I don't know facebook's intentions, but I would bet dollars to donuts that the "target audience" option was there *specifically* to allow advertisers to target women, not exclude them. Can I imagine an all-female employer, a female-centric organization, not wanting to spend their dollars on marketing towards men? Maybe they'll make the gender-targeting option only available to self-described women. Yeah, that's fair.
Well, the author (of the article, Peter Davidson) doesn't know mainframes. "...IBM OS390 mainframe running ZOS..." OS390 is an operating system (the precursor to zOS). Maybe they're running on a z900 or z990 or z9 (or 2064-yada yada). So, the way I interpret this, is they are running some CICS (presumably TS 1.3, 2.3, or 3.1) which talks to DB2, does some messaging with WebSphere MQ, and all the web interface (WebSphere, whatever) is on some Unixy (Solaris?) front end. Speaking as a IBM mainframer, any zOS or CICS systems programmer worth their salt would be able to tell you how long all of those transactions took, and where the problem was. I guarantee we run a much smaller box here, and we push through millions of transactions a day (just business hours!), and we guarantee the majority of the transactions are done in 1/4 second or less. So, shitty code, shitty performance, but on the backend they should be able to pinpoint any problems. I would guess that the bottleneck (and money pit) showed up on the front end (or everyone on the project is clueless - 50/50).
This is business computing. You take some data from point A, and move it to point B, with perhaps a little transformation in between. Not that tough, right? They have their way of doing it. You have your way. What's wrong with their way? What great benefit will be drawn from doing it your way? Because you can't do COBOL or Assembler? Why not?
Quite right. In our shop, we guarantee that 75% of our CICS transactions will run in 1/5 of a second. 90% in less than 1/2 second. It flies. And it's not just the hardware. It's the programming. zOS is only as good as the people who run it.
Being a zOS systems programmer in my mid-30s, having done this for about 7 years, I think the blame for any shortage lies with management, with some blame going to universities for dropping the ball. There is plenty of education available out there, either from IBM or other sources, but your boss (or school) has to get you there.
The article stated that there were few young people at Share in Boston. Well, my boss is there, as well as his boss. Not doing much for the technical skills of the team, is it? Management.
Another culprit is the "unsexiness" of the mainframe. I think it kicks ass. It's the best hardware you can buy. It can do everything the new toys can do, and all the old stuff. It's really an amazing box. I'm not sure what to do to entice in-house talent to take up the reigns on the mainframe. For all the new stuff they want to learn ("Hey, there's a new protocol? Why aren't we using it?"), they sure don't want to learn this. There's a touch of hubris involved. I also think they are intimidated.
Hurrah! Without quoting Jane Jacobs or anything like that, I think this quest for accreditation or certification is sad. I recently saw a (what I see as) ridiculous post on the IBM-MAIN listserv, and here is a fragment:
I am planning to do a IBM certification for DB2.
I am not aware of DB2 at all. Presently, I am working on CICS-IDMS combination.
I want to add a feather in the cap by learning DB2. Also, let me know are there different courses for DBA and Application developer.
I am not aware of DB2 at all. I don't know whether to laugh or cry.
You're missing my point. I'm saying that our theories and our knowledge are what they are. Take gravity, for example. If the theory of gravity were to take a wild turn (say, it's implicit in some new particle, or some aliens control it), most "physical" things (as opposed to ideas) based upon gravity would not explode, or float away, upon our realization or discovery of a new scheme. Perhaps the new knowledge would be a great benefit with potential for new applications (new particle), or it would give us pause (aliens - what if the aliens die?!). But until that new knowlege arrives (through hard work or serendipity), it doesn't matter.
Regardless, new ideas - great. Bust 'em out and put them through their paces. I could go for a better life.
Are planes going to drop from the sky? Will we be thrown out of orbit? This sounds like the Bugs Bunny cartoon where Bugs floats on air because he never studied the laws of gravity (I know I've probably got the reference wrong, but you get the idea).
Your experiment fits the model, or it doesn't. If it doesn't then one or both need to be tweaked, or scrapped.
If this is a mainframe, then the data (DASD) is probably elsewhere. I didn't catch anywhere exactly what was taken, but unless one was a Shark or EMC box, there's no data in danger. It's like someone stealing your motherboard.
I'm a mainframe administrator, in his low 30s, and whenever there is a story about mainframe employment, it generally degenerates into the following:
1) COBOL?!? It's so boring! 2) They don't pay newbies enough! (see #3) 3) They expect me to have experience! (see #2) 4) They should move it to a Beowulf cluster!
