I'm glad to see Borland reiterating their commitment to Linux. I have used Delphi since it came out, and Turbo Pascal before that. I have no problem with C (though C++ makes my skin crawl), but Delphi makes the construction of a UI a minor event, not the majority of the app. That's as it should be.
I wonder whether many of the most negative comments on/. stem from fears on the part of the current community of Linux developers that they will soon be drowned in a sea of Windows emigres. It is a well-founded concern. Many of us developing for Windows would like to be developing for Linux, but as commercial developers, cannot simply drop everything to adopt a new OS, new GUI interface, new compiler, and to build a new tool set.
Contrary to the ravings of RMS, I don't find anything wrong with being paid for my work. The single most offensive aspect of the Open Source movement is the tendency of its adherents to froth at the mouth over the notion of any software which is not free. I pay for a distribution, because I prefer to do that, rather than to download for the many hours it takes to obtain one for "free".
As a resident of North America, I am steeped in the principles of a reasonably free market, and in the notion that profit for labor and risk is an honorable thing. OSS folk take note: the market will determine whether or not Borland's commercially offered Linux-based tools survive. I believe that once Kylix hits the streets, we will see a tremendous upswing in the use of Linux, and in the availability of applications for the desktop.
Linux socialists: brace yourselves, the deluge is near.
With the success of commercial tools, as well as free ones, we will see the maturity of Linux in the market, and then MS will truly have something to fear.
WP peaked at 5.1 for DOS, which was an unmitigated piece of crap with a horribly inconsistent and byzantine set of function keys. The show codes function which I always heard praised was there because they never figured out how to clean up after themselves.
WP was, at least through the 6.1 version for Windows, a horrible, bug-ridden disaster which I used only under extreme duress.
Not that Word is the holy grail, either. Actually, the single most productive word processing tool I have ever used was Borland Sprint.
As so often happens, Linux folk seem to be unaware of the realities of non-Linux software and software companies. 1. Inprise, while no M$, is certainly not dying. 2. Interbase is, in fact, in widespread use in enterprise level and embedded applications. 3. The Java IDE may or may not be inferior to that of IBM, but IBM has an abysmal performance record at marketing good software, or of remaining committed to it. Remember OS2?
No, you don't get it. Real commercial software products are not built by macho C programmers who refuse to use a RAD tool, but by programmers who select tools for the productivity they provide, and for the ease of production of solid programs.
Visual environments, for the most part, make the programming of the UI a task so formidable that it dwarfs tha actual application code in many cases. Delphi, C++ Builder, and JBuilder all put the UI development where it belongs: in a simple and quick IDE which lets the programmer focus on coding application code, not UI code.
My only concern with this deal is that Corel may lessen the commitment to continued support for the Windows development language tools. You do remember Windows, don't you? It's that OS which many of us pragmatists do hate, but which still dominates the desktop market.
As to the language issue, he (and I) may find C as ugly as you apparently find Pascal. And C++ bears all the signs of committee work.
Leaving aside personal opinions and aesthetics, however, Pascal is a more straightforward language to parse than C or C++, and is relatively pure on the Borland branch. C would be better than C++, as it would avoid tremendous complexities, but another issue lurks: C, C++, and Java all muddy the same syntactic pool. An approach based on Pascal would not represent yet another pollution of C-flavored syntax.
It's beyond sad when supposedly intelligent operators of a technical site are incapable of a) spelling Uruguay, or b)looking it up, or c) differentiating between the name of the country (Uruguay) and the description of one of its residents (Uruguayan).
Come on guys, I know that America is the most ethnocentic nation on the planet (I grew up here, after all) but even so...
Looking at the rampant negative comments from those who are a) clearly ignorant of the Interbase product, and b) seem to think they have a God-given right to endless free software, it amazes me why any publisher would want to jump into this miasma.
Since the most vocal of you have made it clear, again and again, that you will never buy software, your comments have little relevance to any commercial operation.
What amazes me even more, however, is the ignorance of K&R. Is there no one here but me who is over the age of 25? Code written to K&R standards will be a) relatively straightforward, and b) free from the abominable practices prevalent in C++. Would you feel better if it were written to VC++ practice? Then it would truly be non-portable, inscrutable, and buggy.
Let your brains engage before your keyboards, people. And stop crapping on newcomers to the Open Source "movement." Without new victims, there can be no movement.
The zip code system was publicly funded, ergo the codes are public property. The reproduction of useful lists, however, as with the reproduction of Linux CDs may be commercialized.
The galling part is that I was once told by a USPS official that the codes are not public property, which is, of course, the usual problem with the government forgetting who works for whom.
Throwing a cap on free does not make it have a special meaning.
