I'm too late, so someone beat me in zope worshipping;).
I should say that I was _never_ in the 10 years I program know for preaching my language/plattform of choice, but if you have gone to zope you'll love it and will never look back (at least surely not if you came from php/perl, and I did them both).
Just to add two things which I think randolpho above didn't mention (just skimmed the text), these are transaction safeness (never ever write db->commit() ) and it very friendly community.
I've heard that terrorists have planted mines in several lonely lakes in alaska. You know the picture, recreational angler runs his small boat on the mine. Gets rescued, an BOOM! the really big mine goes up, NOT ONLY KILLING THAT MAN, BUT ALSO TAKING DOWN THE RESCUE HELICOPTER!!!!!!!!
Boy, I won't go near such a lonely lake if you paid me a million $$$$$!!!
And don't get me started about the land mines in mojave desert.
But if it gets rewritten in 1 week, it would be very hard for SCO to argue that this copyright/contract/trade secret/whatever breach is worth 1e+9$, because it has been proven that this code cost at _max_ 1e+6 $ to write (and I know, this number is way to high, but it's 1/1000 of SCO's claims, so let's be generous).
Go to SCO's german website (it's been like that for several days), what you see there is probably caused by the cease and desist letter they got in germany (to stop making claims against stolen code in linux).
The conclusion is that they are more interested in holding these accusations up _and_ secret than in getting customers (go to the site to understand).
If that doesn't open them to shareholder lawsuits, I don't know what will.
Oh, and btw., I hope that http_s_://www.sco.de will cost them money anyway, hehehehe.
Even stranger than that, is UnitedLinux, which people seem to be forgetting about. If SCO knew that there were IP violations in Linux while they were working on and promoting UnitedLinux, something isn't lining up right.
This is absolutely correct, and in turn opens up SCO for lawsuits from the other UnitedLinux partners.
Interesting, because this completely ignores that POST and 301/302 redirects don't play well together according to the HTTP 1.1 rfc. (It says browser SHOULD/MUST NOT follow the redirect without user intervention). That seems the real reason for me to be careful with redirecting non HEAD/GET requests.
Wow, even more braindead laws then I had imagined. Doesn't this make tv guides illegal - or more precise, doesn't this cause prohibit tv channels from publicizing their schedule in advance?
Unlike apache, microsoft and ibm the apple server exists for one purpose - to serve music.
I guess you left out Real on purpose when replying.
I wouldn't put it past them to sue apple or - at the very least - cancel their contract with apple which would cause apple's Music store to fail miserably.
That's it! It's just politics (or business)! The reason I posted my first comment was to underline that we don't (at least atm) live in a world where a company like apple wouldn't be able to market a streaming music server capable of the whole IPv4 adress range without getting sued. And it's not (in this case) corruption of the legal system by the MPAA, it's simply just manus manum lavat(in english maybe "tit for tat"?) between apple and the music industry.
Since a failure to enforce copyright provisions or enabling consumers to share pirated music will cause lawsuits to rain down upon their heads a la kazaa, napster, etc, etc.
Oh puleeze!
Nobody could sue apple for producing a _server_ which could potentially used for distributing mp3s. Mind you, this is _not_ like kazaa, where one could argue that they base their business modell on people sharing stuff which is illegal to share. If this is really the reason, expect real, microsoft, ibm, apache foundation and any fucking one else who produces server software to be targeted, too.
No, I guess this is just a political move to not destroy their onlince music buying (whatever it'S alled) thingy.
In the development of a transactional processing system, once the volume of data becomes large and the needs for data integrity and manageability becomes overwhelming... you had soooo better be using a relational database.
This sounds like you are giving quality-of-implementation reasons for favoring relational models above entity-relationship models.
I'd say that the relational model per se is better fitted for large data sets. As nice as the thought of looking at everything as objects is, it just isn't flexibel enough for reality IMO.
You and geekoid are right about IBM and MS'S relationship in the past, I wish I had been clearer.