Regarding (1), administering a mainframe has very little COBOL, if any at all. That's for programmers. If any of my co-administrators deals with COBOL, it's because she (yes, she in this case) is the one who handles everything to do with configuring the COBOL compilers and related products.
(2 & 3) Why do you think the oldsters are rolling in the bucks? My employer (govt) is not hiring anyone, let alone mainframers. They're firing. But, the oldsters' 401k's have tanked, so they are not going anywhere for a while. Also, when the majority of the youngsters have such a shitty attitude and an unwillingness to learn "legacy", why would we want to hire them? We have tons of Java programmers creating applications for WebSphere running on the mainframe, and they treat the mainframe like a joke. Yeah, I'm gonna want to train one of those bozos. They don't have a clue.
(4) Yeah, right.
Mainframe dying? Doubt it. Sure, you can point to some big server farms out there as "successful" (I don't know... Amazon?). But, who's using mainframes? Big banks. Govt. Big industry. The people who have been using mainframes for decades, and have been making a ton of money for decades. You want them to change how they do business? You're going to walk up to Andrew Carnegie or JP Morgan and say, "You should lose those mainframes!" He'll light his cigar with a thousand dollar bill and say, "Beat it, kid."
This is a person complaining about a mainframer refusing to learn anything new? I'm incensed. Talk about the pot calling the kettle black.
I'm 34, an administrator on a z900 running OS390, and it rocks. Please explain what's tough about creating test data? I think we get to the root of the problem - "your code is wrong".
EVERYTHING you complained about above pertains to your coding and the language (COBOL), NOT the platform. On my box, you can do COBOL, Assembler, Java, C, C++, Perl, you name it. All the "programmers" depend on their IDEs to develop their code - they wouldn't know what's going on in there if you gave them a dump. Ooh! Wait! The dump isn't in Java! I'm confused! It's using numbers! What kind of wacky computer is this?
I've determined that Moore's law is not driven by technical innovation, but simply by the need to keep up with shitty programming.
Well, you do need administrators, but the number is essentially fixed. I guess we have 8 people administering our mainframe (two z900s in a parallel sysplex, about 6 LPARS). If we were to add 10 engines to each of those boxes, or add 10 z900s to the sysplex, or both (a ton of power), how many people would we need to administer it? Those same 8. And needing all 8 is definitely arguable.
We have web content being served both from AIX boxes and from my z900 (a mainframe, ok?). Mainframes are GREAT webservers, and I'm not talking a Linux LPAR. I could take the 20 million hit/month load from the AIX box, put it on my one z900, barely notice it, lose the AIX box (I think it's a S80) and the two people who are administering that AIX box, and go on my merry way. Why doesn't that happen? Because I work for the government and we are all about burning money. Why doesn't private industry do this? I guess that's the magic question.
And, they have been working for AT LEAST a year on transferring their AIX web stuff to the S80 (It's actually on a smaller box now). We moved from 5 G4 boxes to 2 z900 boxes over a weekend.
Sour grapes. The only reason there are not new mainframers is because of the ignorance and arrogance of the up-and-coming programmers, in my opinion. What happened to education? Computers are 1's and 0's. Yes, there will be a learning curve, but it only gets steep for the close-minded.
I have Java programmers who whine for us to get a Linux LPAR, but when I try to talk to them about things such as filesystems, or anything which is fairly universal in the world of computers, and they are clueless, which shows they don't even know their beloved Linux (I love Linux, by the way).
So, is it the frozen mindset of the programmers which is to blame, or the cads who are teaching them?
And, c'mon... COBOL is EASY. Java has a much steeper initial learning curve.
Whenever I try to get my employer to try some open source app, the main thing they scream about is support. If it breaks, they want someone to blame/fix it, and they are willing to pay big bucks for that.
Please don't post news like this... I work on a S/390 administering WebSphere and USS, and I'm seriously considering bailing to do an AIX job, for viability reasons. Don't tell me Z/OS390 is becoming fashionable. I'm already wishy-washy.
Aaron
Slashdot Mirror
I don't know facebook's intentions, but I would bet dollars to donuts that the "target audience" option was there *specifically* to allow advertisers to target women, not exclude them. Can I imagine an all-female employer, a female-centric organization, not wanting to spend their dollars on marketing towards men? Maybe they'll make the gender-targeting option only available to self-described women. Yeah, that's fair.
Cary Fowler is engaged in the Noah-like task of gathering the seeds of some two million varieties of food plants
Read more http://www.newyorker.com/reporting/2007/08/27/070827fa_fact_seabrook
Think of it as setting a good example for your kids. Sounds like your real problem is your job. Think about it, and fix it.