The zealotry issue arises in part because people foam at the mouth about ideology. When a public corporation invests time and money, and then offers the result without charge, it should be applauded, not disparaged.
Tossing negative comments at a free offering such as this is a great way to dry up the well. And how stupid might that be?
Besides, you may have noticed that the DoJ has some odd views on the subject of free.
Come on, people, let's be as intelligent as we think we are. If a company is run by a Mormon, does that make their software Mormon software? The US (mistakenly supposed to have been founded on principles of religious freedom -- it wasn't -- it was founded to let people practice their own flavor of persecution) has long been a hotbed of narrow and provincial thinking. The small town in which I was raised (in Michigan, not the south) was, until the 30s, restricted (no blacks, Jews, or Catholics, thank you.) It only changed because the law made it do so.
There are good and bad people in every company, country, and religion. Many religions look pretty cockeyed to one not raised within them. All religions look cockeyed to an atheist. Focus on purpose, not packaging.
Supposedly we hate M$ for their bloated and buggy software, not for their money. Yet the current DOJ action is money-motivated, not anything to do with either merchantibility or fitness of purpose, either of which might provide clear grounds for action against them.
I have German ancestry. Having said that, I can also say that my German grandfather was the most narrow-minded, hidebound, and opinionated man I knew while I was growing up. Does that make all Germans the same? Of course not.
On the matter of religion, I believe in a Creator. I do not believe in organized religion. I dislike Churches for the same reason I dislike M$ and governments: large organizations typically give rise to petty tyrants and corruption. In smaller organizations they have nowhere to hide while amassing their power.
Software is written by people, not by churches. From what I know, Executive Software is a company motivated by commercial desires, as is M$, as is Be, Inc., as is Red Hat. If we do not care about the religious beliefs of Bob Young, or of Jean Louis Gasee, why then should we care about the religious beliefs of Craig Jensen?
A mind is a terrible thing to waste. Let's not waste ours on the ignorance of persecution and paranoia.
Doesn't anyone here recall what the recording industry was able to do to DAT technology? The film industry has been paranoid of recording technologies for a long time, and will do the same to DVD. VHS poses little threat, as the quality is so poor. DVD, on the other hand, has the potential (if the encoding is good) to provide a much cleaner signal for both sound and picture.
I anticipate that the film industry will crank up the lobbyists, and that our politicians, ever vigilant for the highest bidder, will do something to damage the value of this technology, as they did with DAT.
I can appreciate that the film industry has IP rights to protect, but when the protection of such rights is used to justify legislating against someone else's IP rights (as was effectively the case with DAT and the notch filter), then justice is not a component of the action, and we should all scream long and loud to our various pols in the hope that sanity may prevail this time.
USB has a horrible weakness with regard to cable length. I won't give a number, because it isn't worth pulling out the spec, but when I looked closely at implementing some peripherals for USB, I found the actuality of the specs to be disappointing, at best.
Nothing of any size or complexity can be put together in USB without hubs, and as others have pointed out:
1. USB requires the presence of a PC 2. USB sucks CPU
If USB is the future of interconnect, we better start placing our hopes in systems which are self-contained.
Even considering that it is on msn.com, this article is overrun with propaganda.
- easier than BeOS??? NOT - implies NTFS is not O/S specific - quarterly service packs? then how is NT 4.0 only on SP5 by now? - almost impossible to screw up on a single O/S machine? um... not in my experience
As one who has lived in Canada, and therefore has some experience with the round robin of union strikes which has long plagued Canada Post, I have to wonder whether this turn of events will provoke yet another strike.
For those who haven't had the pleasure, a postal strike is truly a delight -- no junk mail. On the other hand, paying bills is tough, and creditors expect you to do so in a timely fashion, without using the mail.
Re:In the end, it all comes down to OEM support
on
CNN Installs Linux
·
· Score: 1
Thereby implying that not all the dependencies were met, which takes us back to the point. Your argument is bogus for proper RPMs.
Sounds nice, but I'm talking about RH 6.0 out of the box. If the RPMs aren't proper, then RH has screwed up.
I have installed RHL 6.0 in full, selective, or minamalist configurations, and I have not had a single problem with hangs at boot. Your information is old.
My experience is quite recent, with 6.0, and is at odds with your own. My information is not old, and as is so often the case in talking about Linux deficiencies, your response is arrogant and glib.
If this is not a comparative issue, then why do you keep comparing Linux to Windows? If you want to take Linux on its own, then fine. If you want to compare it to Windows, that is fine, too. But do not compare it to Windows, and then dodge the reply.
I'm not dodging, and my references to Windows were in response to comments in the message to which I was replying.
