### They each have different goals and should be evaluated based on how well each type of game realizes these goals.
So the goal is to look sucky? Come on, good graphics don't hurt, at worst they might attract a whole bunch of new players. As said, I don't expect OSS games to look like the latest blockbuster, just good enough so that not everybody jumps back and screams "Eww... UGLLLYYYY!!!" on the first try, it doesn't take much to get better then that, but still more then a bunch of OSS projects can offer.
### It takes $$$ and lots of it to create a Halo, a Counterstrike or a Sims. That's just how it is.
Halo and Sims, yes. Counterstrike, absolutly not. Counterstrike started as a simple mod, no big $$$ involved in the beginning, only later it got commercially developed but even then I doubt that the cost where very high. The amount of content required for CounterStrike is rather low, a few quite small levels, a few models, thats it, really nothing if you compare it to a 20h playtime game. They reused the HalfLife engine, which in turn is based on the Quake1 engine (which is together with Quake2 GPL since quite a while) so no money involved there either. Counterstrike did the gameplay right, thats all, I doubt that it needed very much money for that.
Wake me up when we have OSS games with 1'000'000 downloads every few weeks or even every few days and that probally short after the release, not with games that are a decade old or older. As long as it is something that is only happening every few years and with rather ancient games its actually not much a good sign for OSS gaming, not even much of a start. Especially when the quality of OSS games still is rather low, no matter how good the gameplay is of BZFlag, nethack or any other OSS game actually is, the overall impression of the games is often rather low[1]. Nobody can tell me that under all the 1'000'000 downloaders there wasn't a single artists which would have been able to produce better art and improve the overall look of the game a lot. So either the tools for creating art are missing, the project coordination is flawed, the game isn't good enough that anybody cares enough to improve it or maybe those Linux folks are really a whole bunch of non-artists types, but maybe its just that OSS model of games isn't all that much attractive to artists who knows.
Overall OSS gaming still has a long long long way to go, 1'000'000 downloads might sound nice, but don't really tell you much at all about the overall state of OSS gaming.
[1] "low" as in "I have seen better art on my C64", not as in "can't compete with latest multimillion dollor blockbuster game"
To improve security even more I recomment to put the password not under the keyboard but in your wallet. If somebody gets access to your wallet you are screwed quite a bit anyway (money, creditcards, idcards, driverlicense, etc.). Its of course a bit less convinient then under the keyboard, but quite a bit more secure for sure.
Sure you might say that 8 characters are easy to remember and if you form sentences with them you have some good help remembering them. However this is only true if you have to remember few passwords and use them frequently. As soon as you start to have lots of passwords (remember internet, where each and every webforum wants to have a password of its own), some of which you don't use all that frequently, you are basically lost. For sure you will forget at least some of them, people simply arn't build for long term memorizion of long obscure combinations of letters or sentences, simply doesn't work.
The throuble is really that hardware hasn't catched up, or well it has, but not on the common mans computer. Things like HBCI demonstrate that a relativly easy to use secure way to handle logins is possible. Why not have a smartcard to handle all your logins to all webpages and stuff on the internet? Have the smartcard protected with one passphrase, not a different passphrase for each webpage and people would have much less throuble keeping stuff secure. The card would simply something to be carried arround like a key. Sure, you might need a dedicated reader for them to have it really secure (ie. protected against keyloggers), but if produced for the masses it really shouldn't be that difficult to make it rather cheap, with new PCs it could even be build in.
Technically its really not that difficult to allow people to have a secure way to handle logins, the throuble is only that people would need to agree on a standard, which is as always way more difficult then to build such a thing in the first place.
1) Disc space will most likly be to limited to be full of OSS software for Windows and for Linux
2) There is already the Ubuntu live cd, which contains a full Ubuntu Linux live distri and in addition to that a bunch of popular OSS software for windows, but due to 1) most likly quite a bit less then the OpenCD.
