I know I am a traditionalist, but I would say drag+drop is not "coding". they may be programming. But when I think of coders I think of people actually typing in code at a keyboard.
And although I don't agree with the slant the person you repsoned to was taking (I see macro programming very similar to the high level IDE programming going on these days), I can see not calling it coding.
And so yes by my defnition, if they are typing in macro scripts all by themselevs, then yes by my definition I see them as coding.
One of the things he doesn't seem to get is that he seems to think of Sun and Java as being on the good side. I actually trust Microsoft more than I trust Sun (in a twisted sort of way).
At least with Microsoft we know they are anti-Linux. Sun trys to seem all buddy-buddy when around the Linux community but in a lot of their press releases and documentation they bash Linux.
Shoot some people have claimed SCO is just a Microsoft pawn. And although I am not doubting Microtosft is somewhat cheering for SCO, I wouldn't be surprised if Sun wasn't doing the same.
Being in on the original site (wow that was a while ago...), the domain was run by me, and the first version didn't have a hyphen. I cracked up and pointed out the alternative words. hence they launched with the hyphen:).
now give credit where credit is due. IBM may have started it (don't know but I am assuming by your claim that you have done prior research) but Microsoft really is the one to have mastered it. Don't you recall all of the jokes saying there was a lawyer for every programmer at Micosoft. during the 80's and 90's Microsoft had a knack for lawsuits.
While I will give you that there are a lot of people who act "like lunatics" on/. (I would prefer the term childish -- and I don't think it is limited to/.), I think there is a more practical reason to see that sort of bias in this particular case.
personal experience. I am fairly sure most people reading slashdot have have far more sucess with linux being stable, reliable, and able to interoperate with things rather than windows.
I know in my experience most windows services need to be babysat a WHOLE LOT more than my linux services. the base OS can be stable (I generally keep my NT workstation up for weeks at a time without an issue) but combine it with the apps written for it, and probably most importanly the mindset of automation-enabled isn't important.
maybe they were better on the replay (I caught it via wednesday 7pm PST -- recorded on a VCR). (for the 2 hours, I only recall fast-forwarding through about 20-25 minutes of commercials)
The commercial breaks were somewhat longer than usual, but I seem to recall that the initial part went on for I think around 10 minutes before a commercial break, it may have been even closer to 15-20 minutes.
on the veritas thing. Sun did approve it, so I really don't consider it that 3rdparty. they used to have a joint effort site up at www.vosinitiative.com. And yes I will give you it may be unfair, but we do all tend to base viewpoints based on our experiences.
- hardware domains: IBM has been doing a lot of work in this area. And this is realyl only usefull if you have a big box doing mainframe style stuff. I guess I don't put a whole lot of value on it as you generally get more bang/buck on clusters of smaller servers.
- ECC: funny you mention that, Sun had to add parity checking on the Ex500 procs due to a nasty chip fault (they force some customers to sign an NDA, it stayed hidden from the press for about 18 months before some folks blew the whistle and it beamce public).
- error reporting. that is a factor more to do with the hardware than the OS. and you even seem to admit to it being a sparc hardware thing.
- firmware -- again hardware
on the DR (dynamic reconfigure) -- yes I ahev heard of it (it was actually those manuals/addendums I was referring two when they (sun) basically said don't use it.
I will grant you that the Sparc hardware has some interesting features, and with the latest generations for some applications the performance is great. But the point was about Solaris. I have far more issues with Solaris than I do Sparc hardware. And yes there is a difference between the two. I will say this, I wasn't that thrilled with the performance on the Sun boxes I was exposed to. Our intel boxes smoked them (running Oracle doing OLTP). Sounds like the newer generation is better.
And again back to Solaris vs. Sun Saprc boxen. My gripes are about the huge Solaris is so wonderfull and so superior to Linux without much (if any) concrete backing.
hmmm... I think you are trolling, but I'll entertain:), how about:
- a real reliable automatable install (speaking of RedHat's kickstart here) -- jumpstart is a joke by comparison.
- CIDR compliance. the funniest comemnt I ever read in a Solaris config file was in Solaris8 (back in 2001ish) about how maybe we should support that CIDR thing soon. -- goes along with decent network conifguration scripts
- workable software raid out of the box. not to mention *gasp* shipping a C compiler.