You are perfectly correct here. My point was that you were jumping up and down and pointing at Linux, when the problem lies with OEMs, not Linux itself. If you want to point fingers, point at your hardware vendor who is not supporting Linux. Linux is practically helpless without their support. I find it completely amazing that it has come as far as it has with practically no OEM support.
OEMs will not support until they perceive a demand. OEMs are, by definition, money-driven, so they cannot simply decide to develop drivers for Linux on a lark. In the beginning, there is never OEM support for a new OS. I won't blame the OEMs because they have no obligation to support any OS, in fact. For a new OS to build interest it is necessary for developers outside the OEM sphere to undertake driver development.
Blaming Linux is not the answer.
You misread me completely. I do not blame Linux; I blame weak and/or absent documentation. In particular I blame distro vendors, such as Red Hat, for failing to provide better docs.
Now there you may well have a legit complaint. I do not have data on how accurate distribution compatability lists are. I just know that I looked for good equipment that was praised as Linux friendly, and Linux worked perfectly.
I have a number of legitimate complaints, whether or not you happen to agree. I also have made sure that my hardware is on the list of supported devices for all of the OSes in which I have an interest. Except for the issue of monitor modes under all but Caldera, I have not found a problem with hardware support under Linux.
Let me state it as clearly as I can here and now: My only complaint with Linux is not with the software -- it is with the altogether inadequate documentation.
I want an alternative to Windows. I would like very much to make Linux my choice, but I can't do so while the documentation leaves me scrambling for answers to problems everywhere I turn.
The first topic on which I looked to the HOWTOs for support was programming a serial port. The document is outdated, and was written by someone who freely admits he doesn't fully understand the issues. It raised more questions for me than it answered.
My sole interest in Linux is as a tool, a platform. I have zero interest in the politics or religion of Linux or Open Source. If the tool gets in my way, I'm not happy. Windows gets in my way every day. That's why I'm driven to seek an alternative.
Too often I find responses here from people who have already achieved a level of comfort and expertise with Linux and who deny the existence of problems. The problems are real. I have so far evaluated four different distros, and each has shortcomings. If I am to commit to Linux as a platform, I need to evaluate on the basis of lost productivity during the learning period. To date that estimate makes the price too high.
Re:In the end, it all comes down to OEM support
on
CNN Installs Linux
·
· Score: 2
RPM may handle dependencies, but that doesn't make things work. Install everything, and the boot goes well. Be selective, and things hang the boot for long periods. I clocked one particular fail-over at over 8 minutes. This is NOT what a newbie needs. In-depth problem solving is an acquired skill. When you're starting, you need something which works, with a minimum of tweaking.
The LDP needs coherent management. It currently suffers from what appears from the outside to be do-nothing management. Linux is good, but it could be great.
Your opinions on documentation are your own. Mine likewise do not yield to your own. We agree to disagree. This is not a comparative issue. Windows docs suck, too. But that does not make the Linux docs acceptable.
Whether the support is from OEMs or MS, the result remains: Windows installation mostly handles adapters better. When installation is unsuccessful, no one cares whose fault it is. Linux wears the black eye, or the distro does. Not the OEM. That's life; that's reality.
I never buy computers pre-assembled or installed. I configure my own, and have for many years. Windows is easier to install. That's based on many installations. It's not good, and I don't like it, and yes, I have a battery of tricks I have learned painfully, but even so, without resorting to those tricks, it's still easier.
I use only components which are well supported. I have pre-checked the claims of the distros, and all of my components and peripherals are claimed to be supported. Of course, that doesn't speak to how well supported they may be.
I have installed (successfully) on my machines: Win95, Win98, WinNT, Caldera, Mandrake, Red Hat 6.0, and BeOS. I multiple boot, and I use the boot manager which comes with BeOS (simpler and more sensible than most, and trivial to install and set up.
My preference? If allthings were equal (and they never are) I prefer BeOS. It installs in just over 5 minutes, and has more of what I want than any of the others, and good docs, besides. (I include 3rd party docs as acceptable alternatives, in all cases.)
Re:There are real installation issues
on
CNN Installs Linux
·
· Score: 3
Let's settle for inexperienced but intelligent newbies. As to your math and science majors, perhaps they, like so many others, fail to RTFM.
I've been working with small computers since 1975, and have written in C, C++, Pascal, assembler, and even Forth. I've designed embedded processor systems and designed and written serial protocols.
In spite of my experience, I have suffered tremendous frustration with Linux. Also, as one who writes user docs, I do RTFM. The source of the problems is as follows:
1) Linux docs are, to be kind, less than wonderful. They make too many assumptions about the background of the reader. I have yet to find a good presentation of disk partitioning strategies, for example.
