err why shouldn't that be patentable? It involves a lot of research, in both hard & software combined. It's not smth you come up with after one night of good sleep, it costs a lot of money to get there. No company will invest in such a research if they can't patent it, otherwise another car company just reverse-engineers it's breaking system - and without much efford they would have copied the whole system for a fraction of the price, without a way for the original company of getting any of it's research money back except the stuff they get from selling their own cars... That would be the same as selling a disguised GPL software as closed-source package. Stealing other ppl's work, but then with no possibility to stop them. Patents are there for a reason, only patents applied to software are pure bullshit...
Last time I checked I still required SCSI-emulation in my kernel to get K3B or CDRecord working. There is the "dev=ATAPI:x,y,z" option, but I never got that working correctly.
And yes K3B is free, I never said it wasn't, Nero IS and will stay commercial software, and a lot of/.'ers think commercial soft is evil, I don't. It has a purpose, a market - whatever. If it's worth buying, and there's no equivalent free software, I'll buy it, like Opera i.e. (dont get me started on "why"). I won't buy NeroLinux - but I'll encourage the development, maybe some day it implements new popular features that K3B doesn't have, and competition in software is still a good thing...
> Off-topic in the Nero/k3b on GNU/Linux issue. You cannot have k3b in Windows, neither kd3, or amarok, lot of software. So, I don't understant you point.
The point is - usually linux users don't switch to windows, but windows users more and more start to switch to linux, and sometimes they miss their windows programs, or want a familiar interface. A lot of alternatives are available, but not always as good as their windows competitor. K3B here is an exception, but there aren't that much linux packages that clearly outclass their windows-equivalents - and that's the sad truth.
Not a very objective review?
on
NeroLinux vs. K3b
·
· Score: 5, Interesting
I read the article, and really got the impression that the author was a frequent K3B user, and didn't give Nero a chance. On windows, I always liked the Nero interface, simple, clean, not too much bullshit (or what the author calls "eyecandy"). It seems the linux version is more limited, but it's a good start, it supports things K3B (cdrecord etc) don't support, like CDwriting without scsi-emulation support, it should support USB writers without too much hassle (don't know what the status is on that-one with CDRecord etc). That has not been covered at all, and basicly the main bad thing about NeroLinux according to the author is that he likes the K3B interface more...
Also nero's licencing may be strange, but don't forget that a lot of CDwriters bundle Nero 6 with em. Most people using Nero in windows got it together with their writer. I myself have 3 or 4 official Nero cd's. It's nice that they also want to deliver a linux version in such scenario. Also another big point there is, you have to register, so they can very well track how many users are actually interested, it's their first time that they make a linux version, if they have enough interest, they probably will continue the development and start supporting it officially.
I do like K3B simply because it works, and the whole family can work with it, it's the perfect example for the dummy-user application in Linux, and it simply works, but I remember the times where I still prefered to use cdrecord on commanline than starting K3B, and it had to come a long way... Right now - K3B is high quality software, but there were times where the picture was different.
This is Ahead's first attempt to release linux software, and no instability was encountered. They should be encouraged, it took balls to do this - but reviews like this won't encourage or convince other software firms to write software for linux. Yes there is a lot of "free" software, but a lot of companies can't go without support or reliable and supported windows interoperability - preferably the same software running on Windows & Linux, so they don't have to train their personell to use new and different software.
Also a lot of software is almost not available for linux, like project management software, decent cross-platform agenda sharing. PLEASE don't mention Sunbird, it sucks - we tried it - and lost all agenda's doing so, suddenly all files on webdav were truncated to 0-byte size, after which sunbird freaked out and could not recover. Also it's way to complicated to setup, you need to do extensive configuration in apache, webdav and the client side (which is the worst thing). Sure - one day it will be better, but as long as you don't have these things, no company will completely switch over to linux. Yes there are other packages doing this kind of stuff in linux, but the support in windows is or unexisting or extremely limited. This is a giant step forward for Ahead, way to go I say!
Let's turn that around now - play advocate of the devil:)
How many people develop a web application that will be run by multiple persons, on multiple servers with different database engines? If you have an ordinary website which has only mysql support - that is no problem - since any PHP hosting i've encountered offered mysql support.
Database abstraction is good when you have a product or software package that should be distributed, but the vast majority of PHP-enthousiasts, for which this book is intended won't need another database than Mysql. In the future this may be SQLite - but time will tell...
Also - database abstraction doesn't work very well anyway. A big problem is differences in SQL between database engines - which is still present. Sollutions for this are simply too complicated for a simple web-application...
I agree that from the moment you need transactions & stuff or a seriously complicated database, you need Postgres. I'm not a Mysql fan at all, I use the best tool for the worst job but for web applications I seriously doubt Postgres has any chances in the mass-market because:
1) Mysql has a much smaller memory-print. Try giving a database 4 or 8mb or ram to work with in both Postgres and Mysql and compare the performance. Postgres loses BIG time, while mysql runs without a glitch. Postgres is nice if you have 1 server dedicated to 1 application. Mysql is so popular because for low-end sollutions, it uses very little resources, which is exactly what mass-hosting companies need. Why do you think you see so little web-hosting with Postgres support?
2) 95% of the queries performed by websites are typicly SELECT's - which is Mysql's strongest point.
3) Most websites are not mission-critical. If you lose data - too bad, that are some blog's you lost, or maybe some reactions... Big deal. Btw - postgres's recovery after a powerfailure isn't perfect, I know a company that is switching to Oracle after a small postgres disaster half a year ago... They lost a rather critical table with +- 7 million records, and daily thousands are added.. Backup was from the day before, but they lost a whole day of pritty important transactions (crash happened at 21u in the evening, backup started at 21u30). No system is perfect, but now they have someone to blame...
4) Postgres's duplication can't tip mysql's. In postgres quick hacks are available, Mysql supports load balancing and mirroring over multiple servers by default, and it's far superior to anything Postgres has.
5) Postgres should be optimized for 1 application. There is no generic optimal configuration, differences in performance between applications can be extreme depending on the configuration.
Mysql has it's purposes, so does Postgres, but I don't think they are competitors actually. It all depends on what you really need. I have experience with both, and for web-purposes - Mysql please... I plan looking into SQLite, which I already use for embedded purposes, but don't know how that is going to react in a multithreaded environment like a webserver.
Imho - extentions are not easy to use, and tend to become a mess... The "let's try this-one" can fuck up your entire browser, wouldn't be the first time that I could delete all my firefox/mozilla settings cause I just got a gray window with no buttons on it. Some are handy, but a lot of those are abandoned in an unfinished state.