Re:While this is great for open source advocates..
on
TheOpenCD 2.0 Released
·
· Score: 1
### what open source really needs is a marketing plan.
What OpenSource needs is better software, not a marketing plan. You can't 'talk' people into using OpenSource Software when they have already purcased superior non-free software. Lets face it, in lots of areas OpenSource is simply inferior to proprietary software and will stay so for a long time to come, better marketing won't change that, IMHO it already is over-hyped way to much. What OpenSource needs is better software, doesn't even need to be better then the proprietary software, even so that would be of course a great plus, it just needs to be a little bit better than "good enough" and people will switch automatically on the next updated when they can save a few bugs and still get there work done equally easy. OSS isn't there yet, but slowly moving into the right direction, just give it some time and keep the hype low.
Spoiling people now with marketing lies (Gimp: Your Photoshop killer for 0$, OpenOffice:...and you will never touch MS Office again, Inkscape: Revolutionary Vector Graphics...) will just make it much harder to get people to use OSS software tomorrow when it finally might be 'good enough' for the average user.
The way some webpages use Flash is evil, Flash itself is actually quite usefull and provides something that you simply can't do today by other means or at least not without wasting way more bandwidth. Maybe SVG will catch up in another decade or so, but until then Flash will be the tool of choice when it comes to interactive vector graphic web application.
### And where are those people? Almost every single post flames Gimp down for being a "toy" or "totally unusable".
Well, the FOSS community is making progress latly and getting closer to reality. Look back a few years, there was much less flaming and far more 'Gimp is Photoshop killer'-hype. Why? Because back then FOSS community was with itself, not much outsiders involved and hardly anybody even know Photoshop from day to day experience, today however people outside the FOSS community actually try to use Gimp for serious work and many of them fail and fallback to Photoshop.
16bit support isn't a prepress feature, its a feature that everybody who ever does a little brightness correction or blur would love. 16bit isn't usefull for the final image, but it reduces clipping artefacts that you get with 8bit color in the editng process a lot. There simply is no way to get some effects when you work with just 8bit integer.
Look for example at http://www.debevec.org/Research/HDR/ the second image is what Gimp can give you, the fourth is what the reality looks like, the third is what Gimp would be able to do with support for higher color ranges. Quite a different, isn't it?
### However, in the real world, any non-trivial app is going to take at least a couple of seconds to launch.
Because most apps are pretty shitty written. In times where GEdit, the thing that is similar to notepad, starts a whole lot SLOWER then my Emacs, thing are really fucked up alot. These days Emacs is actually one of my fasted starting apps. Hardly any of the Gnome/KDE apps gets away with less then two or three seconds boottime, and thats on a 1GHz machine, anything beside 'instant startup' is really not much acceptable. You think complicated apps needs to have slow startup time? Look at Blender, the app that can cut videos, render pictures, edit text, paint pictures and do a heapload of other things, how long does it take to startup, I don't know because its almost instantly, less then a second.
Startup times really seem to be an issue that nobody of the programmers cares about, but even for complicated applications there are ways to boost startup speed or at least hide it from the user. In the old days there where programms that basically saved a screenshot of the GUI and showed that to the user at startup, it was indistingushiable from the real interface, but had the advantage that it was instantly there, so while the user took a few seconds adopt to the interface (today one would move the windows into the right positions, on another workspace or whatever) the app silently loaded in the background and replaced the screenshot-gui silently. Hardly any user ever noticed that, it feld as if the app was there instantly.
Choice might be good sometime, but often its not. The "Choice is Good" is half of the time really just an excuse for bad userinterfaces, bad defaults or otherwise flawed software. Gnome tries to offer choice only where it makes sense and is necessary, where it does not they either not provide an option at all or move it into gconf.
### If users are feeling overwhelmed by options, it's generally a sign that the choices have been poorly presented, not that there are too many of them.