2) Distros which use the RPM installation offer only one safe option: Install Everything. Selective package installation leads to things which don't work, boots which stall for many minutes, and a grand variety of other mysteries.
3) Red Hat, in particular, has an installation process which is easily fouled by the user making an unexpected response. Why would the user do that? Because the instructions and/or prompts are inadequate.
4) X server installation and setup is a very interesting source of problems. My video cards and monitors are capable of 1600x1200, but Caldera is the only distro to have correctly set up that configuration on my machines. And talk about user-hostile: tweaking in Hz, KHz, and microseconds is definitely not for newbies. I'm a hardware designer with a lot of experience with deflection systems, and I still don't want to go there.
5) Sending a newbie to the HOWTOs is a great way to send someone postal. The HOWTOs I have attempted to use have not been very well written, some have been horribly out of date, and for the most part, I was left (after an hour or more) with no solution to the problem they were supposed to help me resolve.
Now before the flaming begins, I freely grant that Windows docs suck. That is irrelevant. the subject is Linux, and Linux docs suck, too. The difference between Linux and Windows at installation time is that Windows (mostly) does a better job of handling the routine setup issues.
I've used Windows since 2.0 (which only stayed on my system for a few hours), and routinely install and manage NT installations. I'm getting better with Linux, but only after having invested nearly full time activity for several weeks, and am not by any means ready to commit my company to using Linux in product.
With all respect to everyone here, the proliferation of Linux will soon be limited by precisely these issues: documentation and setup.
Caldera has it mostly right, and I will soon install their 2.3, which will likely be still better. Let the others look to their model.
Documentation is my biggest issue. All the Linux users insist that it's anywhere from adequate to great, but those of us who didn't cut our teeth on *nix know it just ain't so. Until the Linux cognoscenti come out of denial on this, things won't improve.
My complaints are usually greeted with suggestions that I write some HOWTOs. Hello? I've already stated that I have problems in need of answers. How does that qualify me to write instruction for others? The people who need to write HOWTOs are the Linux gurus who insist there is no need. See the problem?
I live in hope, but I can only work with reality, so for now, Linux is only a dream.
And here I thought I was the only one who noticed. It's a shame (s)he didn't try 2.2 or 2.3, as the article would have had a completely different theme. I've tried RH, Suse, Mandrake, and Caldera, and Caldera has far and away the best installation package I've seen.
It's also a shame that there is no e-mail path given for the author. Sorta puts the article in the troll-bait class, doesn't it?
I would not choose firewire, either, but at least (unlike USB) it can make it to the next room. USB is as sorry a concept as I have seen. Did we really need a new technology for connecting keyboards and mice? I think not.
true, but most others can make it off my desk
on
Is firewire dying?
·
· Score: 1
The point is that USB is a toy, not a tool, and Firewire, on the other hand, has greater capability.
The reality is that as poor as USB is, it could be outdone by asynchronous communications over RS-485. Lest you feel too empowered by your prose, I would point out that I have been designing hardware for 30 years now, and USB is the most disappointing "technology" I have seen.
If MS has placed one or more backdoor entrances in Windows, whether for themselves or for NSA, can we really expect them to give a straight answer in public about that is clearly a matter of security?
Of course they deny it. If they acknowledged placing such keys, they would embarass themselves and the NSA (and would then have to concoct some new scheme for the future.)
By definition, public statements regarding security issues are suspect.
Firewire has (had?) much greater potential than the lame USB. I was at one time excited by USB, until I discovered how limiting it is in cable length. USB has potential only in home computing. Firewire could have become an adjunct to 100Mb Ethernet. Perhaps the shenanigans of Apple with respect to royalties have brought us to this point.
I suppose now we must hope for gigabit Ethernet to become cost-effective and pervasive. Perhaps then we can consign the wimpy USB to the dusty closets of history.
Yanno, when I was in school (many years ago), we were taught that whereas she had a specific and constant meaning, he was context dependent in its meaning, as it referred to either gender. Likewise with mankind, chairman, and so many others. I recall thinking (even in grade school) that it was unfair for women and girls to have their own pronouns, while we had to share.
PC language is a semantic nightmare which makes a disgusting (or impersonal) mess of an otherwise very capable language: English.
I'm quite willing to refrain from using any of the 7 words, but I'm not about to start typing he or she or s/he.
Heinlein said it best: Whenever women have insisted on absolute equality with men, they have invariably wound up with the dirty end of the stick. What they are and what they can do makes them superior to men, and their proper tactic is to demand special privileges, all the traffic will bear. They should never settle merely for equality. For women, 'equality' is a disaster.
My first response above was submitted without preview, and maybe it's just me, but shouldn't the confirmation have shown the message text, even though I did not preview? As it did not, I modified and resubmitted, with preview first....
