Whether or not a person can correctly edit a document is hardly a realistic measure of the quality of a software package.
For instance, how many people can solve partial differential equations? The vast majority of people will look at you with a dopey look in response to such a question. Writing software to solve such mathematical wonders is definitely possible though, with 100% correctness (software bugs withstanding - the point is that the intended behaviour is simply modelled incorrectly by the developer, not the intended behaviour representing something that cannot be modelled perfectly).
Whether or not someone needs to look up grammatical terms for clarification is irrelevant when dealing with software, as computers tend to be quite good at 'looking things up'.
Models able to tout 100% accuracy for classification simply do not exist. Computational linguists have been working on this for a long time, and they've not produced a solution yet.
For models that do approach 90% accuracy and better, the sheer complexity (both space and time) is too high for consumer applications. Requiring hours of computing time to initialise a model from training data is normal. To save the internal state of the model so that the training stage can be avoided requires massive amounts of storage and thus a gnarly loading time to contend with.
By building any application in.NET, you're committing professional negligence. You're locking your client into an entire product line. To quote Gosling,.NET is a product - not a platform.
Should your client ever decide to move to another platform or seek an alternative developer to take care of the code base you're producing for them, choosing.NET will immediately limit their choices.
1. UnixWare will crash 2. UnixWare will tell you that no such call exists
Either way, you could completely and totally use it as an excuse to make whichever person you installed UnixWare look like a complete dick, and then shoot them in the groin with a nailgun.
From the first post's linked
Computer World article, we learn that dear old Darl has decided to call their brand new, ground breaking product OpenServer Legend. Huh? What the fuck are they smoking over there in Utah? And they reckon IBM is giving out the happy Kool Aid...
This makes no sense at all. They've got two products, one even more useless than the other and they put their weight behind the useless one! OpenServer is the worst possible OS one could install on an x86 machine these days! I'd sooner install Windows (or Microsoft's flagship OS, DOS 2.11) than OpenServer. The person quoted in the linked story saying that they've been testing their apps on NT machines due to worries about having to move away from OpenServer should be shot repeatedly in the groin with a nailgun for both moving from a former UNIX to NT and for saying that SCO have good things coming.
Has this "person" not heard of Solaris? Or Linux? Or FreeBSD? Or OpenBSD? Or NetBSD? Or DOS 2.11??? All of the mentioned OSes have one thing in common - they are infinitely better UNIXes than that crap pile called OpenServer. Having to support it still causes me nightmares.
So, what's the next UnixWare release going to be called? How about "UnixWare puncture repair kit shitfaced cockmaster"?
SCO says "heaps of our code from the old SysV codebase, the USL codebase and the UnixWare codebase was copied into Linux!"
Are there any kernel gurus out there that would care to tell me about what level of work is involved in bringing code from another OS (say, one of the BSDs) into the Linux kernel? Is it even reasonable to expect that code from a UNIX kernel rooted in the past could have it's code shoehorned into the Linux kernel line-for-line as SCO are alledging?
Back to the OpenServer->Linux migrations! Damned SCO, I am not paying for UnixWare so that I can tell it to emulate OpenServer.
And to think, Chris Sontag said told the world "don't be surprised" if the US government appears with SCO in going after IBM due to IBM going around US export laws with the enterprise features they've contributed to the funky penguin OS. There goes that partnership!
Regardless of what each person thinks, there are a few worrying aspects here that few people seem to have looked at. I am a big Linux fan. I have migrated several SCO installations to Linux and I, along with my clients, have never been hapier. SCO have decided, against all common sense, to take on IBM. IBM is one of the "big three" when it comes to enterprise UNIX (along with HP and Sun). I like IRIX as well, not to mention the unforunately dying Tru64. Then there's VMS and QNX.....
This is where it gets interesting, and I implore the./ crowd to point out any referencing material to help me out here, but why on earth would IBM team up with SCO for Project Monterey? IBM have experience in developing UNIX (and other enterprise class Operating Systems) operating systems for 64-bit architectures, as well as the supporting tools to go along with such a beast.
What exactly did SCO bring to the table? Considering that the answer to the last question above is quite possibly "Nothing", one must ask why it was that SCO claims that enterprise-strength features are the bulk of the misappropriated property here? Most of the products SCO are claiming ownership of don't exist in any SCO product line except OpenLinux (if you can call that a product). SCO don't have any experience in developing for anything other than x86, and no one would call that an advantage when developing code for the Itanium. It seems that the only thing SCO have that gives them reason to go on in this game is their legal team - people who know about how IBM works.