Depends on the options, when they do something usefull that a large number of people will use, then it should of course be there. If it however is something that basically none will be using and the only thing you can do with the option is to break things, it should better be removed. Same applies to options where the computer can automatically determinate the correct value, ie. no need to set you cdrom device in each and every application, since the OS provides enough info to find out where it is.
That said, Gnome overdid it a bit with the "remove options" in the past, since they either ended up moving common options into gconf or only provided flawed defaults, its after all a act of balance and only a little try and error will show if some option is really needed and if it is if there isn't another way to provide a 'right' default.
### Nautilus on the whole is VERY confusing to the users that I have introduced to it.
The throuble really is that spatial only works on very flat directory trees, there however it works great. I for one loved the spatial nature of Workbench back then on the Amiga, it made perfect sense there since directories where always pretty flat, after all I didn't had a HD, just a single floppy drive and directories where basically never deeper then two levels.
Today on a PC however its much more throublesome, the whole world in form of the internet is just a few clicks away and even my local PC has directory structures with many many levels. Even getting to the top-level directory of a smb-share takes around four levels and my screen is then already full of windows and thats only the start, the file I want might be a whole bunch of levels deeper. Spatial doesn't work in such situations and if it does only with the Ctrl-Click trick that auto-closes the current window, which I personally find hell of a lot confusing since the new window pops up completly at a random position. Yes, I know the position is the same as last one but on deep directories there is no way I can remember all the position, so it looks and feels completly random.
In the end I would say that spatial is really not the one-true-way how browsing will work in the future, but neither is the explorer-style all that perfect. My current favorite is really the Rox-style of handling things, at default I get everything in one window, but if I middle click I basically get a spatial window and the best of all I can press '/' and then get a text-input field which lets me tab-completed browse the file structure, which happens to be a damn lot faster then mousing around in some areas. Nautilus Ctrl-L is similar, but due to the spatial 'mess', it feels rather different.
BTW. The NortonCommander 'double' view is something I consider relativly useless, but thats a different story...
### I'll take this oppertunity to complain about GNOME's current love affair with spatial browsing, in the hope that it will get noticed.
In Gnome2.8.1 there is a easy to reach option to switch back to the normal non-spatial browsing behaviour, so no more gconf searching for the right option. About making it default, its of course questionable, however spatial has its benefits when your directory structures are flat, which it most likly will be for most new users, rest of the users shouldn't have much throuble to switch back to the old behaviour now.
Complain a lot, write bugreports, cross your fingers and wait a release or two and they might add back a useable textinput/typeahead support. In the past Gnome developers have frustrated me quite a lot, especially in the switch from Gnome1.4 to Gnome2.0 where a lot of usefull features have gone missing, however most of the needed features have found there way back again sooner or later. So I have good hopes that they will fix the filedialog too in the future, just give it a bit time. Gnome developers tend to overshoot their goal of simplicity, it just takes some time to find the right balance between 'crowded', 'simply good' and 'too simple'.
### I can't understand why they won't even offer the old one as an option, except that it would mean admitting that they might be wrong.
They follow more or less the principal of doing it right, instead of flooding the screen with options. And as basically everybody will agree the old dialog was just plain awfull (beside the tab-completion, which was really good), so I think they prefered to dump it completly to have it finally dead, instead of dragging it around for another few releases. Until they get proper typeahead implemented, it will be of course a bit painfull, since 'Ctrl-L' is really a rather ugly hack, however it gets the job done and the dialog is already much more pleasent to use with the mouse, so the damage isn't that big and time will most likly fix the rest.
### Unless you know the source is small by some other means, the fact that it is small makes no difference to how easy it is to reverse the hash.
If I run a dictonary attack on that 'hello'-hash I will quite quickly get the result that 'hello' is a input that matches to the hash. You are right that there are unlimited other values that match to the same hash, so I don't have actually reversed it, but only found a value that matches to the same one. However since its almost impossible to find a hash-collision in md5 by pure luck, I would say that the chance that I found the original input to the hash function is rather high. And for many common cases having a input that matches to the same hash is as good as having the original input, ie. for passwords it doesn't matter if I know the real password, its already enough when I have a password that matches to the hash.
