The one thread per tab idea is nice, but the usual usecase is only one tab doing something. Nice for tab isolation, but not of great benefit for performance. It would seem to me that real benefit on a multicore box would come from parallelizing the bottleknecks, if possible... ie, the page rendering. Splitting the rendering on contained dom elements, perhaps. I'm not sure where or how this is best implemented, but addressing the cpu bottleknecks would seem to be the way to a faster browser.
Yes, because blanket broadcasting crap that statistically nobody wants is such a good way. Internet viewing, if anything, stands a much better chance at producing a workable revenue model.
Listening to techno, coding either from lazy boy with datahands in front of 1080p projection, or datahands on aeron in back of truck camper next to creek 20km off main roads in back country British Columbia. Code, code, code, fill generator, walk around looking at trees and butterflies, have a beer in the sun, code, code, code. Will be doing more of that this summer;)
The lockin forces the app-store, which forces the dev eula which forces the non-competition, which perpetuates apples distribution monopoly, which keeps the dump trucks full of cash coming in. Artist 'reward' as a justification for repression, etc. They don't care too much about jailbreaking because the avg buyer is not going to bother... and most people will continue to fork over cash, instead of stripping the nasty, avoiding the cash grab and using it as the more flexible and useful device it could be.
How old is x86? Could it not be considered a defacto standard by now? How would that affect the licensing? There is a chance to ditch x86 in general... perhaps the linux market is big enough? Could this be intel's grab for quick cash in down markets?
I think that's a bit naive... using network theory, the more people that use something, the more it will be used. If someone texts you, you're probably going to text back. If many of your friends send texts, you're probably going to be a texter... sure, texting is not a neccessity, but it's a high probability... an inevitable expensive upsell.
There is only power in Gahndi'esque non-participation if it is adopted in critical mass levels. This level is difficult to hit if there is a continual friend-backed participation push. Even if there is mass attempted vestigilization via concerted non-participation, there's still enough idiots who don't care. This, compounded with familiarity means there's no great escape from expensive texts.
You kids! Get off my lawn. Kitchen sink apps are a great, if often bloated, showcase of ideas. In a perfect world the good ideas stick around in new versions or interpretations and the cruft dies out.
great choice for things done well, or for things not understood
reuse means for more maintainable code. It'll be far easier for another developer to come in and help out.
more stable and optimized
saves time (and, possibly, money)
Homebrew;
great if you have time and dedication
you'll learn alot, which will help you with future evaluation. You'll also become a better coder.
your solution will fit the situation exactly
The reasons behind reuse should make it your first choice, but at the same time, it's hard to be a great coder without coding. Turn that desire to code loose on the situations where the existing approaches are insufficient.
If nvidia wanted to promote 3D, they'd release the stereoscope tech of the quadro workstation cards onto the gaming cards. Stereo 3D projection is great and would be so much better with more users behind it and more vendors enhancing and growing the tech. As it stands now, they've been sitting on the workstation tech for years and it's been going nowhere, as far as mass deployment goes.
Glider and bots wouldn't be used. A game should not need to be a second job in order to make a decent character in a reasonable amount of time.
They, as developers, are undermining their own position by supressing development. They've created a marvelous world, and people want to play in it. Part of the play is discovering through automation... while Blizz may not like the bots, they'll get nowhere by supressing the same type of innovation that lead to their own system.
This hypocritical behavior seems to not be out of place in the entertainment industry in general. Shut up, enjoy and pay you uncreative leech is the message, I guess. Time for the actual creators to overthrow the management and lawyers, methinks.
I picked up a HTPC with onboard nvidia gfx and while it's great for everything else, it has a hard time with 1080p. I just kind of assumed it'd be able to do fullscreen video at 1920x1080, but it is very choppy. Something to consider when looking for an HTPC. There must be reviews of onboard graphics out there...
Or you convert it to binary xml and compress it before it goes across network. XML is standard, easy to understand and easy to debug, and its verboseness goes away with a bit of work on either end.