Yes the first two features I mentioned are not as important if you only have one or two boxes (although I still feel they do matter as relying on tape backups to reproduct the config of a machine is asking for trouble). But what about networks on machines. I can't imagine having to spend the hours tweaking every single box I install manually. -- Yes you could home brew your own, but it is nice when you can use standard scripts that do a lot of the work for you. (And before you say it -- yes I can do the scripts myself if I needed to, and I have in the past on older systems, but I enjoy being able to improve other areas rather than having to re-invent the wheel on so much basic stuff)
And even if I COULDN'T name any (I only spent a minute thinking so far) the point I am making is the previous poster claimed Solaris has had and continues to have "enterprise" features 5-10 years before Linux.
Plus, you also can get down to the "what is meant by Enterprise features?" question.
And if you read soem of my previous posts on this thread, you will notice I mentioned a nasty experience with Solaris on Sparc. We had two dedicated admins on 2 solaris boxes. And they still couldn't keep up and stuff went boom. around the same time we had a 30:1 ratio of linux servers to admins and (knock on wood) have yet to have a similar "experience".
Ummm, Last I checked this century includes 2001. The E[3456]500 line was the main sun line in the "mid range" market (with the E10K's) in the high end up through that time. I don't recall exactly when these no longer were the headline candidate, but I know at least in mid 2001 (when we had the fun in question).
And so that still doesn't address the point. okay, say it has CPU hot plug NOW. the poster I responded to (like a lot of Solaris zealots I see on/.) talk in hand wavy terms about all of these "enterprise" featutes Solaris has had and continues to have 5-10 years before Linux does. And I haven't really seen any facts to back these claims. It so far has less basis than Microsoft's claims about Windows vs. Linux Security.
try 2001/2. the E[3456]500 line was the main sun line in the "mid range" market (with the E10K's) in the high end up through that time. I know at least in mid 2001 (when we had the fun in question).
oh, and your analogy is flawed as well, try it would be like you saying "linux is still for technical OS hobbyists" that is where Linux started.
Oh, and you still aren't answering the main question I asked about "what features"?
okay, I think you are trolling, but I will give you the benefit of the doubt.
What stability/reliability? worst data disaster we ever had was due to Sun techs on Solaris 8 on E4500's running Veritas clustering.
And care to give some tangible examples of the "advanced features" that are actually present/work? (don't say things like hot plugging of CPUs -- read your E4500 manual if you think Sun Hardware + Solaris supports this).
now I am a command line guy for more things that most folks. (I like being able to use things like a pipe or a single command to do a fileop that would take thousands of clicks in a GUI, but that is another discussion) back to CD burning.
In both rh6.2 and rh7.3, cd burning has been cake!. used both scsi and ide burners under rh6.2, ide under 7.3. In all cases just have mkisofs and cdrecord instaleld and then do: mkisofs -J -T -R -v -o cdrecord -v dev=0,0,0 (the dev may vary depending on your hardware config, than is what cdrecord -scanbus is for)
Have no idea about burning under a RH GUI but see above:)
still in use routing packets. one of these days will replace with a mini-itx system (mainly to drop power usage), but been running like a champ (running I think 2.0.38 and ipfwadm)
so why the FSCK don't java programmers use the "security model". I have yet to see one java server application using it, and I don't know of any client applications using it correctly.
And that assumes that the security code is perfect in Java. if there is a bug in that security code you are screwed. And given the (lack of) quality programming in other core JVM subsystems (like memory management, threading, etc.) in various JVM versions (personal experience in the 1.1,1.2 and 1.3 series) I strongly doubt the security code is something to hold a lot of faith in.
Shott I seem to recall nasty brokeness in the security isolation of earier netscape JVM's.
I hear a LOT of talk about oh but with Java you will be secure, but in practice I just haven't seen any results.
What has worked is having good (including meaning they are smart and technical -- aka can actually understand what a computer does) programmers working on the code.
As mentioned in other forums, the article's summary is not accurate. sitefinder is NOT being shut down. and it spite of VeriSlime's press releases to the contrary they were not asked to.
what they WERE asked to do (and have now done) is to drop the.net and.com wildcards. aka they have (for now at least) release.com and.net from being hostages.
I always have to laugh when I here people say linux documentation sucks.