... or some other delivery company, get them make all left turns, and then they can rotate out their tires to each other.
I can see their program hang when it tries to process a required left-hand turn from one one-way street to another.
Well, the author (of the article, Peter Davidson) doesn't know mainframes. "...IBM OS390 mainframe running ZOS ..." OS390 is an operating system (the precursor to zOS). Maybe they're running on a z900 or z990 or z9 (or 2064-yada yada). So, the way I interpret this, is they are running some CICS (presumably TS 1.3, 2.3, or 3.1) which talks to DB2, does some messaging with WebSphere MQ, and all the web interface (WebSphere, whatever) is on some Unixy (Solaris?) front end. Speaking as a IBM mainframer, any zOS or CICS systems programmer worth their salt would be able to tell you how long all of those transactions took, and where the problem was. I guarantee we run a much smaller box here, and we push through millions of transactions a day (just business hours!), and we guarantee the majority of the transactions are done in 1/4 second or less. So, shitty code, shitty performance, but on the backend they should be able to pinpoint any problems. I would guess that the bottleneck (and money pit) showed up on the front end (or everyone on the project is clueless - 50/50).
This is business computing. You take some data from point A, and move it to point B, with perhaps a little transformation in between. Not that tough, right? They have their way of doing it. You have your way. What's wrong with their way? What great benefit will be drawn from doing it your way? Because you can't do COBOL or Assembler? Why not?
Quite right. In our shop, we guarantee that 75% of our CICS transactions will run in 1/5 of a second. 90% in less than 1/2 second. It flies. And it's not just the hardware. It's the programming. zOS is only as good as the people who run it.
Aaron
Being a zOS systems programmer in my mid-30s, having done this for about 7 years, I think the blame for any shortage lies with management, with some blame going to universities for dropping the ball. There is plenty of education available out there, either from IBM or other sources, but your boss (or school) has to get you there.
The article stated that there were few young people at Share in Boston. Well, my boss is there, as well as his boss. Not doing much for the technical skills of the team, is it? Management.
Another culprit is the "unsexiness" of the mainframe. I think it kicks ass. It's the best hardware you can buy. It can do everything the new toys can do, and all the old stuff. It's really an amazing box. I'm not sure what to do to entice in-house talent to take up the reigns on the mainframe. For all the new stuff they want to learn ("Hey, there's a new protocol? Why aren't we using it?"), they sure don't want to learn this. There's a touch of hubris involved. I also think they are intimidated.
Management should wake up. zSeries is incredible.
Aaron
Hurrah! Without quoting Jane Jacobs or anything like that, I think this quest for accreditation or certification is sad. I recently saw a (what I see as) ridiculous post on the IBM-MAIN listserv, and here is a fragment:
I am planning to do a IBM certification for DB2.
I am not aware of DB2 at all. Presently, I am working on CICS-IDMS combination.
I want to add a feather in the cap by learning DB2. Also, let me know
are there different courses for DBA and Application developer.
I am not aware of DB2 at all. I don't know whether to laugh or cry.
I'm venting. I apologize.
You're missing my point. I'm saying that our theories and our knowledge are what they are.
Take gravity, for example. If the theory of gravity were to take a wild turn (say, it's implicit in some new particle, or some aliens control it), most "physical" things (as opposed to ideas) based upon gravity would not explode, or float away, upon our realization or discovery of a new scheme. Perhaps the new knowledge would be a great benefit with potential for new applications (new particle), or it would give us pause (aliens - what if the aliens die?!). But until that new knowlege arrives (through hard work or serendipity), it doesn't matter.
Regardless, new ideas - great. Bust 'em out and put them through their paces. I could go for a better life.
Are planes going to drop from the sky? Will we be thrown out of orbit? This sounds like the Bugs Bunny cartoon where Bugs floats on air because he never studied the laws of gravity (I know I've probably got the reference wrong, but you get the idea).
Your experiment fits the model, or it doesn't. If it doesn't then one or both need to be tweaked, or scrapped.
If this is a mainframe, then the data (DASD) is probably elsewhere. I didn't catch anywhere exactly what was taken, but unless one was a Shark or EMC box, there's no data in danger. It's like someone stealing your motherboard.
Thanks,
Aaron
I'm a mainframe administrator, in his low 30s, and whenever there is a story about mainframe employment, it generally degenerates into the following:
1) COBOL?!? It's so boring!
2) They don't pay newbies enough! (see #3)
3) They expect me to have experience! (see #2)
4) They should move it to a Beowulf cluster!
