of the originals again...)
Most music reviewers asked for such an agreement would tell the label to just fuck off. The goal is to be one of the few albums making it into the limited space for reviews; it's a reviewer's market.
Record reviewers generally have nothing of the sort. Procedure: labels drop off waaaay too many CDs; reviewer reviews some of them. Any label trying to get a reviewer to sign an NDA or equivalent will be told to fuck off and given no cooperation whatsoever in short order.
It's the labels vying for the attention of the reviewers, not the reviewers vying for the favour of the labels.
Record companies can only get away with this sort of thing in extremely few cases. Have you SEEN the piles and piles of CDs a reviewer typically gets? Do you know how little annoyance it takes to put a CD in the 'too much effort' pile?
Record companies are filled with drooling fuckwits.
"In a situation like that, you have to have a pretty damned good reason for going through all that - and as of yet, for the common man, there isn't such a reason."
For me, the pretty damned good reason was that 128kbps Oggs beat 192kbps MP3s, and my music HD was nearly full.
A confirm dialogue on quitting was committeed to death as 'modal = bad'. Of course, that didn't mean they removed that instant-death Ctrl+Q for quit...
I forget which, but there is an XUL file you can edit to remove the Ctrl+Q combination.
What happens is that the performance is improved incrementally. 1% faster here, 2% faster there. Things that hurt performance 3% get backed out almost immediately. And so on.
What this means over the course of a year or so is a smaller, faster browser. Mozilla 1.0 was better in every way than Mozilla M18, but was much faster and actually had a smaller memory footprint.
Developers are looking out for bigger performance wins, but the incremental approach is producing results without stability tradeoffs.
Fair enough. The problem is (as I understand it) that doing DOM properly takes a lot of memory in itself. The Mozilla application itself is not all that big.
IE6 is not light on the memory usage doing similarly complicated pages. (The total memory Windows tells you it is using does not include the 'System' memory it is using.)
Opera 6 cuts corners on its CSS support, and these make its speed and size better. I hope for better support in Opera 7.
One thing that's particularly annoying about Opera 6 is patchy CSS2 support. Which is quite surprising, considering they basically wrote the spec.
CSS2 and DOM are hard problems - IE's rendering engine needed a huge amount of work to get it halfway right in IE6. A lot of Opera's size and speed advantage comes from cutting corners.
is that the interface is complete shit. Embarrassingly awful. Talk to an Ericsson UI designer some time and listen to them apologise because they're so embarrassed about the phones.
In late 2000, two-thirds of Ericsson Australia employees used Nokias. It got to the point where the 'Christmas bonus' for employees in the development organisation was an R320s (the two years old model) with a free Vodafone contract (Vodafone being a complete joke for coverage in Australia).
I just moved to the UK. First thing I did was buy a prepaid phone. (Vodafone, admittedly, but their coverage here is actually pretty good.) I went for the Sony. You wouldn't believe how pissed off I was when I found out it was really another bloody Ericsson.
>If their interest is filesharing, then obviously that's not going to work as well.
Sure it will. Smaller files -> faster download times.
I mean in terms of critical mass of users, and of program defaults. A lot of filesharing apps and networks either don't share.ogg at all as yet, or don't recognise it as an audio format by default.
"With the 1.0 release of ogg vorbis, I believe the docs are now quite complete (that was one of the main tasks for the 1.0 release) so while I think at the time many of the points made in the kuro5hin.org article were valid and relevent then, I feel they are significantly less so now."
Here's to that hypothetical team of six crazy Russians then:-)
"I was under the impression they already had the integer decoder done, and were charging some kind of licensing or purchasing cost (not sure how it is licensed) to try to scrape up money to fund continued development (which I feel is quite reasonable btw; anyone else can build an integer decoder if they want to)."
"I think I'd be a bigger fool to attempt to convert all 40 gigs of my mp3s into ogg vorbis, which I would then subsequently NOT be able to load onto my Nomad Jukebox and take with me, everywhere I go."
Yeah. xiph.org really has to get that integer-maths Vorbis decoder done for the hardware players. That's a BIG problem with the Vorbis 1.0 package as released.
(And you realise of course that you never rip mp3s to Ogg, because the result sounds really crap. You rerip only from source CDs.)
"A format would have to come out with a _significantly_ improved compression to quality ratio, not to mention hardware support, for me to even blink on this one."
My data bucket is 40 gig, and saving 30% of that by reripping my CDs to.ogg is sufficiently significant to me. YMMV, of course.
Slashdot Mirror
Think of it as Mozilla forking, but the fork being blessed by mozilla.org. It was clear in the runup to 1.0 that there were tensions within the project; running forks internally keeps the developers happy and interested, and produces interesting things that can be adapted to the main trunk.