I will agree if you say in general ALL documentation sucks. Now I personally find Linux docs to actually *gasp* suck less (actually a lot less for some) than a lot of other docs I have used.
And in case you are wondering: windows admin, windows programming, dos programming, solaris admin, oracle admin to name a few.
And some docs/books in the various categories are better than others. Like using 3rd party libraries when you are programming the key is find the libs (or docs in this case) what are good.
The big part of this is in learning how to learn. I guess I am lucky, I have a engineering mindset, and so I am able to do self teaching fairly easily.
I will agree though that the core cultures are different. In the MS world admins are trained to just point and lick through canned solutions and not really understand the low level of how things work (and thus not have much real control). Whereas in unix (and even more so in Linux -- if you don't buy that try reading the docs for VCS on Solaris sometime) you are not prevented from knowing what is really going on so you can tweak, tune, cutomize, automate to your hearts content.
One question, you say you can do "any reasonable system administration task in Windows and/or DOS". how can you do DOS admin without using a command line: "As for *ix, I know there exists a command line prompt, and the operator can do many things provided he knows what to do there. I am not one of these people"
So were you lumping DOS into your skillset simply becuase windows (3.1 and 9x) runs on it, or have you done DOS admin?
Reason I ask is I came from the DOS world (good old DOS 1.0 on the original IBM PC -- what you get when yor dad is an IBMer) -- well technically I started on the Commadore Pet, but let's not go there:). and I grew up using a command line. So adopting to Unix wasn't that bad for me (and being a technical engineer guy it comes naturally).
I am not trying to flame, I truely don't understand how someone could have been proficient in DOS and not be conformatable with a command line.
Oh, don't worry I understand the IDEA of ACID. I just know that in pratice you rarely have that "prefect" of an application usage. Some constraints won't have been there from day 1, some get dropped, you have to import data from other companies, new features get added so that older records are missing some now mandatory columns, etc. Also having lots of contraints in the DB can kill you on performance. I know of one large DB company that deals with constraints in the application layer due to performance reasons.
The blow a fuse in the middle of 200 transactions is the D part. I understand that one. From personal experience even some of the "big guys" like Oracle don't hold up to the D. I wasn't personally the DBA on that mess but was working with them. Short story is Oracle wigged out and corrupted DB files (running on on the whole VOS inititive setup) and we had to go to backup files.
So while the thoery makes some sense, in practice I haven't seen the real gain over the long haul (for the ACI aka transaction part). for the D part, the vendor I have actually had the best business luck with has been MySQL first, MSSQL/Sybase second, and Oracle third.
sorry, maybe it is just me, but the whole "ARRGG IT AIN'T ACID" is a lot of hype to me. ACID boils down to transactions. plain and simple.
And I have found in many applications it is easier to deal with transaction type data consistency at the app layer instead of the db one.
knowing that a DB transaction is complete doesn't help you if for in order to move forward you have to have db ops done in mtuliple servers and/or a change happen with an external vendor.
And generally some bad code/process will at some point munge your data in a similar way as if you had a db crash in the middle of a transaction.
I have generally seen for most applciations you are better off just coding things to treat outside input (including data from a db) as evil until you have verified it and cope with the abnormalities.
Yes there are exceptions, but ACID tends to be a knee jerk reaction, and most people realyl need to be askign themselves what it ACTUALLY buys them.
excuse me??? have you ever done code generation in perl?
I have done a lot of it (and I don't think it has every been generating perl code). Mostly been for "code" in the larger sense. See Pragmatic Programmer for details.
I don't know lisp, so I don't know if it would be better,but your statements are not correct. you don't have to write a prse nor a language generator.
we tossed together a very simple system for Kickstart file generation (automation scripts for RedHat's anaconda installeD). basically a bunch of perl code snippets that get require'd in. and soem config files setting variables (using perl syntax) that get eval'd. the system is under 200 lines or perl (including option processing, documentation and the like).
Slashdot Mirror
I know I am a traditionalist, but I would say drag+drop is not "coding". they may be programming. But when I think of coders I think of people actually typing in code at a keyboard.
And although I don't agree with the slant the person you repsoned to was taking (I see macro programming very similar to the high level IDE programming going on these days), I can see not calling it coding.
And so yes by my defnition, if they are typing in macro scripts all by themselevs, then yes by my definition I see them as coding.