Should be easily reversible with a dictonary attack. However in the more common case a hash maps a large input domain, to a much smaller output domain, so yep, hashes are not reversible unless input is somehow smaller then the hash itself.
First of, if there is a backdoor it shouldn't be that hard to fix and it should also be easier to locate when you have the source, with closed source you are neither allowed to fix it and will have a much harder time to find it in the first place.
Second, what makes you believe that there are backdoors in open source any more than in closed source software? Open Source isn't open for everybody, write access to the repository is limited to a small number of people, commit logs get posted to mailing lists and can be easily be verified by other people there. Open source people don't blindly apply every patch they get any more than closed source people do.
### All the low level bit and byte override overflow exploits are very difficult to understand for regular people and harder to protect against.
Yes, but that has nothing todo with open or closed sound.
Open source doesn't gurantee you anything sure, but neither does closed source, just because your contractor says its 'bug and backdoor free' doesn't mean that it is.
The biggest problems is that Nintendos games never really where much strong on storylines or character development, they where however good at gameplay, but thats something that doesn't benefit you much. So I can hardly see how one could turn any Nintendo franchise into a decent movie while still staying something close to the franchise itself. The Mario movie for example had a guy named Mario and a guy named Luigi, but thats all the connection it had to the game.
Better thousands of ours then tens/hundred of thousands of the others. If one dead soldier of ours can save ten dead on the other side then its worth it. Military is already droping bombs on places without checking first what exactly is going on there and thus bombing a wedding every once in a while. And all that just because they didn't want to risk a few soldiers.
Rule of thumb is really to not use Debian stable on system where you acutally want to work on. Debian stable is worth thinking about if you are running a bunch of servers or desktop systems and want to keep them in sync or a router which should only run one or two programms, but even then having a unstable testsystem and rsync might do a better job, since you can't really use a stable without inserting a bunch of unofficial and self backported packages first which in turn make 'stable' not so much 'stable' any more.
Debian stable has a long long history of being full of horrible outdated packages, the pure stable user is still running Gimp1.2 and Kde2.2 after all. I personally have simply given up on Debian stable, it wasn't much useable in the past and probally won't much useable in the future. Hopefully Ubuntu will do a better job at providing a system that I can actually recomment to people, but Debian stable, a few isolated cases asside, is really not much usefull.
The worst of all is really that Debian is completly working against their target audience, sure there are a handfull of people which are using Debian stable and where it makes sense, but I guess there are a hell of a lot more people which are using testing and thus don't get any security updates much in time. The current Debian release cycle does really not work much at all, and simply redirecting people to testing or even unstable is not the solution its the PROBLEM.
Best part of it is when Gimp developer say something along the line of "We are not cloning Photoshop.", while its pretty obvious that it had at least some major inspiration from the Photoshop interface.
Anyway, the point of improving the UI should really not be about getting it closer to Photoshop, the goal should be to make it simply good, if that means to move closer to Photoshop in some areas and further away from it in others, then so be it.
One thing for example that is currently really lacking is a 'toolbar', that thing that you see at the top of the screen. In Gimp all functionallity that would normally be placed in such a toolbar (see Inkscape for example) is hidden in the 'Tool Options' dialog, which like all dialogs is floating wildly around on your screen and gets easily hidden under some window, a fixed placed toolbar would not. Might look like a simple thing, but playing 'hide and seek' with dialog windows is one of the things that make Gimp such a major pain. Things like having a button for automatically aranging dialog windows might of course help to. Oh well, there is a whole bunch of stuff that needs fixing, might be worth to start a Wiki for UI improvment ideas or so.
Slashdot Mirror
### They each have different goals and should be evaluated based on how well each type of game realizes these goals.
So the goal is to look sucky? Come on, good graphics don't hurt, at worst they might attract a whole bunch of new players. As said, I don't expect OSS games to look like the latest blockbuster, just good enough so that not everybody jumps back and screams "Eww... UGLLLYYYY!!!" on the first try, it doesn't take much to get better then that, but still more then a bunch of OSS projects can offer.
