Now I can cause Dukes of Hazzard reruns to be continually interrupted with flash news bulletins in full colour of the accidents I cause while driving and watching Dukes of Hazzard reruns!
This looks like a great tool - even if you normally use a packaged distro, you can get your hands dirty and learn how it all hangs together! It's the sort of thing I'd highly recommend anyone doing if you have the time.
I started with a Slackware distro using the 0.96 kernel, largely archived onto floppies and had to feel my way through making it boot, so I consider myself as having learnt not quite from scratch. But today's folks have it easy, thanks to people who already understand what's under the bonnet.
I bet most people who LFS will gain a better appreciation for configure-and-compile-yourself. Three cheers!
I work for a company which has a single PowerEdge 1850 for trial purposes. We've had a couple of "SMP" problems on this x86_64 box (hint: Can you explain why Red Hat's x86_64.smp kernel panics on boot - at least versions prior to Update 3?), neither of which were explained or fixed by Dell.
On top of this, we have to pass all first and second level RH EL3 calls through Dell. It's such a pity that their engineers appear to have little grasp of the updates to technology, which I suspect is due to lack of training. It may be that Dell need to invest more money in keeping their engineers up to date supporting the software they supply with the hardware they assemble?
On another note, Red Hat haven't particularly impressed anyone with their overall support expertise either (though I've had a couple of quick and useful answers). My first three support calls were given, in order, a completely wrong answer, a succinct and correct answer, and a 6 month plus wait which trailed off to nothingness...
... is when the power's out and the lights are not working at all. Then it seems that everyone approaches the lights at a safe speed, gives way as appropriate, and go when safe. The intersections in question are not usually heavily congested though.
Honestly, if people would all show enough common sense and courtesy on busy roads, we wouldn't need at least half the sets of lights in existence.
Having said that, there are some very useful mechanisms for controlling traffic flow. Like the green arrow to turn across what could otherwise be the oncoming flow of traffic from the other direction - if vehicles stop passing the sensor, or a certain amount of time passes, the arrow changes and oncoming traffic is given the green signal. It gets people out of the turning lane that would otherwise just get in the way.
Now if only we could make people stay in the slow lane unless overtaking, and learn how to merge (You know who you are)!
I saw a small science piece some years ago, not sure if it was a science oriented show or an interest piece on the news, about a "human bat". This guy did his own version of sonar by making clicking sounds with his tongue, and his ears were attuned well enough for him to reliably echo sound. He could even ride his bike on the road, navigating around obstacles, as long as the traffic wasn't too heavy...
That's as I recall - anyone else out there have any further details?
If you want free, flexible and functional, the best of all worlds with emphasis on free is PostgreSQL. I'm interested in seeing what Ingres has that Postgres doesn't already. My guess is not much.
When I joined the company I'm currently with, I found it hard to believe they were using MySQL instead of it, but it looks like MySQL got the foothold first.
Stick PostgreSQL on a Reiser 4 filesystem with full journalling enabled, and barring hardware problems your database server should be as solid as a rock.
Nothing against Ingres here, since it worked for me at Uni in 1994, but it's hard to go past the best of the open source when you're already familiar with it.
I guess the whole Mozilla project could afford the (temporary but oh so long) incompatabilities it suffered during the rewrite. There was one huge advantage though:
They adhere to standards, not trying to keep up with custom hacks to support the whims of the OS. I remember authoring with CSS, having to be very careful not to use certain tags that did exactly what I wanted in one browser, but looked crap in another - compatibility was paramount, and forced me to stick with a rather poor subset of the rich definitions.
The gecko engine is an excellent piece of work - it's great that Netscape can now slap on a different look, add a dictionary, and release it with no need to mess around with the internals. That must have been a bane for Netscape while they had it closed. The changes from the days of Netscape 4.x are a great demonstration of open source achieving its goals and improving the code base. Security bugs are inevitable but infrequent, and the browser seems to keep on getting smaller and faster, yet more feature rich with every release!