One of the things he doesn't seem to get is that he seems to think of Sun and Java as being on the good side. I actually trust Microsoft more than I trust Sun (in a twisted sort of way).
At least with Microsoft we know they are anti-Linux. Sun trys to seem all buddy-buddy when around the Linux community but in a lot of their press releases and documentation they bash Linux.
Shoot some people have claimed SCO is just a Microsoft pawn. And although I am not doubting Microtosft is somewhat cheering for SCO, I wouldn't be surprised if Sun wasn't doing the same.
Being in on the original site (wow that was a while ago...), the domain was run by me, and the first version didn't have a hyphen. I cracked up and pointed out the alternative words. hence they launched with the hyphen :).
Agreed. Although us in the tech crowd know RedHat would do the right thing IF any infringing code is found this is a big warm and fuzzy for PHB's.
And I agree it should show good faith if things get to a court room.
now give credit where credit is due. IBM may have started it (don't know but I am assuming by your claim that you have done prior research) but Microsoft really is the one to have mastered it. Don't you recall all of the jokes saying there was a lawyer for every programmer at Micosoft. during the 80's and 90's Microsoft had a knack for lawsuits.
While I will give you that there are a lot of people who act "like lunatics" on /. (I would prefer the term childish -- and I don't think it is limited to /.), I think there is a more practical reason to see that sort of bias in this particular case.
personal experience. I am fairly sure most people reading slashdot have have far more sucess with linux being stable, reliable, and able to interoperate with things rather than windows.
I know in my experience most windows services need to be babysat a WHOLE LOT more than my linux services. the base OS can be stable (I generally keep my NT workstation up for weeks at a time without an issue) but combine it with the apps written for it, and probably most importanly the mindset of automation-enabled isn't important.
maybe they were better on the replay (I caught it via wednesday 7pm PST -- recorded on a VCR). (for the 2 hours, I only recall fast-forwarding through about 20-25 minutes of commercials)
DS3 is ~45Mbit/sec bi-directional
(so 20 is about 44% utilized)
The commercial breaks were somewhat longer than usual, but I seem to recall that the initial part went on for I think around 10 minutes before a commercial break, it may have been even closer to 15-20 minutes.
on the veritas thing. Sun did approve it, so I really don't consider it that 3rdparty. they used to have a joint effort site up at www.vosinitiative.com. And yes I will give you it may be unfair, but we do all tend to base viewpoints based on our experiences.
- hardware domains: IBM has been doing a lot of work in this area. And this is realyl only usefull if you have a big box doing mainframe style stuff. I guess I don't put a whole lot of value on it as you generally get more bang/buck on clusters of smaller servers.
- ECC: funny you mention that, Sun had to add parity checking on the Ex500 procs due to a nasty chip fault (they force some customers to sign an NDA, it stayed hidden from the press for about 18 months before some folks blew the whistle and it beamce public).
- error reporting. that is a factor more to do with the hardware than the OS. and you even seem to admit to it being a sparc hardware thing.
- firmware -- again hardware
on the DR (dynamic reconfigure) -- yes I ahev heard of it (it was actually those manuals/addendums I was referring two when they (sun) basically said don't use it.
I will grant you that the Sparc hardware has some interesting features, and with the latest generations for some applications the performance is great. But the point was about Solaris. I have far more issues with Solaris than I do Sparc hardware. And yes there is a difference between the two. I will say this, I wasn't that thrilled with the performance on the Sun boxes I was exposed to. Our intel boxes smoked them (running Oracle doing OLTP). Sounds like the newer generation is better.
And again back to Solaris vs. Sun Saprc boxen. My gripes are about the huge Solaris is so wonderfull and so superior to Linux without much (if any) concrete backing.
hmmm... I think you are trolling, but I'll entertain :), how about:
- a real reliable automatable install (speaking of RedHat's kickstart here) -- jumpstart is a joke by comparison.
- CIDR compliance. the funniest comemnt I ever read in a Solaris config file was in Solaris8 (back in 2001ish) about how maybe we should support that CIDR thing soon. -- goes along with decent network conifguration scripts
- workable software raid out of the box. not to mention *gasp* shipping a C compiler.