I use opera now for 5 years, and I'm very happy with it. Tried to switch to Firefox, but it simply isn't as powerfull as Opera. THE main thing I love in opera is 100% keyboard-shortcut browsing & configurability - something that firefox completely lacks. Also - they all seem to lag behind on opera - who was always the first to come up with handy innovative ideas, which others mimmic - or at least try to, not always that successfull.
M2 is _powerfull_. It may look as it's only a stupid extention to the browser, but it has a lot of possibilities. Something Thunderbird still lacks - and which I miss very much now I got used to M2 is the "Unread" main folder - which is the default view, I really can't do w/o.
Opera also isn't a tab-browser - as all the others out-there, but is a full MDI browser, which has as effect that when a popup is needed, it can just be displayed in it's expected size - INSIDE the opera window without cluttering my precious desktop and taskbars. It is far more superior to anything else out-there on that level.
The popup-blocking is something Opera invented, skinning is hardly usefull - but also there for ages, from long before firefox was born. Options that should be options are "plugins" in firefox, even for the most stupid things. Why the hell are there "tabbed-browsing" enhancement plugins? Basicly because the default simply isn't good enough, and yes - those plugins are an improvement, but still not good enough imho.
It's not that Firefox is a bad browser - but everytime I use it for testing some website design, I get confronted with it's keyboard-usage and user interface limitations - which is my main reason to stick with Opera, together with it's consistent and well-tuned UI which simply hasn't got any competition in the browser market. For me it's like comparing Kate (the KDE "power"-editor) to vim, Kate has it's use, but for my daily job - vim please. My ideal browser would be a lean, mean gecko renderer in the Opera UI. As long as this doesn't change in FF - I will stick to opera.
Sorry - but you really sounded as if you never really tried opera. Once you get used to it - there's simply no way back, ad-supported or not - I don't care.
When you speak like that, it sounds like it's a piece-of-cake to do this. Also this is not java-related at all, this applies to all chipcards in general. All very nice in theory, but in reality - what you describe there requires a lot of multi-million-$ equipment, which only very few ppl (read: no-one) have in their private laboratory at home. Also don't forget you'll need some man-years of work to accomplish exactly what you want:) Most chip-foundries have the equipment, some security-audit centers also of which I know there's one in france and one in switserland, but they are really rare. I think those are the only-ones in Europe, maybe there's also one in Germany, but of that I am not sure. Really don't think you can find this equipment in your average supermarket. Also most of this equipment has military import/export restrictions applied to it, making it even harder to get:)
I work with chipcards on a daily base, and even the "non-intelligent" (asynchronious) cards, just read/write things protected by a fixed pincode are a though nut to crack if you only have the card. Typicly, these cards can store 256bytes (Siemens sle4442 cards) or 1kb of (rewritable) data, of which some parts can be fused to read-only.
The intelligent cards (synchronious cards) usually are a bit larger, 2K, 4K, 8K and 16K are common, 32K is already a lot, and 64K is - as said - extremely luxuary. They also cost a lot more per-card, but can run little applications. Depending on the card, this can be stored in non-volatile memory or burned permanently in rom. These applications usually are very small, but even card-java applications are slightly larger than native code for these cards, so you would need more space - certainly since the applications itself on javacards are usually also loaded into the expensive read/write non-volatile storage. A 10k application for such card is already huge. Most of the data-area is used by keys, which are relatively large for current sizes. People not working with embedded software usually have no idea what can be done with 1mb. You would be amazed what could be stored on a 256byte card, i.e. the social-security (SIS) card in Belgium is a 256byte card (with paranoid encryption applied to it:p)
I don't think this will actually be sold & used a lot at the moment, no-one is waiting for such big cards, they are way to expensive for mass-distribution and usage. It's like jumping from the current hard-disk technology which provides 400gig max disksize to 8000gig disks, and this in an area where storage isn't the real issue. Most of the times, interfacing speed is the real issue, where the serial communication between the chipcard and the other side is simply way to slow.
I don't say that 1mb is completely useless, I know the new Belgian electronic passport should store a picture of the card-holder, which is atm done on (I think) an 8kb java-card, which is way to small (ppl are hardly recognizable on the stored image), but otherwise it would simply cost too much. These new larger cards will put pressure on the prices of these cards, so they'll become cheaper - and hopefully, the eventual cards used for the EID (Electronic-ID) card here will be larger, so the image could be clearer:))
Originally coming from slackware (3.4 or smth back in '96 I think) I tried lots of things, always ending up again at slackware. Tried SuSE, Redhat, Debian and Mandrake. Some of them were very brief, didn't like Redhat at all, Mandrake was just to play with, but was way to limited. Ended up in gentoo, and after 2 years still there.
Ran debian for +- a year, but really didn't have a good feeling with the system, it just sucks monkeyballs as a desktop machine... Sure - I used the latest unstable or even test, but the problem is - the "unstable" or "test" perfectly discribe their state, I personally had a lot of dependency problems, and when a new version of some software package was released, it took months before it was added. The stable is hopelesly old, ok - stable, but secure?? I recently installed it as a firewall (still running), so I thought stable would be best. At that time, in the 2.4 series, 2.4.24 of.26 was just out I think. The debian installer happily installed a 2.4.18 kernel, and if you check the changelogs between those 2 versions, you'll notice they fixed some pritty bad root-exploits between those versions. Then checked the SSH version - same prob... What the hell should I install as a firewall then? Ok - added the security-update source for debian, which fixed the ssh-problems, but what the hell are you doing if you promote as your default stable version a very-exploitable default-install? I'm not even starting about the config-file managment and their updates... Lost settings more than once there..:(
Anyway - I don't like debian - but I can say WHY. This guy just ends up bashing on the users w/o any reason or arguments for "gentoo is bad". He does make gentoo-users look bad because they choose gentoo. Is gentoo bad then? Or is it simply bashing on a small minority inside the gentoo community (which certainly is there, I won't deny that), but which is in no way the whole gentoo community? The others choose their distrib simply cause they liked it and judged on that based on experiences - not on hear-say, but they all get the label of "ricer" like a big yellow star sewed on their clothes, simply because they choose smth a majority feels threatened by or doesn't like (whatever you choose, not gonna go into that-one here). Way to go to critisize a distro. If you do it, do it in a good way. My personal "hesitation" towards debian certainly had to do with how little doc or help I found while installing for the first time. The "doc" (or what it should represent) on the official site was pure garbage at that time. I came from slack, where you just put in a cd or floppy, it boots, and it installs, simple. First problem, where the hell are the installation cd's? I found "unofficial" cdimages, and finally gave one a try, afraid that I had someone's personal vision of a debian system on a cd, so not very sure about what was going to happen when I didn't use some official cdimages. Lateron found out that there weren't any official images - damn... Anyway - ended up installing, followed some instructions, and then the option came up to choose a package installer. It presented 4 choices or smth, dselect as the interactive, easy to use sollution and well - won't say much more than "it's a monster - RUUUUNNN!!!". Ended up reinstalling the whole system 4 times, after which I decided I would simply use apt-get to install separate packages. This still wasn't really the biggest issue with Debian.. The biggest issue was - the users. Ohhh ooeee auch... What was this article about? Right, gentoo users. What users always have the same critics on gentoo? I have to say, I only heard "complaints" about gentoo by other linux-ers that were running eeh almost afraid to say - debian. But at that time that wasn't the problem yet, since Gentoo didn't exist yet. Well - as a starting debian-user, I ended up reading the oh-so confusing doc on the website (was written in the way a chicken would describe how to create and lay an egg - they just already know how to do it, but not to exp
So the average inhabitants per square km in the US is +-31, in sweden it's +-21, in Korea it's about What's your point? That denser populations are harder to serve? I would think it's exactly the opposite no? If I have 3 families living in my street, and I put a cable, the cable is a lot more expensive than if there would be 6 families living there no? Don't blame the size, blame the short-sightness and fear for doing large investments of your ISP's & phone companies.
