I find it invigorating, heartening, and just down right beautiful, to read such a vehement impeachment of an article it's author, and the sight it was posted to that has...
Yes, spoken language is murkey as hell.
When you phrase something, and release it,
it's meaning no longer belongs to you,
but exists in the reciever. Very rarely
does the reciever concieve perfectly your
intention.
However, Mathematics while still dependent
on the knowledge of the reciever, is
absolute if the reciever id knowledgeable
enough to understand the math.
maybe we should all speak in geometric
circles. Shrug.
I can think of lots of reasons not to contribute to hurd.
1) it IS a fundamentally different thing from this.
2) It's design is insane... Not in that it's not a good idea, just that it's very diffacult to grok the interactions correctly.
3)I don't like RMS, he's funny looking.
Actually... the one place where napster is likely to be damaging to RIAA sales, is in the singles department. Most people don't download albums of music, the download _that_ song that they want _right now_. That is the crowd that would normally be buying singles.
Personally, i've never baught a single in my life.
why buy a 8-10 dollar single when you can just wait for the 15 dollar album.
which would be a good point... except this entire battle is over potentials, not realities, and megalomania. Right now, the RIAA is a functional monoply of music content. They, in a very real sence, control all music and musical media... except what napsters giving away for free.
Napster, as a technology, has the potential, to freely trade everything the RIAA makes its stupifyingly large income on. Not right now, but in the next 5 to 10 years. Thats VERY VERY BAD to them, as they want that control over all distribution, so they can continually inflate prices and manipulate contracts jacking the artist on one end, and the consumer on the other.
how about the apple roms that were there for ever and ever that needed to be licensed to make clones... And have been an annoying pain in the ass for things like mklinux and linux ppc in the past.
I've been a consultant since i was 17, and a full time developer since i was 19. I'm now been a consultant, system admin, and developer for over 7 years, and i still run into what is essentially, age descrimination especially when i come into contact with older companies.
I've worked in fortune 500 and 50 companies for the last 5 years, and management consistently ignores me, though i've seen the same mistakes over and over and over....
It got too a point at a previous company that i was working at, where a co-worker recognized my ability, and championed my ideas... they were mutch better recieved from him...
At the current rate of legislation, copyrights won't expire. They've (relatively) recently moved from 20 years, to 75 years or 95 + life of creator... all of these were effective immediatly, essentially granting large windows... things copywritten up to 15 years before i was born will probably still be more than 15 years after I die.
by "two dimentional" i mean logical. With an rdbms you can't make an intellegent nested structure without a myriad of joins and lookups. Flattening a tree into X tables (or a single self referencing table) drives me batty.
The native format issue is mostly an elegance issue. Good engineering makes complex structures easy to deal with. So a well engineered view centric data orgainzation system should make dealing with data as either a tree, or a table equally elegant. So where do you start...
Well... here's an attempt.
It would be written for Microsoft QuickBasic 4.5 for dos... I don't have a copy any more, or a machine i could run it on... but i _think_ this'd work...
Giggle... here's a new hello world quest... write hello world in every language using an object oriented paradigm.
TYPE app
DIM name AS STRING
END TYPE
DECLARE SUB app_run( BYREF theApp AS app )
DECLARE SUB app_init( BYREF theApp as app )
DECLARE SUB app_helloWorld( BYREF theApp as app )
DIM myApp as app
app_init( myApp )
app_run( myApp )
END
SUB app_init( BYREF theApp AS app )
theApp.name = "Hello World"
END SUB
SUB app_run( BYREF theApp AS app )
PRINT theApp.name
END SUB
Most RDBMS stuff i've run into is trying to flex trees into two dimentional structures...
All one to many associations can be mapped this way.
The difficulty is that pure trees become exercises in walking algorithms.
Directory services should have mixed views because very often it's far more flexible (cascading permissions, organizational grouping etc.) to deal with a tree for those items. However as often as not you want to say "give me all employees with level 2 security clearence", and pulling that from a table would be far more efficient than walking the organizational trees for atlanta, langley, washington, and annapolis.