### It takes $$$ and lots of it to create a Halo, a Counterstrike or a Sims. That's just how it is.
Halo and Sims, yes. Counterstrike, absolutly not. Counterstrike started as a simple mod, no big $$$ involved in the beginning, only later it got commercially developed but even then I doubt that the cost where very high. The amount of content required for CounterStrike is rather low, a few quite small levels, a few models, thats it, really nothing if you compare it to a 20h playtime game. They reused the HalfLife engine, which in turn is based on the Quake1 engine (which is together with Quake2 GPL since quite a while) so no money involved there either. Counterstrike did the gameplay right, thats all, I doubt that it needed very much money for that.
Wake me up when we have OSS games with 1'000'000 downloads every few weeks or even every few days and that probally short after the release, not with games that are a decade old or older. As long as it is something that is only happening every few years and with rather ancient games its actually not much a good sign for OSS gaming, not even much of a start. Especially when the quality of OSS games still is rather low, no matter how good the gameplay is of BZFlag, nethack or any other OSS game actually is, the overall impression of the games is often rather low[1]. Nobody can tell me that under all the 1'000'000 downloaders there wasn't a single artists which would have been able to produce better art and improve the overall look of the game a lot. So either the tools for creating art are missing, the project coordination is flawed, the game isn't good enough that anybody cares enough to improve it or maybe those Linux folks are really a whole bunch of non-artists types, but maybe its just that OSS model of games isn't all that much attractive to artists who knows.
Overall OSS gaming still has a long long long way to go, 1'000'000 downloads might sound nice, but don't really tell you much at all about the overall state of OSS gaming.
[1] "low" as in "I have seen better art on my C64", not as in "can't compete with latest multimillion dollor blockbuster game"
Why break into a biometric database, I for one leave my finger-prints all over the place, doesn't take much to grab it from the wall.
Biometric solutions might be a nice addition to a pin, ie. both bio and pin must match to get login, but bio alone doesn't help much.
To improve security even more I recomment to put the password not under the keyboard but in your wallet. If somebody gets access to your wallet you are screwed quite a bit anyway (money, creditcards, idcards, driverlicense, etc.). Its of course a bit less convinient then under the keyboard, but quite a bit more secure for sure.
Sure you might say that 8 characters are easy to remember and if you form sentences with them you have some good help remembering them. However this is only true if you have to remember few passwords and use them frequently. As soon as you start to have lots of passwords (remember internet, where each and every webforum wants to have a password of its own), some of which you don't use all that frequently, you are basically lost. For sure you will forget at least some of them, people simply arn't build for long term memorizion of long obscure combinations of letters or sentences, simply doesn't work.
The throuble is really that hardware hasn't catched up, or well it has, but not on the common mans computer. Things like HBCI demonstrate that a relativly easy to use secure way to handle logins is possible. Why not have a smartcard to handle all your logins to all webpages and stuff on the internet? Have the smartcard protected with one passphrase, not a different passphrase for each webpage and people would have much less throuble keeping stuff secure. The card would simply something to be carried arround like a key. Sure, you might need a dedicated reader for them to have it really secure (ie. protected against keyloggers), but if produced for the masses it really shouldn't be that difficult to make it rather cheap, with new PCs it could even be build in.
Technically its really not that difficult to allow people to have a secure way to handle logins, the throuble is only that people would need to agree on a standard, which is as always way more difficult then to build such a thing in the first place.
1) Disc space will most likly be to limited to be full of OSS software for Windows and for Linux
2) There is already the Ubuntu live cd, which contains a full Ubuntu Linux live distri and in addition to that a bunch of popular OSS software for windows, but due to 1) most likly quite a bit less then the OpenCD.