So if you compare Sweden to California and Korea to Indiana, also compare them with the numbers of California plz...
For South-Korea I think the size is incorrect, since this would result in a stunning 494 inhabitants per square km...
This is about reading in streams from magnetic media as far as I understood - but couldn't you store sequential data non-sequential and read it with different heads to archieve higher speeds? Seems logical to me... Just need a logic then to re-assemble all the data from the different streams into one stream.
Even if we would still use magnetism for storing data when we hit this limit, I think there are plenty sollutions...
Then you have far better speed than java, more readable code, in less time. Yah! multiple languages win again:)
You have clearly never 1) seen java code 2) written java code, since java is in general much more readable than C(++) code - even as pure C(++) programmer without any Java knowledge, and a lot faster to write if you know Java as well as you know C(++). I write both for living - and both have their purposes and applications. It's simple - when writing an application - you should use the best tool for the job, and in a lot of cases, Java is that tool - but isn't used because of practical limitations (small JVM userbase most of the time) or the general believe that Java is slow.
I had this problem at work some time ago. The server was eating memory and could not accept more than 5 connections or it would eat 1gig+ of memory. It was simple, the server was written in Java, so Java was the problem. After 6 months someone found an *oops* in a.so that the server used through a JNI interface. Fixing that *oops* allowed the server to accept more than 1000 connections at the same time, a _lot_ faster than before - with no memory problems... I never believed it was Java's problem there - but I was one of the few.
Bad coding == bad results - in C, Java, Python, Perl - whatever.
Firfox is a small (6mb) download and opera is the gigantic bloatware of the full 3.2mb (no java), which includes a mailprogram, plus all the features you name there and even a lot more... Shame on them.
They even managed to make the user-interface more flexible and customizable without any form of plugins, just a bit of skinning, everything working straight out of the box, sensible shortcuts for really everything, and even something that closely resembles standard compliance (however that fails in maybe 0.5% of the cases).
It also includes by default numerous features which Firefox/Mozilla users thought were cool - so for these browsers, the plugin-hell was created. Do you still know what version of which plugin you have installed? Or where you can find the latest version of that plugin? I don't, and I don't like searching them after (another) reinstall of my windows box.
I am a power user, and Opera simply gives me what I want: an extremely powerfull userinterface with sensible defaults, but still easy to use for a newbie. The renderer now and then fails, but it has cool and handy options like page zooming, custom stylesheets, small screen rendering and much more. Anyhow, it would be ideal to have the opera userinterface with the gecko renderer, but Opera isn't opensource, and a lot of people don't like that. I also would prefer to see it otherwise, but to me it makes sense. These people are in it for the money, they also have to eat, and the last few years, they managed to bring us the most innovative browser around.
What's the last cool new feature in mozilla that wasn't copied from somewhere else? Honestly, I can't find any. Some people say that XUL is the thing I'm looking for then, well, it has some nice possibilities, but it's bloated, and has no place in a webbrowser imho. It is used for a "flexible" userinterface. Flexible? I can't even move my addressbar, buttons or tabbar around? Oh but you could install a plugin that does this for you... You call that flexible? Oh wait, maybe I have to open my favorite text editor and edit some XML source somewhere if I don't feel like downloading a plugin.
When I first heard of the Phoenix project, which later became Firebird, and which is now FireFox, I hoped for a completely XUL-less userinterface which did what I wanted... Boy, what was I dissapointed when I saw that they just created a stripdown of Mozilla.
For a low-profile user, it's great, but I don't understand how so many powerusers can bear using such an unflexible and limited userinterface. Are they blinded by OSS? Is it just that? It's not because it's not OSS that is can't be good. Maybe do they love to switch between mouse/keyboard so often, they have tons of time to investigate and customize their browser... Well - I don't.
This is not meant as a flame on the mozilla project, the Gecko engine is one of the most powerfull around, but face it, the userinterface simply sucks. It's a simple IE on steroids, while opera simply is the steroids:)
Anyway - the original discussion was about webdesigning for IE - not about browsers:)
I think this article really makes sense, when I look at my grandma, this is exactly one of the problems. Last time I was really surprised how far she got on her own. She told me that she wanted to learn a lot of other stuff, she felt like she didn't know enough of the computer's power. She "only" knew how to mail, use google and surf a bit. I only remember telling her more or less how to mail. So she learned how to use a mouse pritty well, but she still panics (well at least call someone else to help her) when some error pops up. She simply doesn't know what happens, and indeed, in a multi-window environment, it is certainly not that easy to know which program gave the exact error.
We have an awfull lot of GUI's around in linux, at the moment, with at the "top" arguably Gnome and KDE, which do a great job for the more experienced desktop user, but in this test, they would fail miserably. I do not think you can limit such people to the commandline forever, graphics are very usefull in many cases, like when browsing the web. They can be made chaotic, but can also be used to structure things in a clean way.
So why don't we see a graphical userinterface that is more like the way the console is used? It will probably be very hard to imagine how this would work, but it should be consistent, same logic everywhere, and as the article says - the user is - as said - only interested in one thing at a time, and wants to be notified discretely about other events without being pulled away from his current task... I think it would be a very challenging and usefull job to design such a userinterface which is very (to the extreme) consistent in use, but I'm afraid there won't be any interest in making such a thing in the linux/geek world which loves the power to customize...