Yes the first two features I mentioned are not as important if you only have one or two boxes (although I still feel they do matter as relying on tape backups to reproduct the config of a machine is asking for trouble). But what about networks on machines. I can't imagine having to spend the hours tweaking every single box I install manually. -- Yes you could home brew your own, but it is nice when you can use standard scripts that do a lot of the work for you. (And before you say it -- yes I can do the scripts myself if I needed to, and I have in the past on older systems, but I enjoy being able to improve other areas rather than having to re-invent the wheel on so much basic stuff)
And even if I COULDN'T name any (I only spent a minute thinking so far) the point I am making is the previous poster claimed Solaris has had and continues to have "enterprise" features 5-10 years before Linux.
Plus, you also can get down to the "what is meant by Enterprise features?" question.
And if you read soem of my previous posts on this thread, you will notice I mentioned a nasty experience with Solaris on Sparc. We had two dedicated admins on 2 solaris boxes. And they still couldn't keep up and stuff went boom. around the same time we had a 30:1 ratio of linux servers to admins and (knock on wood) have yet to have a similar "experience".
Ummm, Last I checked this century includes 2001. The E[3456]500 line was the main sun line in the "mid range" market (with the E10K's) in the high end up through that time. I don't recall exactly when these no longer were the headline candidate, but I know at least in mid 2001 (when we had the fun in question).
/.) talk in hand wavy terms about all of these "enterprise" featutes Solaris has had and continues to have 5-10 years before Linux does. And I haven't really seen any facts to back these claims. It so far has less basis than Microsoft's claims about Windows vs. Linux Security.
And so that still doesn't address the point. okay, say it has CPU hot plug NOW. the poster I responded to (like a lot of Solaris zealots I see on
try 2001/2. the E[3456]500 line was the main sun line in the "mid range" market (with the E10K's) in the high end up through that time. I know at least in mid 2001 (when we had the fun in question).
oh, and your analogy is flawed as well, try it would be like you saying "linux is still for technical OS hobbyists" that is where Linux started.
Oh, and you still aren't answering the main question I asked about "what features"?
okay, I think you are trolling, but I will give you the benefit of the doubt.
What stability/reliability? worst data disaster we ever had was due to Sun techs on Solaris 8 on E4500's running Veritas clustering.
And care to give some tangible examples of the "advanced features" that are actually present/work? (don't say things like hot plugging of CPUs -- read your E4500 manual if you think Sun Hardware + Solaris supports this).
now I am a command line guy for more things that most folks. (I like being able to use things like a pipe or a single command to do a fileop that would take thousands of clicks in a GUI, but that is another discussion) back to CD burning.
:)
In both rh6.2 and rh7.3, cd burning has been cake!.
used both scsi and ide burners under rh6.2, ide under 7.3. In all cases just have mkisofs and cdrecord instaleld and then do:
mkisofs -J -T -R -v -o
cdrecord -v dev=0,0,0
(the dev may vary depending on your hardware config, than is what cdrecord -scanbus is for)
Have no idea about burning under a RH GUI but see above
yup a DX-50 not a DX2-50.
still in use routing packets. one of these days will replace with a mini-itx system (mainly to drop power usage), but been running like a champ (running I think 2.0.38 and ipfwadm)
so why the FSCK don't java programmers use the "security model". I have yet to see one java server application using it, and I don't know of any client applications using it correctly.
And that assumes that the security code is perfect in Java. if there is a bug in that security code you are screwed. And given the (lack of) quality programming in other core JVM subsystems (like memory management, threading, etc.) in various JVM versions (personal experience in the 1.1,1.2 and 1.3 series) I strongly doubt the security code is something to hold a lot of faith in.
Shott I seem to recall nasty brokeness in the security isolation of earier netscape JVM's.
I hear a LOT of talk about oh but with Java you will be secure, but in practice I just haven't seen any results.
What has worked is having good (including meaning they are smart and technical -- aka can actually understand what a computer does) programmers working on the code.
As mentioned in other forums, the article's summary is not accurate. sitefinder is NOT being shut down. and it spite of VeriSlime's press releases to the contrary they were not asked to.
.net and .com wildcards. aka they have (for now at least) release .com and .net from being hostages.
what they WERE asked to do (and have now done) is to drop the
I always have to laugh when I here people say linux documentation sucks.
