This is where "Lights out management" is usefull.
Almost all my servers are HPs with iLO - no need for physical access.
You get separate network port (independed from OS) via witch server can be powered on/off and accessed via virtual serial port.
I'm surprised about this missunderstanding too. There where 3 basic rules: Max of one connection _attempt_ per second per IP (time ban if more) 5 IPs for registrar. One concurrent connection at time.
In perfect world this would be round-robin. However when registry system is loaded it starts to loose connections/timeout/etc. How registrar system behaved on such conditions was very important.
Of course additional accounts changed the picture, and that was discused on EURid mailing lists - however they didn't give a damn about it.
However some registrars moved like turtles and that was mostly their (systems) fault.
King can call any prisoner multiple times, he doesn't even need to use k at all: Say we have 3 prisoners, king would call counter and second prisoner, 3 times both, counter will think it's done (3 flips), but third prisoner wasn't called yet.
Thought I've burned my fingers with IBM "deathstars" (4 died), I'm still fan of IBM now Hitachi. But this idea of using 5x100GB design smells bad: more platters->more heat/parts->more troubles. Considering IBM/Hitachi reputation (only one dealer sells Hitachi drives here) new "deathstar" story would probably kill them.
I'm using PostgreSQL and MySQL, from my experience: 1. I've never encountered corrupted data with mysql (It seems to be urban legend), and I have worked on tables with billions rows for two years. 2. PostgreSQL has more features and/or is more complete (simple example can be auto_increment vs. sequences) 3. PostgreSQL is heavier, and I hate statistics collector subprocess via udp (which seems to be eating 1-2% cpu all the time)* 4. mysql isn't much (if any) faster.
* - it's unlikely but possible my configs are to blame.
I would not agree that polish missing. Problem is underlying technology: how KDE can be hardware friendly when there is (was) no API to control that? Sure we can make KDE daemon which parses dmesg output, but it would be blind and deaf.(hopefully HAL/D-BUS will solve this problem).
How can we solve sound problems to "just plays" when it's completely unclear what vision there is: should apps use ALSA libs directly like they do now? If so, where is abstraction layer like one in DirectSound (you have your buffer, play it, mix it, mute it, but DS will take care of HW and other apps buffers). If not - why ALSA guys don't shout "stop, only daemons/upper layers should use alsa libs directly, not apps"?
It's not polish, but clear lower-end infrastructure/API what is missing.
I have installed and used OO.org many PCs, and on all of them MS Office starts way faster than OO.org. I love OO.org, but these "benchmarks" are simply fake or guys MS Office install is broken in some way.
This is very good news. Yes sarge is, technically, available now, but realease means two important things: security updates, when DSA is issued bug is fixed for stable and unstable versions, thus if one is running sarge at the moment, fixes don't come right away. not realeasing sarge is stopping other development, e.g. there is no KDE 3.4 or X.org in unstable because sarge is not released yet.
Why not: generate random key, encrypt data with it (symmetric), encrypt that key with public one (stored in virus itself), destroy random key, give victim encrypted key. Victim sends encrypted key to author, he decrypts it using his private key and sends it back.
You haven't read article then, big parts are mounted by hand, and then run through soldering machine. In fact, my brother operates such solder, and sees women mounting capacitors/etc. all day long - it is hard and very monotonic job.
It's not dirty, but it's easier to write ugly code using PHP. I've done several rather big sucesfull projects using PHP - using OO design, templates etc. and it's suitable for such thing (you are missing many constructs, but then, you get lots of power because it's weakly typed). I thought for next web project I'm going to use Ruby, yet PHP5 attracts me too.
I've recently tried E17, don't get me wrong, but I don't think you can compare it. It's not eye candy/tricks what is required: these will come with time, the extensible/fast/powerfull architecture - this is the key. Now, E17 is nice, but as far as I could see it just locks root window and draws: shadows aren't drawn on windows beneath, applets are not windows and are always beneath real ones (even when moving), if normal window is moved using alpha blending effect everything else freezes, and so on. When framework is here and working, development will spring up (e.g. nearly every terminal emulator has implemented fake transparency, so demand is here).
High number of viruses/malware should be signal that antivirus software won't be good solution forever.
IMO, instead of brute signature detection, various sandboxes should be used: e.g. one should be able to run any app in virtual environment, any file application deletes or changes is only marked as such in sandbox config, network access should be blocked using custom setup etc.
Thus if I'm happy with application, and start trusting it (by checking what it has tried to do) I merge changes and release it from sandbox, if not - remove it and diffs of any changes it tried to do.
App itself sees merged state at all times, possibilities are endless, like groups of sandboxes etc.
Isn't just evaluating returned data security problem? {It sets up a callback _xmlHttp.onchange event function, that will simply evaluate the (what ends up being a javascript funciton) that gets returned from google...}
Re:Oh Debian, I don't know what to think
on
Updates From Debian
·
· Score: 1
I'm updating my sid (unstable) Debian 5-7 times a week (almost daily).
Nothing serious has happened over past 9 months or so, granted I don't blindly accept "Will REMOVE 200 packages" warnings on upgrades.
Tried (live demo and gallery) on Konqueror 3.3 (from Debian sid) - nothing crashed (my default browser).
I'm pleasantly surprised IE did so well, it's usefull test for mozilla devs to take a look.
Linux Nvidia drivers perform just as good or even better than in windows. There wasn't Quake3 test in this review, but ut2004 was said to run faster. As for Q3, I've tested it myself: runs 5-6% faster.
Slashdot Mirror
This is where "Lights out management" is usefull. Almost all my servers are HPs with iLO - no need for physical access. You get separate network port (independed from OS) via witch server can be powered on/off and accessed via virtual serial port.
It doesn't play Ogg at 100% speed yet, however rockbox does.