The real question, is which format does the data exist in natively....
I'm definitely not arguing against the use and creation of libraries. The problem is... most of the time, that _is_ developing one thing, and then quitting.
When it's not... the cycle usually runs something like. I develope on program. I develop another that does something similiar, i get the project assignment for a third that does something similiar, then i abstract a library, write the third, and back port the two during the next review.
As a blue collar business programmer though, i usually find myself moving on before i get to that third program. Additionally, justifying that time i spent on the library to management is often painfull. Then taking into account my deadlines, and often i'm just stuck writing three programs that have alot of very similiar code.
If some of the abstracted libraries are interesting enough, i might work overtime, or in my own time on them if they're sufficiently interesting, or if i know unequivcably ahead of time that the next two projects are coming down the pipe, and that i'm getting them.
These circumstances don't happen too often however, and more often than not they aren't of much use outside the relatively small context of those three apps. They can't be GPL'd, or BSD'd or, personally, Artistically liscensed and released because they belong to the company and most managers look at you like you're a moron, either out of ignorance or intolerance, and dismiss it out of hand.
Not that they'd be particually useful if released anyway, because they're usually aimed at biz niches.
I'm a mixed paradigm programmer myself, and intend to stay that way. I've walked in to far to many maintenence projects cold to believe any hype about any methodology, as they are all equally capable of producing crap. The proof is in the code, not the theory.
I love the "the object is self contained" argument. Mostly because it's rather moot most of the time.
The only place it makes any difference is with dynamic objects in a component model, where component structures can be shared across processes/namespaces/languages....
And when you do that, you are suddenly doing something that very closely resembles... procedural programming... where the object itself is a convienent way to bundle foriegn data and code, and import it.
Don't get me wrong... It's the best thing in the world... as a windows programmer: because the only decent implementation for any purpose is COM/DCOM.
From a business programming perspective, he's correct. Developing generic libraries is a much slower process than developing application specific code to do the deed. Any coder should be able to see that in the fruit of his own work.
As far as the cobol example... I can't say. But i _CAN_ give you an assembly example.
The AIG Life Companies use a program called LifeComm on their IBM Mainframe. It handles all contract information storage and retrievel in the form of vsam files delineated with trailers.
It is written entirely in OS/390 Assembly, and a very strict coding and design principles and oversight has kept it extremely modular, and the code very re-reusable. It's probably one of the best kept source repositories i've ever seen.
The program itself is quite flexible and extremely reliable, as well as efficient.
All in Assembly.
It's not the language or the methodology, it's the discipline.
It takes a very dedicated and disciplined programmer to turn out quality reusable code. The methodology and language he uses is irrelevent.
Sure.
But those little subroutines tend to be inlined alot more often as they are part of the namespace and are usually jumped to with the proper jsr/ret combinations.
What i'm curious about is the performance of a lot of the applications that we wrote under contract in c. They sorta faked oop with structures holding function pointers used as components.
This gave us much better modelling of dynamic objects because their behaviors could be changed at run time.
First, it would be difficult to make those cache's (which in most cases are already there) persistent, which defeats the whole point of using it for fault tolerance.
This is why you place it in an arbitrary open persistent memory cache.
Slashdot Mirror
Wrong view of Fact.
It's a fact that george bush is a liar and a cheat.
That is a different sentece than
George Bush is a liar and a cheat.
The usage of the term fact, makes the statement seem more true.
I find it invigorating, heartening, and just down right beautiful, to read such a vehement impeachment of an article it's author, and the sight it was posted to that has...
Absolutely no reasons posted with it!!!
So it's a lot of bull shit hrm.... well, why?
-V
never met a libertarian who wasn't bollocks to the eyeballs.
Says abit more about your bollocks make up than it does about libertarians wouldn't you think?
Depends on your conception of language.
Yes, spoken language is murkey as hell.
When you phrase something, and release it,
it's meaning no longer belongs to you,
but exists in the reciever. Very rarely
does the reciever concieve perfectly your
intention.
However, Mathematics while still dependent
on the knowledge of the reciever, is
absolute if the reciever id knowledgeable
enough to understand the math.
maybe we should all speak in geometric
circles. Shrug.