Slashdot Mirror
I'm glad to see Borland reiterating their commitment to Linux. I have used Delphi since it came out, and Turbo Pascal before that. I have no problem with C (though C++ makes my skin crawl), but Delphi makes the construction of a UI a minor event, not the majority of the app. That's as it should be.
/. stem from fears on the part of the current community of Linux developers that they will soon be drowned in a sea of Windows emigres. It is a well-founded concern. Many of us developing for Windows would like to be developing for Linux, but as commercial developers, cannot simply drop everything to adopt a new OS, new GUI interface, new compiler, and to build a new tool set.
I wonder whether many of the most negative comments on
Contrary to the ravings of RMS, I don't find anything wrong with being paid for my work. The single most offensive aspect of the Open Source movement is the tendency of its adherents to froth at the mouth over the notion of any software which is not free. I pay for a distribution, because I prefer to do that, rather than to download for the many hours it takes to obtain one for "free".
As a resident of North America, I am steeped in the principles of a reasonably free market, and in the notion that profit for labor and risk is an honorable thing. OSS folk take note: the market will determine whether or not Borland's commercially offered Linux-based tools survive. I believe that once Kylix hits the streets, we will see a tremendous upswing in the use of Linux, and in the availability of applications for the desktop.
Linux socialists: brace yourselves, the deluge is near.
With the success of commercial tools, as well as free ones, we will see the maturity of Linux in the market, and then MS will truly have something to fear.
WP peaked at 5.1 for DOS, which was an unmitigated piece of crap with a horribly inconsistent and byzantine set of function keys. The show codes function which I always heard praised was there because they never figured out how to clean up after themselves.
WP was, at least through the 6.1 version for Windows, a horrible, bug-ridden disaster which I used only under extreme duress.
Not that Word is the holy grail, either. Actually, the single most productive word processing tool I have ever used was Borland Sprint.
As so often happens, Linux folk seem to be unaware of the realities of non-Linux software and software companies.
1. Inprise, while no M$, is certainly not dying.
2. Interbase is, in fact, in widespread use in enterprise level and embedded applications.
3. The Java IDE may or may not be inferior to that of IBM, but IBM has an abysmal performance record at marketing good software, or of remaining committed to it. Remember OS2?
No, you don't get it. Real commercial software products are not built by macho C programmers who refuse to use a RAD tool, but by programmers who select tools for the productivity they provide, and for the ease of production of solid programs.
Visual environments, for the most part, make the programming of the UI a task so formidable that it dwarfs tha actual application code in many cases. Delphi, C++ Builder, and JBuilder all put the UI development where it belongs: in a simple and quick IDE which lets the programmer focus on coding application code, not UI code.
My only concern with this deal is that Corel may lessen the commitment to continued support for the Windows development language tools. You do remember Windows, don't you? It's that OS which many of us pragmatists do hate, but which still dominates the desktop market.
As to the language issue, he (and I) may find C as ugly as you apparently find Pascal. And C++ bears all the signs of committee work.
Leaving aside personal opinions and aesthetics, however, Pascal is a more straightforward language to parse than C or C++, and is relatively pure on the Borland branch. C would be better than C++, as it would avoid tremendous complexities, but another issue lurks: C, C++, and Java all muddy the same syntactic pool. An approach based on Pascal would not represent yet another pollution of C-flavored syntax.
It's beyond sad when supposedly intelligent operators of a technical site are incapable of a) spelling Uruguay, or b)looking it up, or c) differentiating between the name of the country (Uruguay) and the description of one of its residents (Uruguayan).
Come on guys, I know that America is the most ethnocentic nation on the planet (I grew up here, after all) but even so...
Looking at the rampant negative comments from those who are a) clearly ignorant of the Interbase product, and b) seem to think they have a God-given right to endless free software, it amazes me why any publisher would want to jump into this miasma.
Since the most vocal of you have made it clear, again and again, that you will never buy software, your comments have little relevance to any commercial operation.
What amazes me even more, however, is the ignorance of K&R. Is there no one here but me who is over the age of 25? Code written to K&R standards will be a) relatively straightforward, and b) free from the abominable practices prevalent in C++. Would you feel better if it were written to VC++ practice? Then it would truly be non-portable, inscrutable, and buggy.
Let your brains engage before your keyboards, people. And stop crapping on newcomers to the Open Source "movement." Without new victims, there can be no movement.
no growth = death
The zip code system was publicly funded, ergo the codes are public property. The reproduction of useful lists, however, as with the reproduction of Linux CDs may be commercialized.
The galling part is that I was once told by a USPS official that the codes are not public property, which is, of course, the usual problem with the government forgetting who works for whom.
Throwing a cap on free does not make it have a special meaning.