### what open source really needs is a marketing plan.
...and you will never touch MS Office again, Inkscape: Revolutionary Vector Graphics...) will just make it much harder to get people to use OSS software tomorrow when it finally might be 'good enough' for the average user.
What OpenSource needs is better software, not a marketing plan. You can't 'talk' people into using OpenSource Software when they have already purcased superior non-free software. Lets face it, in lots of areas OpenSource is simply inferior to proprietary software and will stay so for a long time to come, better marketing won't change that, IMHO it already is over-hyped way to much. What OpenSource needs is better software, doesn't even need to be better then the proprietary software, even so that would be of course a great plus, it just needs to be a little bit better than "good enough" and people will switch automatically on the next updated when they can save a few bugs and still get there work done equally easy. OSS isn't there yet, but slowly moving into the right direction, just give it some time and keep the hype low.
Spoiling people now with marketing lies (Gimp: Your Photoshop killer for 0$, OpenOffice:
The way some webpages use Flash is evil, Flash itself is actually quite usefull and provides something that you simply can't do today by other means or at least not without wasting way more bandwidth. Maybe SVG will catch up in another decade or so, but until then Flash will be the tool of choice when it comes to interactive vector graphic web application.
### And where are those people? Almost every single post flames Gimp down for being a "toy" or "totally unusable".
Well, the FOSS community is making progress latly and getting closer to reality. Look back a few years, there was much less flaming and far more 'Gimp is Photoshop killer'-hype. Why? Because back then FOSS community was with itself, not much outsiders involved and hardly anybody even know Photoshop from day to day experience, today however people outside the FOSS community actually try to use Gimp for serious work and many of them fail and fallback to Photoshop.
16bit support isn't a prepress feature, its a feature that everybody who ever does a little brightness correction or blur would love. 16bit isn't usefull for the final image, but it reduces clipping artefacts that you get with 8bit color in the editng process a lot. There simply is no way to get some effects when you work with just 8bit integer.
Look for example at http://www.debevec.org/Research/HDR/ the second image is what Gimp can give you, the fourth is what the reality looks like, the third is what Gimp would be able to do with support for higher color ranges. Quite a different, isn't it?
About the other ones you are perfectly right, but please not this one, its my absolute favorite =:)
### However, in the real world, any non-trivial app is going to take at least a couple of seconds to launch.
Because most apps are pretty shitty written. In times where GEdit, the thing that is similar to notepad, starts a whole lot SLOWER then my Emacs, thing are really fucked up alot. These days Emacs is actually one of my fasted starting apps. Hardly any of the Gnome/KDE apps gets away with less then two or three seconds boottime, and thats on a 1GHz machine, anything beside 'instant startup' is really not much acceptable. You think complicated apps needs to have slow startup time? Look at Blender, the app that can cut videos, render pictures, edit text, paint pictures and do a heapload of other things, how long does it take to startup, I don't know because its almost instantly, less then a second.
Startup times really seem to be an issue that nobody of the programmers cares about, but even for complicated applications there are ways to boost startup speed or at least hide it from the user. In the old days there where programms that basically saved a screenshot of the GUI and showed that to the user at startup, it was indistingushiable from the real interface, but had the advantage that it was instantly there, so while the user took a few seconds adopt to the interface (today one would move the windows into the right positions, on another workspace or whatever) the app silently loaded in the background and replaced the screenshot-gui silently. Hardly any user ever noticed that, it feld as if the app was there instantly.
Choice might be good sometime, but often its not. The "Choice is Good" is half of the time really just an excuse for bad userinterfaces, bad defaults or otherwise flawed software. Gnome tries to offer choice only where it makes sense and is necessary, where it does not they either not provide an option at all or move it into gconf.
### If users are feeling overwhelmed by options, it's generally a sign that the choices have been poorly presented, not that there are too many of them.
Depends on the options, when they do something usefull that a large number of people will use, then it should of course be there. If it however is something that basically none will be using and the only thing you can do with the option is to break things, it should better be removed. Same applies to options where the computer can automatically determinate the correct value, ie. no need to set you cdrom device in each and every application, since the OS provides enough info to find out where it is.