-Tilde
Move definitions to an aribtary place in your head, and work backwards.
Maybe read some david hume.
At anyrate, the definition of a fact, is a social contstruct. It is a fact that the sky is blue.
there for the sky is blue.
It is a fact that the sky is green. is the sky green?
applying the label of "fact" to something is merely stating that most people's perceptions agree.
Blah
-Tilde
Which brings up yet another good point... Mommy SHOULD be there sitting next to little johnny while he clicks mindlessly about.
Filtering software is a piss poor prop to proper parenting.
I can think of lots of reasons not to contribute to hurd.
1) it IS a fundamentally different thing from this.
2) It's design is insane... Not in that it's not a good idea, just that it's very diffacult to grok the interactions correctly.
3)I don't like RMS, he's funny looking.
I'd never touch free drugs.
Actually... the one place where napster is likely to be damaging to RIAA sales, is in the singles department. Most people don't download albums of music, the download _that_ song that they want _right now_. That is the crowd that would normally be buying singles.
Personally, i've never baught a single in my life.
why buy a 8-10 dollar single when you can just wait for the 15 dollar album.
boggle
-T
which would be a good point... except this entire battle is over potentials, not realities, and megalomania. Right now, the RIAA is a functional monoply of music content. They, in a very real sence, control all music and musical media... except what napsters giving away for free.
Napster, as a technology, has the potential, to freely trade everything the RIAA makes its stupifyingly large income on. Not right now, but in the next 5 to 10 years. Thats VERY VERY BAD to them, as they want that control over all distribution, so they can continually inflate prices and manipulate contracts jacking the artist on one end, and the consumer on the other.
-T
how about the apple roms that were there for ever and ever that needed to be licensed to make clones... And have been an annoying pain in the ass for things like mklinux and linux ppc in the past.
I've been a consultant since i was 17, and a full time developer since i was 19. I'm now been a consultant, system admin, and developer for over 7 years, and i still run into what is essentially, age descrimination especially when i come into contact with older companies.
I've worked in fortune 500 and 50 companies for the last 5 years, and management consistently ignores me, though i've seen the same mistakes over and over and over....
It got too a point at a previous company that i was working at, where a co-worker recognized my ability, and championed my ideas... they were mutch better recieved from him...
It burns my buttons hard core.
oh well.
-T
At the current rate of legislation, copyrights won't expire. They've (relatively) recently moved from 20 years, to 75 years or 95 + life of creator... all of these were effective immediatly, essentially granting large windows... things copywritten up to 15 years before i was born will probably still be more than 15 years after I die.
I love the way you assume republicans will zonk LPR.
Traditionally, individual and state rights have been central to republican policy...
Obviously you aren't trying hard enough not to be too dogmatic.
by "two dimentional" i mean logical. With an rdbms you can't make an intellegent nested structure without a myriad of joins and lookups. Flattening a tree into X tables (or a single self referencing table) drives me batty.
The native format issue is mostly an elegance issue. Good engineering makes complex structures easy to deal with. So a well engineered view centric data orgainzation system should make dealing with data as either a tree, or a table equally elegant. So where do you start...
-T
Well... here's an attempt.
It would be written for Microsoft QuickBasic 4.5 for dos... I don't have a copy any more, or a machine i could run it on... but i _think_ this'd work...
Giggle... here's a new hello world quest... write hello world in every language using an object oriented paradigm.
TYPE app
DIM name AS STRING
END TYPE
DECLARE SUB app_run( BYREF theApp AS app )
DECLARE SUB app_init( BYREF theApp as app )
DECLARE SUB app_helloWorld( BYREF theApp as app )
DIM myApp as app
app_init( myApp )
app_run( myApp )
END
SUB app_init( BYREF theApp AS app )
theApp.name = "Hello World"
END SUB
SUB app_run( BYREF theApp AS app )
PRINT theApp.name
END SUB
Ugh.
Trees are very flexable...
Most RDBMS stuff i've run into is trying to flex trees into two dimentional structures...