Slightly off topic, but does anyone know the status of cachefs within the kernel? CacheFS was designed as a transparent generic cache for network files systems... nfs, etc. A read would pull into the cache and subsequent reads would come from cache.. so, a cache. It was designed for network filesystems, but I'm thinking it might apply to local file systems also. Specifically the cache could sit on a ram or flash based drive and pull off local hd's. As such, the commonly used bits (libs, etc) would reside on the cache and no longer be subject to seek and rotational latency... ie it'd be f*cking fast. Keeping the cache or profiling the cache across boots would lead to very fast loading (ie near instant) while still using slower, cheaper backend storage mechanisms. I was all set to try it out, but the most recent kernels are missing the cachefs documentation, leading me to think that cachefs is no longer supported in the kernel. Does anyone know the status of cachefs and if it is supported by recent kernels?
Yes, I know fast, big ssds are coming and they'll be getting faster, but they'll be expensive for a while, but still slower than an iRAM and much slower than memory on a pcie16x slot, should such a thing ever happen.
I've used a datahand for a while, and they're great. Even better on a good chair. Expensive, but worth every penny... and the keyboard can be used across multiple computers. The learning curve is a bit steep, but after a week or two it's quite usable. A downside is that you'll never want to work in an office without one. The personal's are as good as the pro's unless you specifically need macros (which you probably don't). They're only ps2, but datahand sell a usb converter that works well. Good stuff... now if only they'd do a wireless version, already.
Yup, it's a beautiful thing. Screen is great. The pdf's have to be drm free, well formatted, and ideally formatted for small format (ie not 8 1/2x11), mp3 functionality is a bit thin... but for reading books, it's as good as I've found, even if imperfect and too expensive. FYI pdf formatting and drm issues can be circumvented by repackaging the pdf through various tools. I actually recently wrote up a review of the bookeen cybook.
To beat a dead horse, it's also the case that some of the end users are extension developers themselves. Having extension mem tracking would make for easier leak finding by the extension dev and extension users could report excessive mem usage which might otherwise be missed (due to obscure untested condition, etc)
... too bad rationality tends to prevail. That is, with mass ownership self interested media seeding is easier, and while we may be told that X (where X is seeded opportunism) is good, the actual personal worth of X cannot be so easily manufactured. If it's a polished turd, it's still a turd. There is a *huge* profit to be made in spamming... but at a huge cost to the perception of the spammer. Soon they are painted as crap spewers... and have the respect deserved of crap spewers. A shame, perhaps, as mass broadcast can be very good (see BBC, PBS, Discovery (on occasion), etc). Those that know this will find ways to the good stuff, and rationality will prevail when the crap spewers tire of their own sickening reflection.
Slashdot Mirror
The one thread per tab idea is nice, but the usual usecase is only one tab doing something. Nice for tab isolation, but not of great benefit for performance. It would seem to me that real benefit on a multicore box would come from parallelizing the bottleknecks, if possible... ie, the page rendering. Splitting the rendering on contained dom elements, perhaps. I'm not sure where or how this is best implemented, but addressing the cpu bottleknecks would seem to be the way to a faster browser.
Yes, because blanket broadcasting crap that statistically nobody wants is such a good way. Internet viewing, if anything, stands a much better chance at producing a workable revenue model.
Listening to techno, coding either from lazy boy with datahands in front of 1080p projection, or datahands on aeron in back of truck camper next to creek 20km off main roads in back country British Columbia. Code, code, code, fill generator, walk around looking at trees and butterflies, have a beer in the sun, code, code, code. Will be doing more of that this summer ;)
The lockin forces the app-store, which forces the dev eula which forces the non-competition, which perpetuates apples distribution monopoly, which keeps the dump trucks full of cash coming in. Artist 'reward' as a justification for repression, etc. They don't care too much about jailbreaking because the avg buyer is not going to bother... and most people will continue to fork over cash, instead of stripping the nasty, avoiding the cash grab and using it as the more flexible and useful device it could be.
I thought the key to their success was OEM lock-in...
How old is x86? Could it not be considered a defacto standard by now? How would that affect the licensing? There is a chance to ditch x86 in general... perhaps the linux market is big enough? Could this be intel's grab for quick cash in down markets?
I think that's a bit naive... using network theory, the more people that use something, the more it will be used. If someone texts you, you're probably going to text back. If many of your friends send texts, you're probably going to be a texter... sure, texting is not a neccessity, but it's a high probability... an inevitable expensive upsell.