That wasn't the actual intention of the question - it really was a "btw" - was too lazy to check if they were already present, have to upgrade the "family" pc that still runs debian (and no complaining about that btw), and it runs and keeps running (KDE 3.1.5 btw):) Also still have a slack running somewhere, and a SuSE also, maybe a good time to upgrade:)
I'll stick to gentoo, and not for it's "optimization", but for it's transparancy in configuration, superior documentation, and fast package updates. Also the freedom to choose is there more than with any other distrib, I mean, Debian with KDE? Wow man - debian-fanboys look at you in a very frightning way...;)
The installation can take a while (well - if you use the pre-compiled cd's it's just as fast or even faster than Debian) - that's true, but even if you put a noob behind it with the docs, he should get a working system after a while (though maybe his "windows" may be vanished:p) I tested this with my sister, and she had to ask me a couple of things, some things not very clear to any newbie (partitioning, what filesystem to choose,...) I actually didn't hear from her anymore till her machine started flipping (bad ram...), and that was 4-5 months later...
I tried a lot of distribs, started with Slack (back in '96), then tried SuSE after a few years, went back to Slack, tried Debian (installation simply sucks), went back to Slack, tried Mandrake somewhere in between (in VMWare) - and then decided to try Gentoo, and I'll stick to that...
KDE 3.2 packages will be "masked" in portage for a few hours prolly cause the package maintainer is asleep I heard:D
Btw - when can we expect "official" debian/slackware/... packages?
Definitly this game rulez: Soldat
It's a 2d platform alike multiplayer game that clearly had some inspiration from worms - but is is oh so cool and fun to play in lan with 4 persons, the gameplay is just fantastic:)
Truck dismount already had a prequal otherwise it would be my choice:)
The other games I saw really were sequals or non-original things, so I guess there won't be any other...
1) Quake 3 - and video memory... Do you actually believe a nearly 4 year old game is dependand on video memory? Back in those days, more than 32mb was outrageous. No way that Q3 ever uses more than 64mb - I would even be surprised if it ever reaches that... I even think that recent games will use up the full 128mb. The videocards have exactly the same GPU - which should perform exactly the same in the same machine if the application doesn't need that exceptional amount of memory.
2) The Opteron/Athlon64 beats the crap out of anything when it comes to memory access. It has an on-die memory-controller - what do you expect? Larger files would have made the G5 only look worse. In fact - they also tested with an 150mb image - and the difference became relatively bigger.
3) & 5) Are you suggesting to bench and compare completely different pieces of software? What sense would that make for pure performance comparison? The best way to compare these things would be to write an own dedicated bench, not comparing apples with banana's.
Stripping the OS - what the hell are you thinking? As if Windows is the most optimal OS. As far as I know, MacOS-X has a much better task scheduling than any Windows around (which is - I'm afraid - not that hard)
You also seem to miss the point completely that in fact they ARE comparing kiwi's with banana's cause the Mac requires a dual 2Ghz rig to get even close to a single-Opteron/Athlon64. How fair is that? Ok - the AMD riggs have a slight Mhz advantage in most of the cases, but if you look at this from a neutral position, I don't see how you could not say that the AMD beats the hell out of the G5. Sorry.
He looks at the facts from the point of view of a linux-geek. The thing is - most of the servers online are not owned by linux-freaks - most of them are from a gaming-related company or a clan. The servers simply run best and require less resources on linux - and linux-serverhosting is cheaper than windows-servers. THAT are the reasons that so many game-servers run linux - not that some geek runs a server for a windows-game just because it runs in linux. At this moment - there are thousands of halflife servers running linux - when there is no native linux-version - would that all be linux-geeks that like it that there is no native version? Don't think so - just the fact that there are already so much HL servers running linux, and the fact that there were already an awfull lot of them long before I was able to even think of starting HL in Wine - or WineX even existed - proves the author wrong.
Another point - about HL2 not running on linux. OpenGL simply doesn't support the necessary features that the HL2 engine uses - unless it is done thu vendor-specific extentions of nVidia or ATI. The HL developers already didn't like it to write a specific rendering path for the last-generation of nVidia cards because they were to slow when running with the generic DX9 rendering path - so they are certainly not going to write 2 separate renderers using opengl for the few customers that want to run it in linux - because - face it - we are a minority in gaming-land. From a developers point of view - I can perfectly understand that (I don't say I like it). Also they don't have any advantages supporting and developing a linux-version - linux is there - and it runs their servers just fine at this moment. Epic had to develop in linux anyway for the PS2 version of their engine. Owkay - they had to write the OpenGL renderer - that's true - but by experience I know it is easyer when you can port a game and run & test the game itself it in the development environment - and have one guy writing the PS2 rendering engine - and let the others work on the port - while able to test it without the necessary equipment - believe me - I do embedded development - but we test and run the main-applications on pc, simulating the hardware - and have one guy (read: me) do the hardware dependand stuff. Yes - that takes a one-time development of the testing-libraries for pc - but in the business I am - I see another type of hardware every half year or so - so it is worth it. I strongly doubt that we will actually see a PS2 version of HL2 any time soon - so there will be no need for them to port/run the game in linux.
It's putting stuff in a way that *looks* clean but with the large disadvantage that you will be unable to locate something when you really really need it quickly...
So if this code implementing NUMA and RCU is now part of the Sys V codebase - SCO basicly owns code (written by IBM/Sequent) which implement at least 2 patents of IBM? Couldn't IBM sue them for this because SCO doesn't have any licence for these patents but having "their" code commercially used and distributed?
Notes can't be, but that's not a huge loss in my humble opinion. There are possible better alternatives (notepad, vim,...)
I use vim on a daily bases, and I can't say that's a better alternative than the notes in opera 7. In opera 7, you select some text on a webpage, add it as a note (using the menu's or the hotkey), the note is saved, so you can look at it afterwards, but (very important but) the thing that makes it a superb feature is that it remembers on which page it originally was, and what position, meaning that you can immediately jump to the page and position on that page where the note originally came from. I don't really see how you would do this in vim or any other text-editor in this comfortable way:)
And it is like someone already said here somewhere, Opera is imho the most innovative piece of software introducing new handy features with every release. Only the very popular are ported to Mozilla afterwards, which keeps the Opera user with Opera, and the Mozilla users thinking Mozilla has all the features it needs, and has all the features Opera has. Live and let live I would say:)
Slashdot Mirror
err why shouldn't that be patentable? It involves a lot of research, in both hard & software combined. It's not smth you come up with after one night of good sleep, it costs a lot of money to get there. No company will invest in such a research if they can't patent it, otherwise another car company just reverse-engineers it's breaking system - and without much efford they would have copied the whole system for a fraction of the price, without a way for the original company of getting any of it's research money back except the stuff they get from selling their own cars... That would be the same as selling a disguised GPL software as closed-source package. Stealing other ppl's work, but then with no possibility to stop them. Patents are there for a reason, only patents applied to software are pure bullshit...