What makes it different from the OS/2 situation IMO are two things. First, it's not a unilateral thing between MS and IBM (as it was in the OS/2 days), and that the gain for MS is not as big as it was at that time, but the risk is higher.
They gain a chance to hurt the market's adoption of linux, like you say, but how far they succeed is very uncertain. Either because "the market" doesn't care (enough), or because the case gets thrown out of court (not completly unlikely), or both. Their risk is that IBM aggressively pushes linux on the client - or more aggressively, because things like eclipse etc. _are_ defacto helping the adoption of linux on the client. And there are a lot of relatively cheap possibilities for IBM to further push the linux desktop, ranging from improving hardware support (notebooks come to mind), investing in relevant companies, helping out KDE/Gnome or Mono, to even participating with Open Office or Mozilla, etc, etc. And the fact that they would have to collaborate with rival companies like Sun wouldn't stop them, as it didn't stop them on the kernel.
MS is not stabbing IBM in the back.
They are. Everybody knows that linux-as-the-only-unix-we-sell is _the_ core strategy for IBM on server machines (at least sub-mainframe, but who knows), and now they are not only openly attacking this, but also do it by openly supporting a company which attacks IBM with a dubios lawsuit.
I think this is actually a sign how desperate MS is. Yes, I wrote desperate. They are basically stabbing IBM in the back, and that seemingly for no apparent reason, except for the fact they want to hurt IBM's adoption of linux. And that is why I am inclined to call it desperate, because it will hurt them more than it helps. SCO will lose this suit big time, and IBM will be see that another proof that MS is unreliable, which will further underline the importance for them to go with linux. Basically, MS may have declared an end to a business relation with IBM, where both partners demonstrated a good relationship in the public while kicking each others shinbone under the table.
They openly kicked IBM here, and they'll have to expect IBM to do the same when they get the chance. Therefore I think MS wouldn't have done that if they had felt themselves in a strong position against IBM/linux.
Then there's the fact that Microsoft has one keen advantage over OSS. That is the fact that it's centerally managed. In the end, one person has the final say over things, keeping things constant, consistent and flowing smoothly.
This is totally wrong. There are sooo many examples which show that Microsoft constantly changes its way, corrects itself etc. etc. Examples are their early strategy regarding the internet, windows pre 4.0 vs. windows after 4.0 (departure from pure microkernel strategy), and so on. They have no grand technical masterplan, and if I'd have to compare the consistency of the windows line vs. linux, linux wins hands down, because compared to linux, windows has had _far_ bigger paradigm changes.
Ok, but it's not a good idea, IMO. Well at least it can be optimized IMO. (I grant license to anyone to use it, as long as he doesn't patent it, then you own me 1 bazillion $$$$;) )
Here's it:
Just use the last search term the user entered before clicking on a search result, that way you are guaranted that nothing silly will polute your database.
But there's a lot more to ACID than just keeping RAM and disk in sync, and I don't see how RAM would make ACID that much easier, and certainly not "almost trivial". You still have all the transactional semaphores, record locking, potential deadlocks, rollbacks, etc. to worry about.
IMO there's one important argument which shows that this is complete bullshit, that is....
SWAP
Yes, it shows that modern OSs do _not_ allow to second guess about the properties of an interface from what you think might be the implementation. I could add NUMA, transparent memory access throught the network etc..
It feels like people using open() without checking the return value because "the file is always there, and writeable, and the disk is plenty enough".
Slashdot Mirror
I'm too late, so someone beat me in zope worshipping ;).
I should say that I was _never_ in the 10 years I program know for preaching my language/plattform of choice, but if you have gone to zope you'll love it and will never look back (at least surely not if you came from php/perl, and I did them both).
Just to add two things which I think randolpho above didn't mention (just skimmed the text), these are transaction safeness (never ever write db->commit() ) and it very friendly community.
I've heard that terrorists have planted mines in several lonely lakes in alaska. You know the picture, recreational angler runs his small boat on the mine. Gets rescued, an BOOM! the really big mine goes up, NOT ONLY KILLING THAT MAN, BUT ALSO TAKING DOWN THE RESCUE HELICOPTER!!!!!!!!