That said, Gnome overdid it a bit with the "remove options" in the past, since they either ended up moving common options into gconf or only provided flawed defaults, its after all a act of balance and only a little try and error will show if some option is really needed and if it is if there isn't another way to provide a 'right' default.
### Nautilus on the whole is VERY confusing to the users that I have introduced to it.
The throuble really is that spatial only works on very flat directory trees, there however it works great. I for one loved the spatial nature of Workbench back then on the Amiga, it made perfect sense there since directories where always pretty flat, after all I didn't had a HD, just a single floppy drive and directories where basically never deeper then two levels.
Today on a PC however its much more throublesome, the whole world in form of the internet is just a few clicks away and even my local PC has directory structures with many many levels. Even getting to the top-level directory of a smb-share takes around four levels and my screen is then already full of windows and thats only the start, the file I want might be a whole bunch of levels deeper. Spatial doesn't work in such situations and if it does only with the Ctrl-Click trick that auto-closes the current window, which I personally find hell of a lot confusing since the new window pops up completly at a random position. Yes, I know the position is the same as last one but on deep directories there is no way I can remember all the position, so it looks and feels completly random.
In the end I would say that spatial is really not the one-true-way how browsing will work in the future, but neither is the explorer-style all that perfect. My current favorite is really the Rox-style of handling things, at default I get everything in one window, but if I middle click I basically get a spatial window and the best of all I can press '/' and then get a text-input field which lets me tab-completed browse the file structure, which happens to be a damn lot faster then mousing around in some areas. Nautilus Ctrl-L is similar, but due to the spatial 'mess', it feels rather different.
BTW. The NortonCommander 'double' view is something I consider relativly useless, but thats a different story...
### I'll take this oppertunity to complain about GNOME's current love affair with spatial browsing, in the hope that it will get noticed.
In Gnome2.8.1 there is a easy to reach option to switch back to the normal non-spatial browsing behaviour, so no more gconf searching for the right option. About making it default, its of course questionable, however spatial has its benefits when your directory structures are flat, which it most likly will be for most new users, rest of the users shouldn't have much throuble to switch back to the old behaviour now.
Complain a lot, write bugreports, cross your fingers and wait a release or two and they might add back a useable textinput/typeahead support. In the past Gnome developers have frustrated me quite a lot, especially in the switch from Gnome1.4 to Gnome2.0 where a lot of usefull features have gone missing, however most of the needed features have found there way back again sooner or later. So I have good hopes that they will fix the filedialog too in the future, just give it a bit time. Gnome developers tend to overshoot their goal of simplicity, it just takes some time to find the right balance between 'crowded', 'simply good' and 'too simple'.
### I can't understand why they won't even offer the old one as an option, except that it would mean admitting that they might be wrong.
They follow more or less the principal of doing it right, instead of flooding the screen with options. And as basically everybody will agree the old dialog was just plain awfull (beside the tab-completion, which was really good), so I think they prefered to dump it completly to have it finally dead, instead of dragging it around for another few releases. Until they get proper typeahead implemented, it will be of course a bit painfull, since 'Ctrl-L' is really a rather ugly hack, however it gets the job done and the dialog is already much more pleasent to use with the mouse, so the damage isn't that big and time will most likly fix the rest.
### Unless you know the source is small by some other means, the fact that it is small makes no difference to how easy it is to reverse the hash.
If I run a dictonary attack on that 'hello'-hash I will quite quickly get the result that 'hello' is a input that matches to the hash. You are right that there are unlimited other values that match to the same hash, so I don't have actually reversed it, but only found a value that matches to the same one. However since its almost impossible to find a hash-collision in md5 by pure luck, I would say that the chance that I found the original input to the hash function is rather high. And for many common cases having a input that matches to the same hash is as good as having the original input, ie. for passwords it doesn't matter if I know the real password, its already enough when I have a password that matches to the hash.