The zealotry issue arises in part because people foam at the mouth about ideology. When a public corporation invests time and money, and then offers the result without charge, it should be applauded, not disparaged.
Tossing negative comments at a free offering such as this is a great way to dry up the well. And how stupid might that be?
Besides, you may have noticed that the DoJ has some odd views on the subject of free.
Finally, JBF seems a good deal freer than RH.
Come on, people, let's be as intelligent as we think we are. If a company is run by a Mormon, does that make their software Mormon software? The US (mistakenly supposed to have been founded on principles of religious freedom -- it wasn't -- it was founded to let people practice their own flavor of persecution) has long been a hotbed of narrow and provincial thinking. The small town in which I was raised (in Michigan, not the south) was, until the 30s, restricted (no blacks, Jews, or Catholics, thank you.) It only changed because the law made it do so.
There are good and bad people in every company, country, and religion. Many religions look pretty cockeyed to one not raised within them. All religions look cockeyed to an atheist. Focus on purpose, not packaging.
Supposedly we hate M$ for their bloated and buggy software, not for their money. Yet the current DOJ action is money-motivated, not anything to do with either merchantibility or fitness of purpose, either of which might provide clear grounds for action against them.
I have German ancestry. Having said that, I can also say that my German grandfather was the most narrow-minded, hidebound, and opinionated man I knew while I was growing up. Does that make all Germans the same? Of course not.
On the matter of religion, I believe in a Creator. I do not believe in organized religion. I dislike Churches for the same reason I dislike M$ and governments: large organizations typically give rise to petty tyrants and corruption. In smaller organizations they have nowhere to hide while amassing their power.
Software is written by people, not by churches. From what I know, Executive Software is a company motivated by commercial desires, as is M$, as is Be, Inc., as is Red Hat. If we do not care about the religious beliefs of Bob Young, or of Jean Louis Gasee, why then should we care about the religious beliefs of Craig Jensen?
A mind is a terrible thing to waste. Let's not waste ours on the ignorance of persecution and paranoia.
Doesn't anyone here recall what the recording industry was able to do to DAT technology? The film industry has been paranoid of recording technologies for a long time, and will do the same to DVD. VHS poses little threat, as the quality is so poor. DVD, on the other hand, has the potential (if the encoding is good) to provide a much cleaner signal for both sound and picture.
I anticipate that the film industry will crank up the lobbyists, and that our politicians, ever vigilant for the highest bidder, will do something to damage the value of this technology, as they did with DAT.
I can appreciate that the film industry has IP rights to protect, but when the protection of such rights is used to justify legislating against someone else's IP rights (as was effectively the case with DAT and the notch filter), then justice is not a component of the action, and we should all scream long and loud to our various pols in the hope that sanity may prevail this time.
History, however, offers little hope.
USB has a horrible weakness with regard to cable length. I won't give a number, because it isn't worth pulling out the spec, but when I looked closely at implementing some peripherals for USB, I found the actuality of the specs to be disappointing, at best.
Nothing of any size or complexity can be put together in USB without hubs, and as others have pointed out:
1. USB requires the presence of a PC
2. USB sucks CPU
If USB is the future of interconnect, we better start placing our hopes in systems which are self-contained.
Even considering that it is on msn.com, this article is overrun with propaganda.
- easier than BeOS??? NOT
- implies NTFS is not O/S specific
- quarterly service packs? then how is NT 4.0 only on SP5 by now?
- almost impossible to screw up on a single O/S machine? um... not in my experience
As one who has lived in Canada, and therefore has some experience with the round robin of union strikes which has long plagued Canada Post, I have to wonder whether this turn of events will provoke yet another strike.
For those who haven't had the pleasure, a postal strike is truly a delight -- no junk mail. On the other hand, paying bills is tough, and creditors expect you to do so in a timely fashion, without using the mail.
Thereby implying that not all the dependencies were met, which takes us back to the point. Your argument is bogus for proper RPMs.
Sounds nice, but I'm talking about RH 6.0 out of the box. If the RPMs aren't proper, then RH has screwed up.
I have installed RHL 6.0 in full, selective, or minamalist configurations, and I have not had a single problem with hangs at boot. Your information is old.
My experience is quite recent, with 6.0, and is at odds with your own. My information is not old, and as is so often the case in talking about Linux deficiencies, your response is arrogant and glib.
If this is not a comparative issue, then why do you keep comparing Linux to Windows? If you want to take Linux on its own, then fine. If you want to compare it to Windows, that is fine, too. But do not compare it to Windows, and then dodge the reply.
I'm not dodging, and my references to Windows were in response to comments in the message to which I was replying.