Looking at the Novell situation, it is terribly convenient that SCO happened to "find" a document detailing their ownership of various copyrights after Novell had made a public statement. As Novell were quite methodical and timed in their release of their press statement, one could very well assume two things:
SCO knew that Novell would have something to say
SCO knew that Novell would not have the "first ammendment" on hand
The factualness of the second point above can be further asserted by the fact that Novell basically went quiet soon after SCO responding. To paraphrase saint McBride, SCO's legal "team" have been going over the Linux codebase for over 12 months now. Yet, in that time:
Caldera/SCO got into United Linux in a big way
Caldera/SCO continued to push their own Linux
Caldera/SCO continued to make open source tools and their inclusion major features of new releases of UnixWare and OpenServer
Regardless, SCO has told the nice blokes at SuSE that they'll be on SCO's list of potential victims at some point, they continue to refine LKP for UnixWare and exclaim on the excellence of OpenLinux as the ideal conduit for migration from commercial UNIXes (they also say that UnixWare is the best commercial UNIX though...) Looking at the above, you could be led to beleive that SCO has planned all of these events. This is not a conspiracy of any sort - simply a possible example of excellent planning, with very questionable overtones in each of the three major cases above.
In reality, could a closed-source kernel (such as UnixWare or OpenServer) contain code that was suitable for copying straight into Linux? Everyone seems to forget here that Linux departs majorly from the way that most UNIXes do things. Linux doesn't use working sets for memory management as one prime example. It has two memory allocation mechanisms as another case in point. With the resources that SCO has, one would seriously doubt their ability to make changes to their codebase that depart massively from the "traditional" way of doing things while still maintaining a supportable product with the release cycle they're committed to. Seriously though, OpenServer doesn't even seem to have anything looking like kernel-space threads (no offence to any open source OSes that like their user-space thread implementations. In your case, your threads are that way because they serve you
The way I see it, there is an awful lot that Novell stands to gain out of this:
A native Groupwise client - Novell has publicly stated that the anticipated move away from client software to web interface access to Groupwise was wrong. Looking at the Ximian
press release confirms this part of the deal. The fact that they now have Java clients available to fill this gap seems a bit hollow now really.
Another bit of server software to flaunt - Novell is being a lot more proactive in expanding the Novell NetWare software library as of late. They assisted in the ports of PostgreSQL and MySQL to NetWare after Oracle dropped support. Mono fits rather nicely considering the above.
They can let PHB and BOFH types have their way - Some PHBs will choose Windows as their server platform based on the "strugle" (read slaughter) involving.NOT and J2EE. If Novell throw some resources behind Mono, they can also fill that role for those who must have a commercial OS (yes, all 6 of you).
ZenWorks gets a boost - Red Carpet works and works well now. As ZenWorks is one of the cornerstones of Novell's Linux strategy, anything they can get their hands on to improve it is a definite "yes, gimme!".
"From the trenches" assistance with their Linux push - Even though Novell have plenty of experience developing for UNIX in general and Linux specifically, having some of the bods from Ximian to help out with the porting of things like iPrint to Linux would have to be a plus
Offering the complete package - Even though they tend to work together a fair bit these days (both are committed to J2EE, both are involved in the Liberty alliance), Novell needs a desktop strategy to compete with Sun's desktop Linux plans. Additionally, their sales force will no longer need to say "Sure! We can provide all of your server needs! Err...client side? Umm....the area code for Redmond is...."
Highlighting Novell's commitment to the open source world - Even though Novell have backed down slightly from those dicks who sell stuff that they dare call a UNIX (all who want to see an OpenServer box urinated on, set on fire and then detonated get in touch) this is a pretty decent way of saying "Yes, you own the copyright to a few things. Think we care?"
I may not be from the US, but surely now that the integrity of this election has been so badly comprimised and brought into question, a new ballot makes sense. Although our voting laws are quite different here in Australia, the fact that our laws and procedures ensure everyone's vote always count should be a good example for everyone.
The people of the USA seem to be having very little say in what is happening in the presedential race at the moment, so surely giving them their right to vote again would be the best solution.
Slashdot Mirror
Whether or not a person can correctly edit a document is hardly a realistic measure of the quality of a software package.
For instance, how many people can solve partial differential equations? The vast majority of people will look at you with a dopey look in response to such a question. Writing software to solve such mathematical wonders is definitely possible though, with 100% correctness (software bugs withstanding - the point is that the intended behaviour is simply modelled incorrectly by the developer, not the intended behaviour representing something that cannot be modelled perfectly).
Whether or not someone needs to look up grammatical terms for clarification is irrelevant when dealing with software, as computers tend to be quite good at 'looking things up'.
Models able to tout 100% accuracy for classification simply do not exist. Computational linguists have been working on this for a long time, and they've not produced a solution yet.
For models that do approach 90% accuracy and better, the sheer complexity (both space and time) is too high for consumer applications. Requiring hours of computing time to initialise a model from training data is normal. To save the internal state of the model so that the training stage can be avoided requires massive amounts of storage and thus a gnarly loading time to contend with.
There's an easy way of looking at this.
.NET, you're committing professional negligence. You're locking your client into an entire product line. To quote Gosling, .NET is a product - not a platform.
.NET will immediately limit their choices.