Boy, I won't go near such a lonely lake if you paid me a million $$$$$!!!
And don't get me started about the land mines in mojave desert.
I stared a while at what you calculated there, and I can't see it. Could you please explain what this 0.0025% propability is about?
The Language of Shakespeare in danger
How far will IP terrorism go.
Didn't read the article, but lemme guess:
An attorney announced a class action suit on behalf of a million monkeys, right?
But if it gets rewritten in 1 week, it would be very hard for SCO to argue that this copyright/contract/trade secret/whatever breach is worth 1e+9$, because it has been proven that this code cost at _max_ 1e+6 $ to write (and I know, this number is way to high, but it's 1/1000 of SCO's claims, so let's be generous).
Go to SCO's german website (it's been like that for several days), what you see there is probably caused by the cease and desist letter they got in germany (to stop making claims against stolen code in linux).
The conclusion is that they are more interested in holding these accusations up _and_ secret than in getting customers (go to the site to understand).
If that doesn't open them to shareholder lawsuits, I don't know what will.
Oh, and btw., I hope that http_s_://www.sco.de will cost them money anyway, hehehehe.
This possible SCO "suicide" is happening in real time over the last few days and I'm sure the shareholder suits will duly follow.
Hehe, I have an idea. Just for the kick of it, IBM should acquire SCO stock for like $500 (hurry up IBM, before this gets you the whole company).
And then, after they got rid of that silly lawsuit, go after SCO's managers in a shareholder suit. That'd teach this scum.
Even stranger than that, is UnitedLinux, which people seem to be forgetting about. If SCO knew that there were IP violations in Linux while they were working on and promoting UnitedLinux, something isn't lining up right.
This is absolutely correct, and in turn opens up SCO for lawsuits from the other UnitedLinux partners.
Yeah, I should have written "btw. they are fuckin funny". Oh, and that I mistyped the name doesn't add to the sense of my post.
Maybe I'd add
rainforest puppy and
goobles (they are fruckin funny).
"We're still identifying more and more code from Unix System V that is in Linux, and so we haven't even fully scoped the problem"
[5 months later]
SCO issues a press release, declaring that after intense code reviews they are absolutely sure that in fact, linux _is_ Open Server.
Interesting, because this completely ignores that POST and 301/302 redirects don't play well together according to the HTTP 1.1 rfc. (It says browser SHOULD/MUST NOT follow the redirect without user intervention).
That seems the real reason for me to be careful with redirecting non HEAD/GET requests.
Wow,
even more braindead laws then I had imagined. Doesn't this make tv guides illegal - or more precise, doesn't this cause prohibit tv channels from publicizing their schedule in advance?
Unlike apache, microsoft and ibm the apple server exists for one purpose - to serve music.
I guess you left out Real on purpose when replying.
I wouldn't put it past them to sue apple or - at the very least - cancel their contract with apple which would cause apple's Music store to fail miserably.
That's it! It's just politics (or business)! The reason I posted my first comment was to underline that we don't (at least atm) live in a world where a company like apple wouldn't be able to market a streaming music server capable of the whole IPv4 adress range without getting sued.
And it's not (in this case) corruption of the legal system by the MPAA, it's simply just manus manum lavat(in english maybe "tit for tat"?) between apple and the music industry.
Since a failure to enforce copyright provisions or enabling consumers to share pirated music will cause lawsuits to rain down upon their heads a la kazaa, napster, etc, etc.
Oh puleeze!
Nobody could sue apple for producing a _server_ which could potentially used for distributing mp3s.
Mind you, this is _not_ like kazaa, where one could argue that they base their business modell on people sharing stuff which is illegal to share.
If this is really the reason, expect real, microsoft, ibm, apache foundation and any fucking one else who produces server software to be targeted, too.
No, I guess this is just a political move to not destroy their onlince music buying (whatever it'S alled) thingy.
. But i wont be using it since Firebird runs fine
It can't hurt to have a web browser besides you database.