Very nice and has simple look.
Icons don't really fit design thought (imho, new icon theme should be created).
I'm surprised about this missunderstanding too. There where 3 basic rules:
Max of one connection _attempt_ per second per IP (time ban if more)
5 IPs for registrar.
One concurrent connection at time.
In perfect world this would be round-robin.
However when registry system is loaded it starts to loose connections/timeout/etc. How registrar system behaved on such conditions was very important.
Of course additional accounts changed the picture, and that was discused on EURid mailing lists - however they didn't give a damn about it.
However some registrars moved like turtles and that was mostly their (systems) fault.
Konqueror 3.5.1 (almost): Acid2 test image
King can call any prisoner multiple times, he doesn't even need to use k at all:
Say we have 3 prisoners, king would call counter and second prisoner, 3 times both, counter will think it's done (3 flips), but third prisoner wasn't called yet.
Thought I've burned my fingers with IBM "deathstars" (4 died), I'm still fan of IBM now Hitachi.
But this idea of using 5x100GB design smells bad: more platters->more heat/parts->more troubles.
Considering IBM/Hitachi reputation (only one dealer sells Hitachi drives here) new "deathstar" story would probably kill them.
I'm using PostgreSQL and MySQL, from my experience:
1. I've never encountered corrupted data with mysql (It seems to be urban legend), and I have worked on tables with billions rows for two years.
2. PostgreSQL has more features and/or is more complete (simple example can be auto_increment vs. sequences)
3. PostgreSQL is heavier, and I hate statistics collector subprocess via udp (which seems to be eating 1-2% cpu all the time)*
4. mysql isn't much (if any) faster.
* - it's unlikely but possible my configs are to blame.
I'm always considering WL-500G or WRT54GS for modest routers: drop openwrt and standart linux system is ready for use.
Granted, they can't push more than 4-6MB/s, but you've got 5-ports + wifi [+ usb + lpt] in small box.
I would not agree that polish missing. Problem is underlying technology: how KDE can be hardware friendly when there is (was) no API to control that?
Sure we can make KDE daemon which parses dmesg output, but it would be blind and deaf.(hopefully HAL/D-BUS will solve this problem).
How can we solve sound problems to "just plays" when it's completely unclear what vision there is: should apps use ALSA libs directly like they do now? If so, where is abstraction layer like one in DirectSound (you have your buffer, play it, mix it, mute it, but DS will take care of HW and other apps buffers).
If not - why ALSA guys don't shout "stop, only daemons/upper layers should use alsa libs directly, not apps"?
It's not polish, but clear lower-end infrastructure/API what is missing.
I have installed and used OO.org many PCs,
and on all of them MS Office starts way faster than OO.org.
I love OO.org, but these "benchmarks" are simply fake or guys MS Office install is broken in some way.
This is very good news.
Yes sarge is, technically, available now, but realease means two important things:
security updates, when DSA is issued bug is fixed for stable and unstable versions, thus if one is running sarge at the moment, fixes don't come right away.
not realeasing sarge is stopping other development, e.g. there is no KDE 3.4 or X.org in unstable because sarge is not released yet.
Why not:
generate random key, encrypt data with it (symmetric),
encrypt that key with public one (stored in virus itself), destroy random key, give victim encrypted key.
Victim sends encrypted key to author, he decrypts it using his private key and sends it back.
You haven't read article then, big parts are mounted by hand, and then run through soldering machine.
In fact, my brother operates such solder, and sees women mounting capacitors/etc. all day long - it is hard and very monotonic job.
It's not dirty, but it's easier to write ugly code using PHP.
I've done several rather big sucesfull projects using PHP - using OO design, templates etc. and it's suitable for such thing (you are missing many constructs, but then, you get lots of power because it's weakly typed).
I thought for next web project I'm going to use Ruby, yet PHP5 attracts me too.
I've recently tried E17, don't get me wrong, but I don't think you can compare it.
It's not eye candy/tricks what is required: these will come with time, the extensible/fast/powerfull architecture - this is the key.
Now, E17 is nice, but as far as I could see it just locks root window and draws: shadows aren't drawn on windows beneath, applets are not windows and are always beneath real ones (even when moving), if normal window is moved using alpha blending effect everything else freezes, and so on.
When framework is here and working, development will spring up (e.g. nearly every terminal emulator has implemented fake transparency, so demand is here).
It appears to be already solved
I'm wondering, why xprint isn't deprecated by now?
High number of viruses/malware should be signal that antivirus software won't be good solution forever.
IMO, instead of brute signature detection, various sandboxes should be used: e.g. one should be able to run any app in virtual environment, any file application deletes or changes is only marked as such in sandbox config, network access should be blocked using custom setup etc.
Thus if I'm happy with application, and start trusting it (by checking what it has tried to do) I merge changes and release it from sandbox, if not - remove it and diffs of any changes it tried to do.
App itself sees merged state at all times, possibilities are endless, like groups of sandboxes etc.
Isn't just evaluating returned data security problem?
{It sets up a callback _xmlHttp.onchange event function, that will simply evaluate the (what ends up being a javascript funciton) that gets returned from google...}
I'm updating my sid (unstable) Debian 5-7 times a week (almost daily). Nothing serious has happened over past 9 months or so, granted I don't blindly accept "Will REMOVE 200 packages" warnings on upgrades.
Tried (live demo and gallery) on Konqueror 3.3 (from Debian sid) - nothing crashed (my default browser). I'm pleasantly surprised IE did so well, it's usefull test for mozilla devs to take a look.
Linux Nvidia drivers perform just as good or even better than in windows.
There wasn't Quake3 test in this review, but ut2004 was said to run faster. As for Q3, I've tested it myself: runs 5-6% faster.