You are perfectly correct here. My point was that you were jumping up and down and pointing at Linux, when the problem lies with OEMs, not Linux itself. If you want to point fingers, point at your hardware vendor who is not supporting Linux. Linux is practically helpless without their support. I find it completely amazing that it has come as far as it has with practically no OEM support.
OEMs will not support until they perceive a demand. OEMs are, by definition, money-driven, so they cannot simply decide to develop drivers for Linux on a lark. In the beginning, there is never OEM support for a new OS. I won't blame the OEMs because they have no obligation to support any OS, in fact. For a new OS to build interest it is necessary for developers outside the OEM sphere to undertake driver development.
Blaming Linux is not the answer.
You misread me completely. I do not blame Linux; I blame weak and/or absent documentation. In particular I blame distro vendors, such as Red Hat, for failing to provide better docs.
Now there you may well have a legit complaint. I do not have data on how accurate distribution compatability lists are. I just know that I looked for good equipment that was praised as Linux friendly, and Linux worked perfectly.
I have a number of legitimate complaints, whether or not you happen to agree. I also have made sure that my hardware is on the list of supported devices for all of the OSes in which I have an interest. Except for the issue of monitor modes under all but Caldera, I have not found a problem with hardware support under Linux.
Let me state it as clearly as I can here and now: My only complaint with Linux is not with the software -- it is with the altogether inadequate documentation.
I want an alternative to Windows. I would like very much to make Linux my choice, but I can't do so while the documentation leaves me scrambling for answers to problems everywhere I turn.
The first topic on which I looked to the HOWTOs for support was programming a serial port. The document is outdated, and was written by someone who freely admits he doesn't fully understand the issues. It raised more questions for me than it answered.
My sole interest in Linux is as a tool, a platform. I have zero interest in the politics or religion of Linux or Open Source. If the tool gets in my way, I'm not happy. Windows gets in my way every day. That's why I'm driven to seek an alternative.
Too often I find responses here from people who have already achieved a level of comfort and expertise with Linux and who deny the existence of problems. The problems are real. I have so far evaluated four different distros, and each has shortcomings. If I am to commit to Linux as a platform, I need to evaluate on the basis of lost productivity during the learning period. To date that estimate makes the price too high.
RPM may handle dependencies, but that doesn't make things work. Install everything, and the boot goes well. Be selective, and things hang the boot for long periods. I clocked one particular fail-over at over 8 minutes. This is NOT what a newbie needs. In-depth problem solving is an acquired skill. When you're starting, you need something which works, with a minimum of tweaking.
The LDP needs coherent management. It currently suffers from what appears from the outside to be do-nothing management. Linux is good, but it could be great.
Your opinions on documentation are your own. Mine likewise do not yield to your own. We agree to disagree. This is not a comparative issue. Windows docs suck, too. But that does not make the Linux docs acceptable.
Whether the support is from OEMs or MS, the result remains: Windows installation mostly handles adapters better. When installation is unsuccessful, no one cares whose fault it is. Linux wears the black eye, or the distro does. Not the OEM. That's life; that's reality.
I never buy computers pre-assembled or installed. I configure my own, and have for many years. Windows is easier to install. That's based on many installations. It's not good, and I don't like it, and yes, I have a battery of tricks I have learned painfully, but even so, without resorting to those tricks, it's still easier.
I use only components which are well supported. I have pre-checked the claims of the distros, and all of my components and peripherals are claimed to be supported. Of course, that doesn't speak to how well supported they may be.
I have installed (successfully) on my machines: Win95, Win98, WinNT, Caldera, Mandrake, Red Hat 6.0, and BeOS. I multiple boot, and I use the boot manager which comes with BeOS (simpler and more sensible than most, and trivial to install and set up.
My preference? If allthings were equal (and they never are) I prefer BeOS. It installs in just over 5 minutes, and has more of what I want than any of the others, and good docs, besides. (I include 3rd party docs as acceptable alternatives, in all cases.)
Let's settle for inexperienced but intelligent newbies. As to your math and science majors, perhaps they, like so many others, fail to RTFM.
I've been working with small computers since 1975, and have written in C, C++, Pascal, assembler, and even Forth. I've designed embedded processor systems and designed and written serial protocols.
In spite of my experience, I have suffered tremendous frustration with Linux. Also, as one who writes user docs, I do RTFM. The source of the problems is as follows:
1) Linux docs are, to be kind, less than wonderful. They make too many assumptions about the background of the reader. I have yet to find a good presentation of disk partitioning strategies, for example.
2) Distros which use the RPM installation offer only one safe option: Install Everything. Selective package installation leads to things which don't work, boots which stall for many minutes, and a grand variety of other mysteries.