In the development of a transactional processing system, once the volume of data becomes large and the needs for data integrity and manageability becomes overwhelming... you had soooo better be using a relational database.
This sounds like you are giving quality-of-implementation reasons for favoring relational models above entity-relationship models.
I'd say that the relational model per se is better fitted for large data sets.
As nice as the thought of looking at everything as objects is, it just isn't flexibel enough for reality IMO.
You and geekoid are right about IBM and MS'S relationship in the past, I wish I had been clearer.
What makes it different from the OS/2 situation IMO are two things. First, it's not a unilateral thing between MS and IBM (as it was in the OS/2 days), and that the gain for MS is not as big as it was at that time, but the risk is higher.
They gain a chance to hurt the market's adoption of linux, like you say, but how far they succeed is very uncertain. Either because "the market" doesn't care (enough), or because the case gets thrown out of court (not completly unlikely), or both.
Their risk is that IBM aggressively pushes linux on the client - or more aggressively, because things like eclipse etc. _are_ defacto helping the adoption of linux on the client.
And there are a lot of relatively cheap possibilities for IBM to further push the linux desktop, ranging from improving hardware support (notebooks come to mind), investing in relevant companies, helping out KDE/Gnome or Mono, to even participating with Open Office or Mozilla, etc, etc. And the fact that they would have to collaborate with rival companies like Sun wouldn't stop them, as it didn't stop them on the kernel.
MS is not stabbing IBM in the back.
They are. Everybody knows that linux-as-the-only-unix-we-sell is _the_ core strategy for IBM on server machines (at least sub-mainframe, but who knows), and now they are not only openly attacking this, but also do it by openly supporting a company which attacks IBM with a dubios lawsuit.
I think this is actually a sign how desperate MS is. Yes, I wrote desperate.
They are basically stabbing IBM in the back, and that seemingly for no apparent reason, except for the fact they want to hurt IBM's adoption of linux.
And that is why I am inclined to call it desperate, because it will hurt them more than it helps. SCO will lose this suit big time, and IBM will be see that another proof that MS is unreliable, which will further underline the importance for them to go with linux.
Basically, MS may have declared an end to a business relation with IBM, where both partners demonstrated a good relationship in the public while kicking each others shinbone under the table.
They openly kicked IBM here, and they'll have to expect IBM to do the same when they get the chance. Therefore I think MS wouldn't have done that if they had felt themselves in a strong position against IBM/linux.
Seems fitting, even more so with a small addition:
"Never attribute to malice that which can be adequately explained by [the combination of malice and] stupidity"
Then there's the fact that Microsoft has one keen advantage over OSS. That is the fact that it's centerally managed. In the end, one person has the final say over things, keeping things constant, consistent and flowing smoothly.
This is totally wrong. There are sooo many examples which show that Microsoft constantly changes its way, corrects itself etc. etc.
Examples are their early strategy regarding the internet, windows pre 4.0 vs. windows after 4.0 (departure from pure microkernel strategy), and so on.
They have no grand technical masterplan, and if I'd have to compare the consistency of the windows line vs. linux, linux wins hands down, because compared to linux, windows has had _far_ bigger paradigm changes.
Ok, but it's not a good idea, IMO. Well at least it can be optimized IMO. ;) )
;)
(I grant license to anyone to use it, as long as he doesn't patent it, then you own me 1 bazillion $$$$
Here's it:
Just use the last search term the user entered before clicking on a search result, that way you are guaranted that nothing silly will polute your database.
HTH
How about killing just the worker process which hangs?
But there's a lot more to ACID than just keeping RAM and disk in sync, and I don't see how RAM would make ACID that much easier, and certainly not "almost trivial". You still have all the transactional semaphores, record locking, potential deadlocks, rollbacks, etc. to worry about.
....
IMO there's one important argument which shows that this is complete bullshit, that is
SWAP
Yes, it shows that modern OSs do _not_ allow to second guess about the properties of an interface from what you think might be the implementation. I could add NUMA, transparent memory access throught the network etc..
It feels like people using open() without checking the return value because "the file is always there, and writeable, and the disk is plenty enough".