Regarding (1), administering a mainframe has very little COBOL, if any at all. That's for programmers. If any of my co-administrators deals with COBOL, it's because she (yes, she in this case) is the one who handles everything to do with configuring the COBOL compilers and related products.
(2 & 3) Why do you think the oldsters are rolling in the bucks? My employer (govt) is not hiring anyone, let alone mainframers. They're firing. But, the oldsters' 401k's have tanked, so they are not going anywhere for a while. Also, when the majority of the youngsters have such a shitty attitude and an unwillingness to learn "legacy", why would we want to hire them? We have tons of Java programmers creating applications for WebSphere running on the mainframe, and they treat the mainframe like a joke. Yeah, I'm gonna want to train one of those bozos. They don't have a clue.
(4) Yeah, right.
Mainframe dying? Doubt it. Sure, you can point to some big server farms out there as "successful" (I don't know... Amazon?). But, who's using mainframes? Big banks. Govt. Big industry. The people who have been using mainframes for decades, and have been making a ton of money for decades. You want them to change how they do business? You're going to walk up to Andrew Carnegie or JP Morgan and say, "You should lose those mainframes!" He'll light his cigar with a thousand dollar bill and say, "Beat it, kid."
Like Mr. Sutton said, "It's where the money is."
This is a person complaining about a mainframer refusing to learn anything new? I'm incensed. Talk about the pot calling the kettle black.
I'm 34, an administrator on a z900 running OS390, and it rocks. Please explain what's tough about creating test data? I think we get to the root of the problem - "your code is wrong".
EVERYTHING you complained about above pertains to your coding and the language (COBOL), NOT the platform. On my box, you can do COBOL, Assembler, Java, C, C++, Perl, you name it. All the "programmers" depend on their IDEs to develop their code - they wouldn't know what's going on in there if you gave them a dump. Ooh! Wait! The dump isn't in Java! I'm confused! It's using numbers! What kind of wacky computer is this?
I've determined that Moore's law is not driven by technical innovation, but simply by the need to keep up with shitty programming.
I apologize, but this really cheeses me off.
Well, you do need administrators, but the number is essentially fixed. I guess we have 8 people administering our mainframe (two z900s in a parallel sysplex, about 6 LPARS). If we were to add 10 engines to each of those boxes, or add 10 z900s to the sysplex, or both (a ton of power), how many people would we need to administer it? Those same 8. And needing all 8 is definitely arguable.
We have web content being served both from AIX boxes and from my z900 (a mainframe, ok?). Mainframes are GREAT webservers, and I'm not talking a Linux LPAR. I could take the 20 million hit/month load from the AIX box, put it on my one z900, barely notice it, lose the AIX box (I think it's a S80) and the two people who are administering that AIX box, and go on my merry way. Why doesn't that happen? Because I work for the government and we are all about burning money. Why doesn't private industry do this? I guess that's the magic question.
And, they have been working for AT LEAST a year on transferring their AIX web stuff to the S80 (It's actually on a smaller box now). We moved from 5 G4 boxes to 2 z900 boxes over a weekend.
What's the problem?
I'm as thick as the Complete Large-Print Dickens
Sour grapes. The only reason there are not new mainframers is because of the ignorance and arrogance of the up-and-coming programmers, in my opinion. What happened to education? Computers are 1's and 0's. Yes, there will be a learning curve, but it only gets steep for the close-minded.
I have Java programmers who whine for us to get a Linux LPAR, but when I try to talk to them about things such as filesystems, or anything which is fairly universal in the world of computers, and they are clueless, which shows they don't even know their beloved Linux (I love Linux, by the way).
So, is it the frozen mindset of the programmers which is to blame, or the cads who are teaching them?
And, c'mon... COBOL is EASY. Java has a much steeper initial learning curve.
And COBOL is faster.
I'm thick as a whale omelette.
"Robert Novak is representing himself in this lawsuit, and thus it is effectively costing him nothing to persue this campaign of harassment. "
You just made the list, buddy! I don't want anybody touchin' my stuff. Anyone touches my stuff, I'll kill ya!
Whenever I try to get my employer to try some open source app, the main thing they scream about is support. If it breaks, they want someone to blame/fix it, and they are willing to pay big bucks for that.
Please don't post news like this... I work on a S/390 administering WebSphere and USS, and I'm seriously considering bailing to do an AIX job, for viability reasons. Don't tell me Z/OS390 is becoming fashionable. I'm already wishy-washy. Aaron
They're also trying to pass a bill to arm public school teachers. Get me out.