All one to many associations can be mapped this way.
The difficulty is that pure trees become exercises in walking algorithms.
Directory services should have mixed views because very often it's far more flexible (cascading permissions, organizational grouping etc.) to deal with a tree for those items. However as often as not you want to say "give me all employees with level 2 security clearence", and pulling that from a table would be far more efficient than walking the organizational trees for atlanta, langley, washington, and annapolis.
The real question, is which format does the data exist in natively....
I'm definitely not arguing against the use and creation of libraries. The problem is... most of the time, that _is_ developing one thing, and then quitting.
When it's not... the cycle usually runs something like. I develope on program. I develop another that does something similiar, i get the project assignment for a third that does something similiar, then i abstract a library, write the third, and back port the two during the next review.
As a blue collar business programmer though, i usually find myself moving on before i get to that third program. Additionally, justifying that time i spent on the library to management is often painfull. Then taking into account my deadlines, and often i'm just stuck writing three programs that have alot of very similiar code.
If some of the abstracted libraries are interesting enough, i might work overtime, or in my own time on them if they're sufficiently interesting, or if i know unequivcably ahead of time that the next two projects are coming down the pipe, and that i'm getting them.
These circumstances don't happen too often however, and more often than not they aren't of much use outside the relatively small context of those three apps. They can't be GPL'd, or BSD'd or, personally, Artistically liscensed and released because they belong to the company and most managers look at you like you're a moron, either out of ignorance or intolerance, and dismiss it out of hand.
Not that they'd be particually useful if released anyway, because they're usually aimed at biz niches.
I'm a mixed paradigm programmer myself, and intend to stay that way. I've walked in to far to many maintenence projects cold to believe any hype about any methodology, as they are all equally capable of producing crap. The proof is in the code, not the theory.
I love the "the object is self contained" argument. Mostly because it's rather moot most of the time.
The only place it makes any difference is with dynamic objects in a component model, where component structures can be shared across processes/namespaces/languages....
And when you do that, you are suddenly doing something that very closely resembles... procedural programming... where the object itself is a convienent way to bundle foriegn data and code, and import it.
Don't get me wrong... It's the best thing in the world... as a windows programmer: because the only decent implementation for any purpose is COM/DCOM.
From a business programming perspective, he's correct. Developing generic libraries is a much slower process than developing application specific code to do the deed. Any coder should be able to see that in the fruit of his own work.
As far as the cobol example... I can't say. But i _CAN_ give you an assembly example.
The AIG Life Companies use a program called LifeComm on their IBM Mainframe. It handles all contract information storage and retrievel in the form of vsam files delineated with trailers.
It is written entirely in OS/390 Assembly, and a very strict coding and design principles and oversight has kept it extremely modular, and the code very re-reusable. It's probably one of the best kept source repositories i've ever seen.
The program itself is quite flexible and extremely reliable, as well as efficient.
All in Assembly.
It's not the language or the methodology, it's the discipline.
It takes a very dedicated and disciplined programmer to turn out quality reusable code. The methodology and language he uses is irrelevent.
Sure. But those little subroutines tend to be inlined alot more often as they are part of the namespace and are usually jumped to with the proper jsr/ret combinations. What i'm curious about is the performance of a lot of the applications that we wrote under contract in c. They sorta faked oop with structures holding function pointers used as components. This gave us much better modelling of dynamic objects because their behaviors could be changed at run time.
Yes, as a matter of fact, it is.
First, it would be difficult to make those cache's (which in most cases are already there) persistent, which defeats the whole point of using it for fault tolerance.
This is why you place it in an arbitrary open persistent memory cache.
I can't believe how perfect this is...
Granted i'd prefer to have two top dogs competing for my market... but to own an "Nvidia Voodoo" so preferctly ironic...
-T
Anyone else remember pulling shitty porn off the frogpond bbs in delaware?
giggle
-T
no.. there was a proliferation of online services well before "the internet" as a public phenomanon
old school bbs'prodigy, compuserve, and yes american online...
The single largest application of these services was messaging... or rather, some form of email.
-T