By building any application in
Should your client ever decide to move to another platform or seek an alternative developer to take care of the code base you're producing for them, choosing
If you do that, one of two things will happen:
1. UnixWare will crash
2. UnixWare will tell you that no such call exists
Either way, you could completely and totally use it as an excuse to make whichever person you installed UnixWare look like a complete dick, and then shoot them in the groin with a nailgun.
From the first post's linked Computer World article, we learn that dear old Darl has decided to call their brand new, ground breaking product OpenServer Legend. Huh? What the fuck are they smoking over there in Utah? And they reckon IBM is giving out the happy Kool Aid... This makes no sense at all. They've got two products, one even more useless than the other and they put their weight behind the useless one! OpenServer is the worst possible OS one could install on an x86 machine these days! I'd sooner install Windows (or Microsoft's flagship OS, DOS 2.11) than OpenServer. The person quoted in the linked story saying that they've been testing their apps on NT machines due to worries about having to move away from OpenServer should be shot repeatedly in the groin with a nailgun for both moving from a former UNIX to NT and for saying that SCO have good things coming. Has this "person" not heard of Solaris? Or Linux? Or FreeBSD? Or OpenBSD? Or NetBSD? Or DOS 2.11??? All of the mentioned OSes have one thing in common - they are infinitely better UNIXes than that crap pile called OpenServer. Having to support it still causes me nightmares. So, what's the next UnixWare release going to be called? How about "UnixWare puncture repair kit shitfaced cockmaster"?
SCO says "heaps of our code from the old SysV codebase, the USL codebase and the UnixWare codebase was copied into Linux!"
Are there any kernel gurus out there that would care to tell me about what level of work is involved in bringing code from another OS (say, one of the BSDs) into the Linux kernel? Is it even reasonable to expect that code from a UNIX kernel rooted in the past could have it's code shoehorned into the Linux kernel line-for-line as SCO are alledging?
Back to the OpenServer->Linux migrations! Damned SCO, I am not paying for UnixWare so that I can tell it to emulate OpenServer.
And to think, Chris Sontag said told the world "don't be surprised" if the US government appears with SCO in going after IBM due to IBM going around US export laws with the enterprise features they've contributed to the funky penguin OS. There goes that partnership!
This is where it gets interesting, and I implore the ./ crowd to point out any referencing material to help me out here, but why on earth would IBM team up with SCO for Project Monterey? IBM have experience in developing UNIX (and other enterprise class Operating Systems) operating systems for 64-bit architectures, as well as the supporting tools to go along with such a beast.
What exactly did SCO bring to the table? Considering that the answer to the last question above is quite possibly "Nothing", one must ask why it was that SCO claims that enterprise-strength features are the bulk of the misappropriated property here? Most of the products SCO are claiming ownership of don't exist in any SCO product line except OpenLinux (if you can call that a product). SCO don't have any experience in developing for anything other than x86, and no one would call that an advantage when developing code for the Itanium. It seems that the only thing SCO have that gives them reason to go on in this game is their legal team - people who know about how IBM works.
Looking at the Novell situation, it is terribly convenient that SCO happened to "find" a document detailing their ownership of various copyrights after Novell had made a public statement. As Novell were quite methodical and timed in their release of their press statement, one could very well assume two things:
The factualness of the second point above can be further asserted by the fact that Novell basically went quiet soon after SCO responding. To paraphrase saint McBride, SCO's legal "team" have been going over the Linux codebase for over 12 months now. Yet, in that time:
Regardless, SCO has told the nice blokes at SuSE that they'll be on SCO's list of potential victims at some point, they continue to refine LKP for UnixWare and exclaim on the excellence of OpenLinux as the ideal conduit for migration from commercial UNIXes (they also say that UnixWare is the best commercial UNIX though...) Looking at the above, you could be led to beleive that SCO has planned all of these events. This is not a conspiracy of any sort - simply a possible example of excellent planning, with very questionable overtones in each of the three major cases above.
In reality, could a closed-source kernel (such as UnixWare or OpenServer) contain code that was suitable for copying straight into Linux? Everyone seems to forget here that Linux departs majorly from the way that most UNIXes do things. Linux doesn't use working sets for memory management as one prime example. It has two memory allocation mechanisms as another case in point. With the resources that SCO has, one would seriously doubt their ability to make changes to their codebase that depart massively from the "traditional" way of doing things while still maintaining a supportable product with the release cycle they're committed to. Seriously though, OpenServer doesn't even seem to have anything looking like kernel-space threads (no offence to any open source OSes that like their user-space thread implementations. In your case, your threads are that way because they serve you
Well, that's me out of ideas.
I may not be from the US, but surely now that the integrity of this election has been so badly comprimised and brought into question, a new ballot makes sense. Although our voting laws are quite different here in Australia, the fact that our laws and procedures ensure everyone's vote always count should be a good example for everyone.
The people of the USA seem to be having very little say in what is happening in the presedential race at the moment, so surely giving them their right to vote again would be the best solution.