(Or, as is their goal, to replace the main trunk.)
Well, the version number in this case is accurate: this is an 0.2 and will act like one from time to time. You can actually expect noticeable changes from day to day.
Beating on nightlys gives immediate feedback on the effects of changes made that day - catch serious bugs early. Being a tester is a way to contribute greatly to a project as Joe User. And if there's a bug that's really been annoying you, you can get the fix straight away instead of having to wait until the next full release.
I think Phoenix is doing it this way because that's how Mozilla does it - and it works very well for Mozilla - and therefore because they can (being in the Mozilla build system).
Not with DC. It's much more complicated than that. Check the book Kimota! - covers the entire saga.
The most likely guess is that when Dez Skinn revived Marvelman/Miracleman in 1981, the rights had lapsed into the public domain. However, after having published the character for a while, he had then created interest in it. And, name and costume aside, the Alan Moore Marvelman/Miracleman was quite different to the Mick Anglo Marvelman.
Kimota! gives the impression that Gaiman would like the series to continue. Moore wouldn't mind seeing it in print again either. Gaiman could FINISH THE DAMN STORY ...
(I just bought Kimota! last Saturday, and was despairing of ever getting hold of the originals again ...)
So how's Happy99.exe going these days? The little turd was very much alive and well by the end of 2000 ...
Most music reviewers asked for such an agreement would tell the label to just fuck off. The goal is to be one of the few albums making it into the limited space for reviews; it's a reviewer's market.
It's the labels vying for the attention of the reviewers, not the reviewers vying for the favour of the labels.
Record companies are filled with drooling fuckwits.
The real purpose of Chatzilla is as a technology demo - to show you what sort of things you can do with XUL and JavaScript.
For me, the pretty damned good reason was that 128kbps Oggs beat 192kbps MP3s, and my music HD was nearly full.
This page will be linked in the next version of the FAQ, which I'm at home working on now ...
A confirm dialogue on quitting was committeed to death as 'modal = bad'. Of course, that didn't mean they removed that instant-death Ctrl+Q for quit ...
I forget which, but there is an XUL file you can edit to remove the Ctrl+Q combination.
What this means over the course of a year or so is a smaller, faster browser. Mozilla 1.0 was better in every way than Mozilla M18, but was much faster and actually had a smaller memory footprint.
Developers are looking out for bigger performance wins, but the incremental approach is producing results without stability tradeoffs.
My problem with Opera 5 is that its rendering is spotty (6 is much better) and that it crashes in a stiff breeze (6 is much better) ...
IE6 is not light on the memory usage doing similarly complicated pages. (The total memory Windows tells you it is using does not include the 'System' memory it is using.)
Opera 6 cuts corners on its CSS support, and these make its speed and size better. I hope for better support in Opera 7.
(disclaimer: I am involved in Mozilla.)
Hmm. With that much RAM, it should be faster than IE6 ... much faster.
"I just type cat | cc and get it right the first time."
With anything under 64MB, Mozilla is a slug. It would like 256MB. It's better than it used to be, but its arse is still incredibly fat.
CSS2 and DOM are hard problems - IE's rendering engine needed a huge amount of work to get it halfway right in IE6. A lot of Opera's size and speed advantage comes from cutting corners.
(Statement of bias: I'm involved in Mozilla.)
Having retracted it after the shitstorm doesn't change the fact that HP will be on the list of DMCA-wielding thugs from now until the end of time.
In late 2000, two-thirds of Ericsson Australia employees used Nokias. It got to the point where the 'Christmas bonus' for employees in the development organisation was an R320s (the two years old model) with a free Vodafone contract (Vodafone being a complete joke for coverage in Australia).
I just moved to the UK. First thing I did was buy a prepaid phone. (Vodafone, admittedly, but their coverage here is actually pretty good.) I went for the Sony. You wouldn't believe how pissed off I was when I found out it was really another bloody Ericsson.
Sure it will. Smaller files -> faster download times.
I mean in terms of critical mass of users, and of program defaults. A lot of filesharing apps and networks either don't share .ogg at all as yet, or don't recognise it as an audio format by default.
Here's to that hypothetical team of six crazy Russians then :-)
Something like that - see the kuro5hin story The Trouble with Vorbis.
Yeah. xiph.org really has to get that integer-maths Vorbis decoder done for the hardware players. That's a BIG problem with the Vorbis 1.0 package as released.
(And you realise of course that you never rip mp3s to Ogg, because the result sounds really crap. You rerip only from source CDs.)
"A format would have to come out with a _significantly_ improved compression to quality ratio, not to mention hardware support, for me to even blink on this one."
My data bucket is 40 gig, and saving 30% of that by reripping my CDs to .ogg is sufficiently significant to me. YMMV, of course.