Microsoft will never achieve that with closed model, no matter how brilliant their programmers are, because they aren't exposing their source to general criticism, and better ideas in the community. I went to Uni with one of the developers at Redmond. He has a brilliant mind, but knowing him, in the end he'll just write according to the other peoples' expectations, no matter how flawed that may be.
I found this article a bit wishy-washy. Yes, a couple of good points are made, but the application depends on more than just the development model. IE is but one application, and it seems to be concentrated on a bit much - you can run Firefox on Windows too! Remember when Netscape went open source? It's been pretty much re-written since then, as it should have been. Does anybody know if IE has ever had a ground-up rebuild since the Win95? From what I understand, the answer is no, though someone may correct me here.
This gives an interesting insight into the open source model through taking over an open source project. It presents lessons learnt, and corresponding cardinal rules when running such a project. It also outlines quite effectively why open source is a viable means to develop quality software, despite the author's initial reservations. In C or C++ even.
Ok, first, I'm no Microsoft fan, but if you are a Microsoft shop, and you want commercial support, there's little choice but to use SQL Server. It's a logical choice given the circumstances.
Second, I've seen a bit of Oracle, and watched our DBAs tearing hair out over mismatches in certifications (we are forced to use RH AS2.1 for their iLearning product where I work, where everything else happily runs on EL3). However, if you're not trying to make a whole bunch of suites of Oracle software work together then it all looks ok. I must agree that they definitely have done databases right. It's hard to go past RAC for ensuring reliability!
My previous job, I was (network|system|database|web) administrator, and (web|software) (developer|troubleshooter) in the early days, until we were able to expand the IT base a bit. That introduced me to PostgreSQL. I will always remember those days fondly. After a few years away I'm certainly very rusty, but there's nothing to stop me printing up the manuals and getting back on the rails if I ever get the time.
With our implementation of PostgreSQL we had almost no downtime that wasn't caused by the boss, and experienced one whole inconsistency issue. That issue was patched within 48 hours of being reported. And our hideously inefficient queries just kept chugging along when I felt sure the lack of grunt would cause SOMETHING to snap... IMHO MySQL is still eating its dust in features and reliability. Stick it on reiser4 filesystems, then I reckon virtually all risk becomes genuine hardware failure.
But hang on... it's not just open source. You can buy a license if you want commercial support! You can even run it on Windows! Not that I would, but... wow, right?
The only thing I can say for CA is I hope they maintain the back end better than they create their front ends. The company I work for had CA Unicenter, and since the license expired we switched to an Oracle application and haven't regretted it. It's good to keep the current vendors on their toes, but somehow I doubt they'll grab a good market share. We will see...
Slashdot Mirror
I told the google execs to patent all that, but would they listen? Noooooo!
Now I can cause Dukes of Hazzard reruns to be continually interrupted with flash news bulletins in full colour of the accidents I cause while driving and watching Dukes of Hazzard reruns!
I started with a Slackware distro using the 0.96 kernel, largely archived onto floppies and had to feel my way through making it boot, so I consider myself as having learnt not quite from scratch. But today's folks have it easy, thanks to people who already understand what's under the bonnet.
I bet most people who LFS will gain a better appreciation for configure-and-compile-yourself. Three cheers!
On top of this, we have to pass all first and second level RH EL3 calls through Dell. It's such a pity that their engineers appear to have little grasp of the updates to technology, which I suspect is due to lack of training. It may be that Dell need to invest more money in keeping their engineers up to date supporting the software they supply with the hardware they assemble?
On another note, Red Hat haven't particularly impressed anyone with their overall support expertise either (though I've had a couple of quick and useful answers). My first three support calls were given, in order, a completely wrong answer, a succinct and correct answer, and a 6 month plus wait which trailed off to nothingness...
Of course, the best way to do this is throw gold dust at them... lots of gold dust.
Honestly, if people would all show enough common sense and courtesy on busy roads, we wouldn't need at least half the sets of lights in existence.