Last time I checked I still required SCSI-emulation in my kernel to get K3B or CDRecord working. There is the "dev=ATAPI:x,y,z" option, but I never got that working correctly.
/.'ers think commercial soft is evil, I don't. It has a purpose, a market - whatever. If it's worth buying, and there's no equivalent free software, I'll buy it, like Opera i.e. (dont get me started on "why"). I won't buy NeroLinux - but I'll encourage the development, maybe some day it implements new popular features that K3B doesn't have, and competition in software is still a good thing...
And yes K3B is free, I never said it wasn't, Nero IS and will stay commercial software, and a lot of
> Off-topic in the Nero/k3b on GNU/Linux issue. You cannot have k3b in Windows, neither kd3, or amarok, lot of software. So, I don't understant you point.
The point is - usually linux users don't switch to windows, but windows users more and more start to switch to linux, and sometimes they miss their windows programs, or want a familiar interface. A lot of alternatives are available, but not always as good as their windows competitor. K3B here is an exception, but there aren't that much linux packages that clearly outclass their windows-equivalents - and that's the sad truth.
I read the article, and really got the impression that the author was a frequent K3B user, and didn't give Nero a chance. On windows, I always liked the Nero interface, simple, clean, not too much bullshit (or what the author calls "eyecandy"). It seems the linux version is more limited, but it's a good start, it supports things K3B (cdrecord etc) don't support, like CDwriting without scsi-emulation support, it should support USB writers without too much hassle (don't know what the status is on that-one with CDRecord etc). That has not been covered at all, and basicly the main bad thing about NeroLinux according to the author is that he likes the K3B interface more...
Also nero's licencing may be strange, but don't forget that a lot of CDwriters bundle Nero 6 with em. Most people using Nero in windows got it together with their writer. I myself have 3 or 4 official Nero cd's. It's nice that they also want to deliver a linux version in such scenario. Also another big point there is, you have to register, so they can very well track how many users are actually interested, it's their first time that they make a linux version, if they have enough interest, they probably will continue the development and start supporting it officially.
I do like K3B simply because it works, and the whole family can work with it, it's the perfect example for the dummy-user application in Linux, and it simply works, but I remember the times where I still prefered to use cdrecord on commanline than starting K3B, and it had to come a long way... Right now - K3B is high quality software, but there were times where the picture was different.
This is Ahead's first attempt to release linux software, and no instability was encountered. They should be encouraged, it took balls to do this - but reviews like this won't encourage or convince other software firms to write software for linux. Yes there is a lot of "free" software, but a lot of companies can't go without support or reliable and supported windows interoperability - preferably the same software running on Windows & Linux, so they don't have to train their personell to use new and different software.
Also a lot of software is almost not available for linux, like project management software, decent cross-platform agenda sharing. PLEASE don't mention Sunbird, it sucks - we tried it - and lost all agenda's doing so, suddenly all files on webdav were truncated to 0-byte size, after which sunbird freaked out and could not recover. Also it's way to complicated to setup, you need to do extensive configuration in apache, webdav and the client side (which is the worst thing). Sure - one day it will be better, but as long as you don't have these things, no company will completely switch over to linux. Yes there are other packages doing this kind of stuff in linux, but the support in windows is or unexisting or extremely limited. This is a giant step forward for Ahead, way to go I say!
"SLI, or Scalable Link Interface..."
And I always thought this was "Scan Line Interleave"... ??? At least with the Voodoo 2 cards, it was like that...
Let's turn that around now - play advocate of the devil :)
How many people develop a web application that will be run by multiple persons, on multiple servers with different database engines? If you have an ordinary website which has only mysql support - that is no problem - since any PHP hosting i've encountered offered mysql support.
Database abstraction is good when you have a product or software package that should be distributed, but the vast majority of PHP-enthousiasts, for which this book is intended won't need another database than Mysql. In the future this may be SQLite - but time will tell...
Also - database abstraction doesn't work very well anyway. A big problem is differences in SQL between database engines - which is still present. Sollutions for this are simply too complicated for a simple web-application...
I agree that from the moment you need transactions & stuff or a seriously complicated database, you need Postgres. I'm not a Mysql fan at all, I use the best tool for the worst job but for web applications I seriously doubt Postgres has any chances in the mass-market because:
1) Mysql has a much smaller memory-print. Try giving a database 4 or 8mb or ram to work with in both Postgres and Mysql and compare the performance. Postgres loses BIG time, while mysql runs without a glitch. Postgres is nice if you have 1 server dedicated to 1 application. Mysql is so popular because for low-end sollutions, it uses very little resources, which is exactly what mass-hosting companies need. Why do you think you see so little web-hosting with Postgres support?
2) 95% of the queries performed by websites are typicly SELECT's - which is Mysql's strongest point.
3) Most websites are not mission-critical. If you lose data - too bad, that are some blog's you lost, or maybe some reactions... Big deal. Btw - postgres's recovery after a powerfailure isn't perfect, I know a company that is switching to Oracle after a small postgres disaster half a year ago... They lost a rather critical table with +- 7 million records, and daily thousands are added.. Backup was from the day before, but they lost a whole day of pritty important transactions (crash happened at 21u in the evening, backup started at 21u30). No system is perfect, but now they have someone to blame...
4) Postgres's duplication can't tip mysql's. In postgres quick hacks are available, Mysql supports load balancing and mirroring over multiple servers by default, and it's far superior to anything Postgres has.
5) Postgres should be optimized for 1 application. There is no generic optimal configuration, differences in performance between applications can be extreme depending on the configuration.
Mysql has it's purposes, so does Postgres, but I don't think they are competitors actually. It all depends on what you really need. I have experience with both, and for web-purposes - Mysql please... I plan looking into SQLite, which I already use for embedded purposes, but don't know how that is going to react in a multithreaded environment like a webserver.
So how old are you? 3? :D Yeah - then I'm probably old :)
Imho - extentions are not easy to use, and tend to become a mess... The "let's try this-one" can fuck up your entire browser, wouldn't be the first time that I could delete all my firefox/mozilla settings cause I just got a gray window with no buttons on it. Some are handy, but a lot of those are abandoned in an unfinished state.