I will agree if you say in general ALL documentation sucks. Now I personally find Linux docs to actually *gasp* suck less (actually a lot less for some) than a lot of other docs I have used.
And in case you are wondering: windows admin, windows programming, dos programming, solaris admin, oracle admin to name a few.
And some docs/books in the various categories are better than others. Like using 3rd party libraries when you are programming the key is find the libs (or docs in this case) what are good.
The big part of this is in learning how to learn. I guess I am lucky, I have a engineering mindset, and so I am able to do self teaching fairly easily.
I will agree though that the core cultures are different. In the MS world admins are trained to just point and lick through canned solutions and not really understand the low level of how things work (and thus not have much real control). Whereas in unix (and even more so in Linux -- if you don't buy that try reading the docs for VCS on Solaris sometime) you are not prevented from knowing what is really going on so you can tweak, tune, cutomize, automate to your hearts content.
One question, you say you can do "any reasonable system administration task in Windows and/or DOS". how can you do DOS admin without using a command line: "As for *ix, I know there exists a command line prompt, and the operator can do many things provided he knows what to do there. I am not one of these people"
:). and I grew up using a command line. So adopting to Unix wasn't that bad for me (and being a technical engineer guy it comes naturally).
So were you lumping DOS into your skillset simply becuase windows (3.1 and 9x) runs on it, or have you done DOS admin?
Reason I ask is I came from the DOS world (good old DOS 1.0 on the original IBM PC -- what you get when yor dad is an IBMer) -- well technically I started on the Commadore Pet, but let's not go there
I am not trying to flame, I truely don't understand how someone could have been proficient in DOS and not be conformatable with a command line.
Oh, don't worry I understand the IDEA of ACID. I just know that in pratice you rarely have that "prefect" of an application usage. Some constraints won't have been there from day 1, some get dropped, you have to import data from other companies, new features get added so that older records are missing some now mandatory columns, etc. Also having lots of contraints in the DB can kill you on performance. I know of one large DB company that deals with constraints in the application layer due to performance reasons.
The blow a fuse in the middle of 200 transactions is the D part. I understand that one. From personal experience even some of the "big guys" like Oracle don't hold up to the D. I wasn't personally the DBA on that mess but was working with them. Short story is Oracle wigged out and corrupted DB files (running on on the whole VOS inititive setup) and we had to go to backup files.
So while the thoery makes some sense, in practice I haven't seen the real gain over the long haul (for the ACI aka transaction part). for the D part, the vendor I have actually had the best business luck with has been MySQL first, MSSQL/Sybase second, and Oracle third.
well saying lisp can do everything that perl can do is like saying everything can be boiled down to a turing machine.
I am sure lisp can do some things better than perl, but I would be hard pressed to beleive lisp can do everything better than perl can do it.
sorry, maybe it is just me, but the whole "ARRGG IT AIN'T ACID" is a lot of hype to me. ACID boils down to transactions. plain and simple.
And I have found in many applications it is easier to deal with transaction type data consistency at the app layer instead of the db one.
knowing that a DB transaction is complete doesn't help you if for in order to move forward you have to have db ops done in mtuliple servers and/or a change happen with an external vendor.
And generally some bad code/process will at some point munge your data in a similar way as if you had a db crash in the middle of a transaction.
I have generally seen for most applciations you are better off just coding things to treat outside input (including data from a db) as evil until you
have verified it and cope with the abnormalities.
Yes there are exceptions, but ACID tends to be a knee jerk reaction, and most people realyl need to be askign themselves what it ACTUALLY buys them.
excuse me??? have you ever done code generation in perl?
I have done a lot of it (and I don't think it has every been generating perl code). Mostly been for "code" in the larger sense. See Pragmatic Programmer for details.
I don't know lisp, so I don't know if it would be better,but your statements are not correct. you don't have to write a prse nor a language generator.
we tossed together a very simple system for Kickstart file generation (automation scripts for RedHat's anaconda installeD). basically a bunch of perl code snippets that get require'd in. and soem config files setting variables (using perl syntax) that get eval'd. the system is under 200 lines or perl (including option processing, documentation and the like).
wow. If you think that post was bad you should see some of my previous ones. (And I didn't see much outside of minor grammer/puncuation issues.)
oh and it takes real balls to post an "insult" as an AC. Wow, I am so impressed. why don't you wait until you are grown up before you post again.