I would certainly agree with this sentiment. I use Protel EDA software as a major part of my day job - it's *much* faster to memorise the key-sequences than to choose the correct menu item. Example: 'P-T' (place-track), to start placing tracks. This is much faster than choosing the 3-levels down on the menu to place a track, not to mention that this loses the mouse position, so you have to find it again.
I would contend that Commercial Pressures (tm) cause a *lot* of closed source programs to be released without this kind of error checking. I'm not saying it's OK to write code without error checking, but that the reality of it is that a lot of commercial software is released without error checking.
Open Source software is quite often developed by the programmer with a view toward 'correctness' rather than a view toward a July 16 release date (or whatever). It is this software that I would expect to have a lot of 'good' error checking. A lot of other Open Source software, however, is developed by a programmer for personal use, then released to the community because 'others might find it useful'. It is this software that is likely to have less stringent error checking (not always, but less likely).
So, sure, some OSS will have lax error checking, but there is an awful lot of OSS with exceptional (pun intended) error handling.
I think the generalization is gross at best, and inaccurate and wrong at worst.
This is all "Unix vs Windows", I don't see how this applies to "Open Source vs Closed Source". But, I will reply anyway...
1. Staff will need to be retrained: Yes, in a windows shop, staff *will* need to be retrained. In a Unix shop, the retraining to run Linux is less than the retraining to run WinNT.
2. You will only find competent staff... I don't think you meant this the way it reads. I'm sure there are competent staff for MS Windows; it's just that in my experience the solutions proposed by them involve buying more software and/or servers, rather than running more processes on the Linux box.
3. Downtime costs and refitting costs... Yes, ther e will be some downtime (I'm assuming in changing from Windows to Linux), but how much downtime is there to change from WinNT Server to Win2000 Server? Just as much, if not more, I'd imagine. How long does it take to apply an MS security patch, and do you have to take the server down? I know Win2kPro requires a reboot to install a service pack...
4. Most open source software unproven... Well, I guess that's the case. But then, there is much more open source software to be unproven. That software which *is* proven has proven to be *very* stable and reliable (examples: apache, perl, samba, *BSD, etc)
5. An Open Source OS is vulnerable to Script kiddies. ROFL. Yes, but so is that 'other' OS from Redmond. You must have been living under a rock when code red and nimda were around... I think you'll find that *any* OS is vulnerable, open source or not. Some OSes are more vulnerable than others - OpenBSD, in spite of the name, is one of the most secure OSes around (or, at least, it has that reputation). Windows {NT|Win2k} server does not have a reputation as being *particularly* secure (it's not noted for it, at any rate).
My Dad used to interview and hire a lot of people. He always said that when you look at a reference, you're not looing at what's there - you're looking at what's *not* there. No-one will ever pass on a bad reference.
Slashdot Mirror
I would certainly agree with this sentiment. I use Protel EDA software as a major part of my day job - it's *much* faster to memorise the key-sequences than to choose the correct menu item. Example: 'P-T' (place-track), to start placing tracks. This is much faster than choosing the 3-levels down on the menu to place a track, not to mention that this loses the mouse position, so you have to find it again.
I would contend that Commercial Pressures (tm) cause a *lot* of closed source programs to be released without this kind of error checking. I'm not saying it's OK to write code without error checking, but that the reality of it is that a lot of commercial software is released without error checking.
Open Source software is quite often developed by the programmer with a view toward 'correctness' rather than a view toward a July 16 release date (or whatever). It is this software that I would expect to have a lot of 'good' error checking. A lot of other Open Source software, however, is developed by a programmer for personal use, then released to the community because 'others might find it useful'. It is this software that is likely to have less stringent error checking (not always, but less likely).
So, sure, some OSS will have lax error checking, but there is an awful lot of OSS with exceptional (pun intended) error handling.
I think the generalization is gross at best, and inaccurate and wrong at worst.
This is all "Unix vs Windows", I don't see how this applies to "Open Source vs Closed Source". But, I will reply anyway...
1. Staff will need to be retrained: Yes, in a windows shop, staff *will* need to be retrained. In a Unix shop, the retraining to run Linux is less than the retraining to run WinNT.
2. You will only find competent staff... I don't think you meant this the way it reads. I'm sure there are competent staff for MS Windows; it's just that in my experience the solutions proposed by them involve buying more software and/or servers, rather than running more processes on the Linux box.
3. Downtime costs and refitting costs... Yes, ther e will be some downtime (I'm assuming in changing from Windows to Linux), but how much downtime is there to change from WinNT Server to Win2000 Server? Just as much, if not more, I'd imagine. How long does it take to apply an MS security patch, and do you have to take the server down? I know Win2kPro requires a reboot to install a service pack...
4. Most open source software unproven... Well, I guess that's the case. But then, there is much more open source software to be unproven. That software which *is* proven has proven to be *very* stable and reliable (examples: apache, perl, samba, *BSD, etc)
5. An Open Source OS is vulnerable to Script kiddies. ROFL. Yes, but so is that 'other' OS from Redmond. You must have been living under a rock when code red and nimda were around... I think you'll find that *any* OS is vulnerable, open source or not. Some OSes are more vulnerable than others - OpenBSD, in spite of the name, is one of the most secure OSes around (or, at least, it has that reputation). Windows {NT|Win2k} server does not have a reputation as being *particularly* secure (it's not noted for it, at any rate).
6. You may have a point there.
My Dad used to interview and hire a lot of people. He always said that when you look at a reference, you're not looing at what's there - you're looking at what's *not* there. No-one will ever pass on a bad reference.