I use opera now for 5 years, and I'm very happy with it. Tried to switch to Firefox, but it simply isn't as powerfull as Opera. THE main thing I love in opera is 100% keyboard-shortcut browsing & configurability - something that firefox completely lacks. Also - they all seem to lag behind on opera - who was always the first to come up with handy innovative ideas, which others mimmic - or at least try to, not always that successfull.
M2 is _powerfull_. It may look as it's only a stupid extention to the browser, but it has a lot of possibilities. Something Thunderbird still lacks - and which I miss very much now I got used to M2 is the "Unread" main folder - which is the default view, I really can't do w/o.
Opera also isn't a tab-browser - as all the others out-there, but is a full MDI browser, which has as effect that when a popup is needed, it can just be displayed in it's expected size - INSIDE the opera window without cluttering my precious desktop and taskbars. It is far more superior to anything else out-there on that level.
The popup-blocking is something Opera invented, skinning is hardly usefull - but also there for ages, from long before firefox was born. Options that should be options are "plugins" in firefox, even for the most stupid things. Why the hell are there "tabbed-browsing" enhancement plugins? Basicly because the default simply isn't good enough, and yes - those plugins are an improvement, but still not good enough imho.
It's not that Firefox is a bad browser - but everytime I use it for testing some website design, I get confronted with it's keyboard-usage and user interface limitations - which is my main reason to stick with Opera, together with it's consistent and well-tuned UI which simply hasn't got any competition in the browser market. For me it's like comparing Kate (the KDE "power"-editor) to vim, Kate has it's use, but for my daily job - vim please.
My ideal browser would be a lean, mean gecko renderer in the Opera UI. As long as this doesn't change in FF - I will stick to opera.
Sorry - but you really sounded as if you never really tried opera. Once you get used to it - there's simply no way back, ad-supported or not - I don't care.
When you speak like that, it sounds like it's a piece-of-cake to do this. Also this is not java-related at all, this applies to all chipcards in general. All very nice in theory, but in reality - what you describe there requires a lot of multi-million-$ equipment, which only very few ppl (read: no-one) have in their private laboratory at home. Also don't forget you'll need some man-years of work to accomplish exactly what you want :) :)
:p)
:))
Most chip-foundries have the equipment, some security-audit centers also of which I know there's one in france and one in switserland, but they are really rare. I think those are the only-ones in Europe, maybe there's also one in Germany, but of that I am not sure.
Really don't think you can find this equipment in your average supermarket. Also most of this equipment has military import/export restrictions applied to it, making it even harder to get
I work with chipcards on a daily base, and even the "non-intelligent" (asynchronious) cards, just read/write things protected by a fixed pincode are a though nut to crack if you only have the card. Typicly, these cards can store 256bytes (Siemens sle4442 cards) or 1kb of (rewritable) data, of which some parts can be fused to read-only.
The intelligent cards (synchronious cards) usually are a bit larger, 2K, 4K, 8K and 16K are common, 32K is already a lot, and 64K is - as said - extremely luxuary. They also cost a lot more per-card, but can run little applications. Depending on the card, this can be stored in non-volatile memory or burned permanently in rom. These applications usually are very small, but even card-java applications are slightly larger than native code for these cards, so you would need more space - certainly since the applications itself on javacards are usually also loaded into the expensive read/write non-volatile storage. A 10k application for such card is already huge. Most of the data-area is used by keys, which are relatively large for current sizes.
People not working with embedded software usually have no idea what can be done with 1mb. You would be amazed what could be stored on a 256byte card, i.e. the social-security (SIS) card in Belgium is a 256byte card (with paranoid encryption applied to it
I don't think this will actually be sold & used a lot at the moment, no-one is waiting for such big cards, they are way to expensive for mass-distribution and usage. It's like jumping from the current hard-disk technology which provides 400gig max disksize to 8000gig disks, and this in an area where storage isn't the real issue. Most of the times, interfacing speed is the real issue, where the serial communication between the chipcard and the other side is simply way to slow.
I don't say that 1mb is completely useless, I know the new Belgian electronic passport should store a picture of the card-holder, which is atm done on (I think) an 8kb java-card, which is way to small (ppl are hardly recognizable on the stored image), but otherwise it would simply cost too much. These new larger cards will put pressure on the prices of these cards, so they'll become cheaper - and hopefully, the eventual cards used for the EID (Electronic-ID) card here will be larger, so the image could be clearer
Originally coming from slackware (3.4 or smth back in '96 I think) I tried lots of things, always ending up again at slackware. Tried SuSE, Redhat, Debian and Mandrake. Some of them were very brief, didn't like Redhat at all, Mandrake was just to play with, but was way to limited. Ended up in gentoo, and after 2 years still there.
.26 was just out I think. The debian installer happily installed a 2.4.18 kernel, and if you check the changelogs between those 2 versions, you'll notice they fixed some pritty bad root-exploits between those versions. Then checked the SSH version - same prob... What the hell should I install as a firewall then? Ok - added the security-update source for debian, which fixed the ssh-problems, but what the hell are you doing if you promote as your default stable version a very-exploitable default-install? I'm not even starting about the config-file managment and their updates... Lost settings more than once there.. :(
Ran debian for +- a year, but really didn't have a good feeling with the system, it just sucks monkeyballs as a desktop machine... Sure - I used the latest unstable or even test, but the problem is - the "unstable" or "test" perfectly discribe their state, I personally had a lot of dependency problems, and when a new version of some software package was released, it took months before it was added. The stable is hopelesly old, ok - stable, but secure?? I recently installed it as a firewall (still running), so I thought stable would be best. At that time, in the 2.4 series, 2.4.24 of
Anyway - I don't like debian - but I can say WHY. This guy just ends up bashing on the users w/o any reason or arguments for "gentoo is bad". He does make gentoo-users look bad because they choose gentoo. Is gentoo bad then? Or is it simply bashing on a small minority inside the gentoo community (which certainly is there, I won't deny that), but which is in no way the whole gentoo community? The others choose their distrib simply cause they liked it and judged on that based on experiences - not on hear-say, but they all get the label of "ricer" like a big yellow star sewed on their clothes, simply because they choose smth a majority feels threatened by or doesn't like (whatever you choose, not gonna go into that-one here). Way to go to critisize a distro. If you do it, do it in a good way. My personal "hesitation" towards debian certainly had to do with how little doc or help I found while installing for the first time. The "doc" (or what it should represent) on the official site was pure garbage at that time. I came from slack, where you just put in a cd or floppy, it boots, and it installs, simple. First problem, where the hell are the installation cd's? I found "unofficial" cdimages, and finally gave one a try, afraid that I had someone's personal vision of a debian system on a cd, so not very sure about what was going to happen when I didn't use some official cdimages. Lateron found out that there weren't any official images - damn... Anyway - ended up installing, followed some instructions, and then the option came up to choose a package installer. It presented 4 choices or smth, dselect as the interactive, easy to use sollution and well - won't say much more than "it's a monster - RUUUUNNN!!!". Ended up reinstalling the whole system 4 times, after which I decided I would simply use apt-get to install separate packages.