3) Red Hat, in particular, has an installation process which is easily fouled by the user making an unexpected response. Why would the user do that? Because the instructions and/or prompts are inadequate.
4) X server installation and setup is a very interesting source of problems. My video cards and monitors are capable of 1600x1200, but Caldera is the only distro to have correctly set up that configuration on my machines. And talk about user-hostile: tweaking in Hz, KHz, and microseconds is definitely not for newbies. I'm a hardware designer with a lot of experience with deflection systems, and I still don't want to go there.
5) Sending a newbie to the HOWTOs is a great way to send someone postal. The HOWTOs I have attempted to use have not been very well written, some have been horribly out of date, and for the most part, I was left (after an hour or more) with no solution to the problem they were supposed to help me resolve.
Now before the flaming begins, I freely grant that Windows docs suck. That is irrelevant. the subject is Linux, and Linux docs suck, too. The difference between Linux and Windows at installation time is that Windows (mostly) does a better job of handling the routine setup issues.
I've used Windows since 2.0 (which only stayed on my system for a few hours), and routinely install and manage NT installations. I'm getting better with Linux, but only after having invested nearly full time activity for several weeks, and am not by any means ready to commit my company to using Linux in product.
With all respect to everyone here, the proliferation of Linux will soon be limited by precisely these issues: documentation and setup.
Caldera has it mostly right, and I will soon install their 2.3, which will likely be still better. Let the others look to their model.
Documentation is my biggest issue. All the Linux users insist that it's anywhere from adequate to great, but those of us who didn't cut our teeth on *nix know it just ain't so. Until the Linux cognoscenti come out of denial on this, things won't improve.
My complaints are usually greeted with suggestions that I write some HOWTOs. Hello? I've already stated that I have problems in need of answers. How does that qualify me to write instruction for others? The people who need to write HOWTOs are the Linux gurus who insist there is no need. See the problem?
I live in hope, but I can only work with reality, so for now, Linux is only a dream.
And here I thought I was the only one who noticed. It's a shame (s)he didn't try 2.2 or 2.3, as the article would have had a completely different theme. I've tried RH, Suse, Mandrake, and Caldera, and Caldera has far and away the best installation package I've seen.
It's also a shame that there is no e-mail path given for the author. Sorta puts the article in the troll-bait class, doesn't it?
Better still, the TCP/IP availability makes possible a whole range of networked peripherals. Who needs USB?
I would not choose firewire, either, but at least (unlike USB) it can make it to the next room. USB is as sorry a concept as I have seen. Did we really need a new technology for connecting keyboards and mice? I think not.
The point is that USB is a toy, not a tool, and Firewire, on the other hand, has greater capability.
The reality is that as poor as USB is, it could be outdone by asynchronous communications over RS-485. Lest you feel too empowered by your prose, I would point out that I have been designing hardware for 30 years now, and USB is the most disappointing "technology" I have seen.
If MS has placed one or more backdoor entrances in Windows, whether for themselves or for NSA, can we really expect them to give a straight answer in public about that is clearly a matter of security?
Of course they deny it. If they acknowledged placing such keys, they would embarass themselves and the NSA (and would then have to concoct some new scheme for the future.)
By definition, public statements regarding security issues are suspect.
Firewire has (had?) much greater potential than the lame USB. I was at one time excited by USB, until I discovered how limiting it is in cable length. USB has potential only in home computing. Firewire could have become an adjunct to 100Mb Ethernet. Perhaps the shenanigans of Apple with respect to royalties have brought us to this point.
I suppose now we must hope for gigabit Ethernet to become cost-effective and pervasive. Perhaps then we can consign the wimpy USB to the dusty closets of history.
Yanno, when I was in school (many years ago), we were taught that whereas she had a specific and constant meaning, he was context dependent in its meaning, as it referred to either gender. Likewise with mankind, chairman, and so many others. I recall thinking (even in grade school) that it was unfair for women and girls to have their own pronouns, while we had to share.
PC language is a semantic nightmare which makes a disgusting (or impersonal) mess of an otherwise very capable language: English.
I'm quite willing to refrain from using any of the 7 words, but I'm not about to start typing he or she or s/he.
Heinlein said it best: Whenever women have insisted on absolute equality with men, they have invariably wound up with the dirty end of the stick. What they are and what they can do makes them superior to men, and their proper tactic is to demand special privileges, all the traffic
will bear. They should never settle merely for equality. For women, 'equality' is a disaster.
My first response above was submitted without preview, and maybe it's just me, but shouldn't the confirmation have shown the message text, even though I did not preview? As it did not, I modified and resubmitted, with preview first....
The response was getting pretty horrible on the old server. The new one is much nicer.
/. is one of my primary references, so anything you can do to improve performance is greatly appreciated.