There is only power in Gahndi'esque non-participation if it is adopted in critical mass levels. This level is difficult to hit if there is a continual friend-backed participation push. Even if there is mass attempted vestigilization via concerted non-participation, there's still enough idiots who don't care. This, compounded with familiarity means there's no great escape from expensive texts.
You kids! Get off my lawn. Kitchen sink apps are a great, if often bloated, showcase of ideas. In a perfect world the good ideas stick around in new versions or interpretations and the cruft dies out.
hmm, that was by me (grumble).
If you want to use my proprietary (lgpl'd) nastieness, check out my persistence framework.
Specifically for the project in question, try to work around the deficiencies... ie go to jdbc using the hibernate tables, etc.
Reuse:
Homebrew;
The reasons behind reuse should make it your first choice, but at the same time, it's hard to be a great coder without coding. Turn that desire to code loose on the situations where the existing approaches are insufficient.
If nvidia wanted to promote 3D, they'd release the stereoscope tech of the quadro workstation cards onto the gaming cards. Stereo 3D projection is great and would be so much better with more users behind it and more vendors enhancing and growing the tech. As it stands now, they've been sitting on the workstation tech for years and it's been going nowhere, as far as mass deployment goes.
IRAMs don't play well with controllers... bad SATA implementation. Good idea, bad implementation, and a costly experiment on my part.
They, as developers, are undermining their own position by supressing development. They've created a marvelous world, and people want to play in it. Part of the play is discovering through automation... while Blizz may not like the bots, they'll get nowhere by supressing the same type of innovation that lead to their own system.
This hypocritical behavior seems to not be out of place in the entertainment industry in general. Shut up, enjoy and pay you uncreative leech is the message, I guess. Time for the actual creators to overthrow the management and lawyers, methinks.
Welcome back demonoid.
I picked up a HTPC with onboard nvidia gfx and while it's great for everything else, it has a hard time with 1080p. I just kind of assumed it'd be able to do fullscreen video at 1920x1080, but it is very choppy. Something to consider when looking for an HTPC. There must be reviews of onboard graphics out there...
Or you convert it to binary xml and compress it before it goes across network. XML is standard, easy to understand and easy to debug, and its verboseness goes away with a bit of work on either end.
Yes, I know fast, big ssds are coming and they'll be getting faster, but they'll be expensive for a while, but still slower than an iRAM and much slower than memory on a pcie16x slot, should such a thing ever happen.
I've used a datahand for a while, and they're great. Even better on a good chair. Expensive, but worth every penny... and the keyboard can be used across multiple computers. The learning curve is a bit steep, but after a week or two it's quite usable. A downside is that you'll never want to work in an office without one. The personal's are as good as the pro's unless you specifically need macros (which you probably don't). They're only ps2, but datahand sell a usb converter that works well. Good stuff... now if only they'd do a wireless version, already.
thanks, yw.
Yup, it's a beautiful thing. Screen is great. The pdf's have to be drm free, well formatted, and ideally formatted for small format (ie not 8 1/2x11), mp3 functionality is a bit thin... but for reading books, it's as good as I've found, even if imperfect and too expensive. FYI pdf formatting and drm issues can be circumvented by repackaging the pdf through various tools. I actually recently wrote up a review of the bookeen cybook.
To beat a dead horse, it's also the case that some of the end users are extension developers themselves. Having extension mem tracking would make for easier leak finding by the extension dev and extension users could report excessive mem usage which might otherwise be missed (due to obscure untested condition, etc)
Greasemonkey plugin to trim ac's with links? Metamod should deal with this.
... too bad rationality tends to prevail. That is, with mass ownership self interested media seeding is easier, and while we may be told that X (where X is seeded opportunism) is good, the actual personal worth of X cannot be so easily manufactured. If it's a polished turd, it's still a turd. There is a *huge* profit to be made in spamming... but at a huge cost to the perception of the spammer. Soon they are painted as crap spewers... and have the respect deserved of crap spewers. A shame, perhaps, as mass broadcast can be very good (see BBC, PBS, Discovery (on occasion), etc). Those that know this will find ways to the good stuff, and rationality will prevail when the crap spewers tire of their own sickening reflection.