### If it's a hash, the message CANNOT be reconstructable from it
Depends on the hash function, if the input length is equal or smaller then output length reconstructing the message might be possible, ie:
$ echo "hello" | md5sum
b1946ac92492d2347c6235b4d2611184 -
Should be easily reversible with a dictonary attack. However in the more common case a hash maps a large input domain, to a much smaller output domain, so yep, hashes are not reversible unless input is somehow smaller then the hash itself.
First of, if there is a backdoor it shouldn't be that hard to fix and it should also be easier to locate when you have the source, with closed source you are neither allowed to fix it and will have a much harder time to find it in the first place.
Second, what makes you believe that there are backdoors in open source any more than in closed source software? Open Source isn't open for everybody, write access to the repository is limited to a small number of people, commit logs get posted to mailing lists and can be easily be verified by other people there. Open source people don't blindly apply every patch they get any more than closed source people do.
### All the low level bit and byte override overflow exploits are very difficult to understand for regular people and harder to protect against.
Yes, but that has nothing todo with open or closed sound.
Open source doesn't gurantee you anything sure, but neither does closed source, just because your contractor says its 'bug and backdoor free' doesn't mean that it is.
The biggest problems is that Nintendos games never really where much strong on storylines or character development, they where however good at gameplay, but thats something that doesn't benefit you much. So I can hardly see how one could turn any Nintendo franchise into a decent movie while still staying something close to the franchise itself. The Mario movie for example had a guy named Mario and a guy named Luigi, but thats all the connection it had to the game.
Better thousands of ours then tens/hundred of thousands of the others. If one dead soldier of ours can save ten dead on the other side then its worth it. Military is already droping bombs on places without checking first what exactly is going on there and thus bombing a wedding every once in a while. And all that just because they didn't want to risk a few soldiers.
If you are the one that started the war, then yep, its a bad argument.
Rule of thumb is really to not use Debian stable on system where you acutally want to work on. Debian stable is worth thinking about if you are running a bunch of servers or desktop systems and want to keep them in sync or a router which should only run one or two programms, but even then having a unstable testsystem and rsync might do a better job, since you can't really use a stable without inserting a bunch of unofficial and self backported packages first which in turn make 'stable' not so much 'stable' any more.
Debian stable has a long long history of being full of horrible outdated packages, the pure stable user is still running Gimp1.2 and Kde2.2 after all. I personally have simply given up on Debian stable, it wasn't much useable in the past and probally won't much useable in the future. Hopefully Ubuntu will do a better job at providing a system that I can actually recomment to people, but Debian stable, a few isolated cases asside, is really not much usefull.
The worst of all is really that Debian is completly working against their target audience, sure there are a handfull of people which are using Debian stable and where it makes sense, but I guess there are a hell of a lot more people which are using testing and thus don't get any security updates much in time. The current Debian release cycle does really not work much at all, and simply redirecting people to testing or even unstable is not the solution its the PROBLEM.
Best part of it is when Gimp developer say something along the line of "We are not cloning Photoshop.", while its pretty obvious that it had at least some major inspiration from the Photoshop interface.
Anyway, the point of improving the UI should really not be about getting it closer to Photoshop, the goal should be to make it simply good, if that means to move closer to Photoshop in some areas and further away from it in others, then so be it.
One thing for example that is currently really lacking is a 'toolbar', that thing that you see at the top of the screen. In Gimp all functionallity that would normally be placed in such a toolbar (see Inkscape for example) is hidden in the 'Tool Options' dialog, which like all dialogs is floating wildly around on your screen and gets easily hidden under some window, a fixed placed toolbar would not. Might look like a simple thing, but playing 'hide and seek' with dialog windows is one of the things that make Gimp such a major pain. Things like having a button for automatically aranging dialog windows might of course help to. Oh well, there is a whole bunch of stuff that needs fixing, might be worth to start a Wiki for UI improvment ideas or so.