Having said that, there are some very useful mechanisms for controlling traffic flow. Like the green arrow to turn across what could otherwise be the oncoming flow of traffic from the other direction - if vehicles stop passing the sensor, or a certain amount of time passes, the arrow changes and oncoming traffic is given the green signal. It gets people out of the turning lane that would otherwise just get in the way.
Now if only we could make people stay in the slow lane unless overtaking, and learn how to merge (You know who you are)!
That's as I recall - anyone else out there have any further details?
When I joined the company I'm currently with, I found it hard to believe they were using MySQL instead of it, but it looks like MySQL got the foothold first.
Stick PostgreSQL on a Reiser 4 filesystem with full journalling enabled, and barring hardware problems your database server should be as solid as a rock.
Nothing against Ingres here, since it worked for me at Uni in 1994, but it's hard to go past the best of the open source when you're already familiar with it.
They adhere to standards, not trying to keep up with custom hacks to support the whims of the OS. I remember authoring with CSS, having to be very careful not to use certain tags that did exactly what I wanted in one browser, but looked crap in another - compatibility was paramount, and forced me to stick with a rather poor subset of the rich definitions.
The gecko engine is an excellent piece of work - it's great that Netscape can now slap on a different look, add a dictionary, and release it with no need to mess around with the internals. That must have been a bane for Netscape while they had it closed. The changes from the days of Netscape 4.x are a great demonstration of open source achieving its goals and improving the code base. Security bugs are inevitable but infrequent, and the browser seems to keep on getting smaller and faster, yet more feature rich with every release!
Microsoft will never achieve that with closed model, no matter how brilliant their programmers are, because they aren't exposing their source to general criticism, and better ideas in the community. I went to Uni with one of the developers at Redmond. He has a brilliant mind, but knowing him, in the end he'll just write according to the other peoples' expectations, no matter how flawed that may be.
A far far better and more informative read IMHO is The Cathedral and The Bazaar. Beware, it's on the long side.
This gives an interesting insight into the open source model through taking over an open source project. It presents lessons learnt, and corresponding cardinal rules when running such a project. It also outlines quite effectively why open source is a viable means to develop quality software, despite the author's initial reservations. In C or C++ even.
Ok, first, I'm no Microsoft fan, but if you are a Microsoft shop, and you want commercial support, there's little choice but to use SQL Server. It's a logical choice given the circumstances.
... wow, right?
Second, I've seen a bit of Oracle, and watched our DBAs tearing hair out over mismatches in certifications (we are forced to use RH AS2.1 for their iLearning product where I work, where everything else happily runs on EL3). However, if you're not trying to make a whole bunch of suites of Oracle software work together then it all looks ok. I must agree that they definitely have done databases right. It's hard to go past RAC for ensuring reliability!
My previous job, I was (network|system|database|web) administrator, and (web|software) (developer|troubleshooter) in the early days, until we were able to expand the IT base a bit. That introduced me to PostgreSQL. I will always remember those days fondly. After a few years away I'm certainly very rusty, but there's nothing to stop me printing up the manuals and getting back on the rails if I ever get the time.
With our implementation of PostgreSQL we had almost no downtime that wasn't caused by the boss, and experienced one whole inconsistency issue. That issue was patched within 48 hours of being reported. And our hideously inefficient queries just kept chugging along when I felt sure the lack of grunt would cause SOMETHING to snap... IMHO MySQL is still eating its dust in features and reliability. Stick it on reiser4 filesystems, then I reckon virtually all risk becomes genuine hardware failure.
But hang on... it's not just open source. You can buy a license if you want commercial support! You can even run it on Windows! Not that I would, but
The only thing I can say for CA is I hope they maintain the back end better than they create their front ends. The company I work for had CA Unicenter, and since the license expired we switched to an Oracle application and haven't regretted it. It's good to keep the current vendors on their toes, but somehow I doubt they'll grab a good market share. We will see...
The last box I hacked (root password needed resetting) was running SCO Unix... :-P