This still wasn't really the biggest issue with Debian.. The biggest issue was - the users. Ohhh ooeee auch... What was this article about? Right, gentoo users. What users always have the same critics on gentoo? I have to say, I only heard "complaints" about gentoo by other linux-ers that were running eeh almost afraid to say - debian. But at that time that wasn't the problem yet, since Gentoo didn't exist yet. Well - as a starting debian-user, I ended up reading the oh-so confusing doc on the website (was written in the way a chicken would describe how to create and lay an egg - they just already know how to do it, but not to exp
So the average inhabitants per square km in the US is +-31, in sweden it's +-21, in Korea it's about What's your point? That denser populations are harder to serve? I would think it's exactly the opposite no? If I have 3 families living in my street, and I put a cable, the cable is a lot more expensive than if there would be 6 families living there no? Don't blame the size, blame the short-sightness and fear for doing large investments of your ISP's & phone companies.
So if you compare Sweden to California and Korea to Indiana, also compare them with the numbers of California plz...
For South-Korea I think the size is incorrect, since this would result in a stunning 494 inhabitants per square km...
Too bad it doesn't fix any of the IE positioning bugs...
And then you find a site like this...
It will be a surprise to a lot of you guys, but Opera is actually the better CSS renderer...
This is about reading in streams from magnetic media as far as I understood - but couldn't you store sequential data non-sequential and read it with different heads to archieve higher speeds? Seems logical to me... Just need a logic then to re-assemble all the data from the different streams into one stream.
Even if we would still use magnetism for storing data when we hit this limit, I think there are plenty sollutions...
Then you have far better speed than java, more readable code, in less time. Yah! multiple languages win again :)
.so that the server used through a JNI interface. Fixing that *oops* allowed the server to accept more than 1000 connections at the same time, a _lot_ faster than before - with no memory problems... I never believed it was Java's problem there - but I was one of the few.
You have clearly never 1) seen java code 2) written java code, since java is in general much more readable than C(++) code - even as pure C(++) programmer without any Java knowledge, and a lot faster to write if you know Java as well as you know C(++). I write both for living - and both have their purposes and applications. It's simple - when writing an application - you should use the best tool for the job, and in a lot of cases, Java is that tool - but isn't used because of practical limitations (small JVM userbase most of the time) or the general believe that Java is slow.
I had this problem at work some time ago. The server was eating memory and could not accept more than 5 connections or it would eat 1gig+ of memory. It was simple, the server was written in Java, so Java was the problem. After 6 months someone found an *oops* in a
Bad coding == bad results - in C, Java, Python, Perl - whatever.
Firfox is a small (6mb) download and opera is the gigantic bloatware of the full 3.2mb (no java), which includes a mailprogram, plus all the features you name there and even a lot more... Shame on them.
:)
:)
They even managed to make the user-interface more flexible and customizable without any form of plugins, just a bit of skinning, everything working straight out of the box, sensible shortcuts for really everything, and even something that closely resembles standard compliance (however that fails in maybe 0.5% of the cases).
It also includes by default numerous features which Firefox/Mozilla users thought were cool - so for these browsers, the plugin-hell was created. Do you still know what version of which plugin you have installed? Or where you can find the latest version of that plugin? I don't, and I don't like searching them after (another) reinstall of my windows box.
I am a power user, and Opera simply gives me what I want: an extremely powerfull userinterface with sensible defaults, but still easy to use for a newbie. The renderer now and then fails, but it has cool and handy options like page zooming, custom stylesheets, small screen rendering and much more.
Anyhow, it would be ideal to have the opera userinterface with the gecko renderer, but Opera isn't opensource, and a lot of people don't like that. I also would prefer to see it otherwise, but to me it makes sense. These people are in it for the money, they also have to eat, and the last few years, they managed to bring us the most innovative browser around.
What's the last cool new feature in mozilla that wasn't copied from somewhere else?
Honestly, I can't find any. Some people say that XUL is the thing I'm looking for then, well, it has some nice possibilities, but it's bloated, and has no place in a webbrowser imho. It is used for a "flexible" userinterface. Flexible? I can't even move my addressbar, buttons or tabbar around? Oh but you could install a plugin that does this for you... You call that flexible? Oh wait, maybe I have to open my favorite text editor and edit some XML source somewhere if I don't feel like downloading a plugin.
When I first heard of the Phoenix project, which later became Firebird, and which is now FireFox, I hoped for a completely XUL-less userinterface which did what I wanted... Boy, what was I dissapointed when I saw that they just created a stripdown of Mozilla.
For a low-profile user, it's great, but I don't understand how so many powerusers can bear using such an unflexible and limited userinterface. Are they blinded by OSS? Is it just that? It's not because it's not OSS that is can't be good.
Maybe do they love to switch between mouse/keyboard so often, they have tons of time to investigate and customize their browser... Well - I don't.
This is not meant as a flame on the mozilla project, the Gecko engine is one of the most powerfull around, but face it, the userinterface simply sucks. It's a simple IE on steroids, while opera simply is the steroids
Anyway - the original discussion was about webdesigning for IE - not about browsers
I think this article really makes sense, when I look at my grandma, this is exactly one of the problems. Last time I was really surprised how far she got on her own. She told me that she wanted to learn a lot of other stuff, she felt like she didn't know enough of the computer's power. She "only" knew how to mail, use google and surf a bit. I only remember telling her more or less how to mail.
So she learned how to use a mouse pritty well, but she still panics (well at least call someone else to help her) when some error pops up. She simply doesn't know what happens, and indeed, in a multi-window environment, it is certainly not that easy to know which program gave the exact error.
We have an awfull lot of GUI's around in linux, at the moment, with at the "top" arguably Gnome and KDE, which do a great job for the more experienced desktop user, but in this test, they would fail miserably. I do not think you can limit such people to the commandline forever, graphics are very usefull in many cases, like when browsing the web. They can be made chaotic, but can also be used to structure things in a clean way.
So why don't we see a graphical userinterface that is more like the way the console is used? It will probably be very hard to imagine how this would work, but it should be consistent, same logic everywhere, and as the article says - the user is - as said - only interested in one thing at a time, and wants to be notified discretely about other events without being pulled away from his current task...
I think it would be a very challenging and usefull job to design such a userinterface which is very (to the extreme) consistent in use, but I'm afraid there won't be any interest in making such a thing in the linux/geek world which loves the power to customize...
That wasn't the actual intention of the question - it really was a "btw" - was too lazy to check if they were already present, have to upgrade the "family" pc that still runs debian (and no complaining about that btw), and it runs and keeps running (KDE 3.1.5 btw) :) Also still have a slack running somewhere, and a SuSE also, maybe a good time to upgrade :)
I'll stick to gentoo, and not for it's "optimization", but for it's transparancy in configuration, superior documentation, and fast package updates. Also the freedom to choose is there more than with any other distrib, I mean, Debian with KDE? Wow man - debian-fanboys look at you in a very frightning way... ;)
:p) I tested this with my sister, and she had to ask me a couple of things, some things not very clear to any newbie (partitioning, what filesystem to choose, ...) I actually didn't hear from her anymore till her machine started flipping (bad ram...), and that was 4-5 months later...
:D
The installation can take a while (well - if you use the pre-compiled cd's it's just as fast or even faster than Debian) - that's true, but even if you put a noob behind it with the docs, he should get a working system after a while (though maybe his "windows" may be vanished
I tried a lot of distribs, started with Slack (back in '96), then tried SuSE after a few years, went back to Slack, tried Debian (installation simply sucks), went back to Slack, tried Mandrake somewhere in between (in VMWare) - and then decided to try Gentoo, and I'll stick to that...
KDE 3.2 packages will be "masked" in portage for a few hours prolly cause the package maintainer is asleep I heard
Btw - when can we expect "official" debian/slackware/... packages?
Definitly this game rulez: Soldat :)
:)
It's a 2d platform alike multiplayer game that clearly had some inspiration from worms - but is is oh so cool and fun to play in lan with 4 persons, the gameplay is just fantastic
Truck dismount already had a prequal otherwise it would be my choice
The other games I saw really were sequals or non-original things, so I guess there won't be any other...
1) Quake 3 - and video memory... Do you actually believe a nearly 4 year old game is dependand on video memory? Back in those days, more than 32mb was outrageous. No way that Q3 ever uses more than 64mb - I would even be surprised if it ever reaches that... I even think that recent games will use up the full 128mb. The videocards have exactly the same GPU - which should perform exactly the same in the same machine if the application doesn't need that exceptional amount of memory.
2) The Opteron/Athlon64 beats the crap out of anything when it comes to memory access. It has an on-die memory-controller - what do you expect? Larger files would have made the G5 only look worse. In fact - they also tested with an 150mb image - and the difference became relatively bigger.
3) & 5) Are you suggesting to bench and compare completely different pieces of software? What sense would that make for pure performance comparison? The best way to compare these things would be to write an own dedicated bench, not comparing apples with banana's.
Stripping the OS - what the hell are you thinking? As if Windows is the most optimal OS. As far as I know, MacOS-X has a much better task scheduling than any Windows around (which is - I'm afraid - not that hard)
You also seem to miss the point completely that in fact they ARE comparing kiwi's with banana's cause the Mac requires a dual 2Ghz rig to get even close to a single-Opteron/Athlon64. How fair is that? Ok - the AMD riggs have a slight Mhz advantage in most of the cases, but if you look at this from a neutral position, I don't see how you could not say that the AMD beats the hell out of the G5. Sorry.
He looks at the facts from the point of view of a linux-geek. The thing is - most of the servers online are not owned by linux-freaks - most of them are from a gaming-related company or a clan.
The servers simply run best and require less resources on linux - and linux-serverhosting is cheaper than windows-servers. THAT are the reasons that so many game-servers run linux - not that some geek runs a server for a windows-game just because it runs in linux. At this moment - there are thousands of halflife servers running linux - when there is no native linux-version - would that all be linux-geeks that like it that there is no native version? Don't think so - just the fact that there are already so much HL servers running linux, and the fact that there were already an awfull lot of them long before I was able to even think of starting HL in Wine - or WineX even existed - proves the author wrong.
Another point - about HL2 not running on linux. OpenGL simply doesn't support the necessary features that the HL2 engine uses - unless it is done thu vendor-specific extentions of nVidia or ATI. The HL developers already didn't like it to write a specific rendering path for the last-generation of nVidia cards because they were to slow when running with the generic DX9 rendering path - so they are certainly not going to write 2 separate renderers using opengl for the few customers that want to run it in linux - because - face it - we are a minority in gaming-land. From a developers point of view - I can perfectly understand that (I don't say I like it). Also they don't have any advantages supporting and developing a linux-version - linux is there - and it runs their servers just fine at this moment. Epic had to develop in linux anyway for the PS2 version of their engine. Owkay - they had to write the OpenGL renderer - that's true - but by experience I know it is easyer when you can port a game and run & test the game itself it in the development environment - and have one guy writing the PS2 rendering engine - and let the others work on the port - while able to test it without the necessary equipment - believe me - I do embedded development - but we test and run the main-applications on pc, simulating the hardware - and have one guy (read: me) do the hardware dependand stuff. Yes - that takes a one-time development of the testing-libraries for pc - but in the business I am - I see another type of hardware every half year or so - so it is worth it.
I strongly doubt that we will actually see a PS2 version of HL2 any time soon - so there will be no need for them to port/run the game in linux.
It's putting stuff in a way that *looks* clean but with the large disadvantage that you will be unable to locate something when you really really need it quickly...
So if this code implementing NUMA and RCU is now part of the Sys V codebase - SCO basicly owns code (written by IBM/Sequent) which implement at least 2 patents of IBM?
;) :p
Couldn't IBM sue them for this because SCO doesn't have any licence for these patents but having "their" code commercially used and distributed?
This would be a rather funny trial I think
Notes can't be, but that's not a huge loss in my humble opinion. There are possible better alternatives (notepad, vim, ...)
:)
:)
I use vim on a daily bases, and I can't say that's a better alternative than the notes in opera 7. In opera 7, you select some text on a webpage, add it as a note (using the menu's or the hotkey), the note is saved, so you can look at it afterwards, but (very important but) the thing that makes it a superb feature is that it remembers on which page it originally was, and what position, meaning that you can immediately jump to the page and position on that page where the note originally came from. I don't really see how you would do this in vim or any other text-editor in this comfortable way
And it is like someone already said here somewhere, Opera is imho the most innovative piece of software introducing new handy features with every release. Only the very popular are ported to Mozilla afterwards, which keeps the Opera user with Opera, and the Mozilla users thinking Mozilla has all the features it needs, and has all the features Opera has. Live and let live I would say