SmartMedia, like CompactFlash, looks like a DOS-formatted disk drive when plugged into a PC with an appropriate adapter. I use a Minolta PCMCIA-to-SCSI adapter, and a SmartMedia-to-PCMCIA adapter. To transfer images under Linux, I just mount the SmartMedia card like any other DOS file system.
I think SmartMedia is actually a better design than CompactFlash. CompactFlash has pin-based connectors and carries a lot of electronics. SmartMedia cards have flat gold connectors (like smart cards) that don't get bent and don't trap dirt and are very easy to insert/remove. Te cards contain essentially just the storage and no extra electronics. SmartMedia cards are very thin and somewhat flexible. While individual SmartMedia cards are always a little behind CompactFlash in terms of capacity, their storage density seems higher, and it's easy to carry a whole bunch of them.
While both SmartMedia and CompactFlash are usable, I actually prefer cameras that use SmartMedia.
Microsoft seems to be adopting a strategy whereby they invest in any company that has some influence on standards.
In the case of Borland, they probably want support for their Java "extensions" in JBuilder, plus more commitment by Borland to COM and MFC.
But despite some forays onto other platforms, Borland seemed largely a Windows company anyway. It's a shame because some of their products would have been ideal for a cross-platform strategy.
I wonder whether these "strategic investments" shouldn't be curtailed. While an investment does not mean the same thing as full ownership, it does guarantee a "seat at the table" and significant influence. It may also be easier to get past antitrust regulators for now.
Microsoft has sufficiently deep pockets to make those kinds of investments in just about any company that matters, and that bodes ill for any kind of real competition.
I suspect actually that piracy enforcement and upgrade policies of software companies are simply driven by the need for differential pricing.
A company like Microsoft wants to get their products onto as many desktops as possible and to extract as much money from any customer as possible. So, at the one end, they have full price versions that just install, then they have "upgrades" with all sorts of cumbersome legal or software requirements (home users may bother, corporate users often don't), and they have "low cost" versions for universities ("hey, buddy, the first one is on me" comes to mind).
Tolerating some degree of piracy among people who couldn't otherwise afford their software would actually make sense for them. Of course, piracy needs to be curtailed among people who could otherwise pay for their products.
Another common example of "piracy" involving Microsoft products is MSDN. Even though the $2500/year MSDN subscription includes Office and a lot of other applications, you are not (in principle) permitted to use that software for anything other than development. If you actually want to write, you need to shell out more money. Of course, many (most?) developers who couldn't afford separate full versions seem to ignore this requirement, while big corporations dutifully license Office and all the other software.
I very much hope software companies will crack down more on piracy. That will make the true cost of their products much clearer to people who right now are getting the impression that something like Windows is "cheap".
I stick religiously to software licensing terms, and I personally simply avoid most of this unpleasant business by just using free software whenever I can. I recommend you do the same...
There are really two questions here. First, do we even want to bring Windows-style computing to Linux. Second, does it make a difference whether that is tied to one Linux distribution or many.
Bringing Windows-style computing to Linux has risks. Most people who use Linux right now probably use it because they like the way it works, not because they don't have access to Windows. But what if people start writing more and more "free" Linux software that relies on some components on the CodeWarrier IDE? What if people stop developing egcs because CodeWarrior Professional is cheap enough and most users are happy enough with it? Commercial software may displace some free software, and in the process, some of the alternative computing paradigms Linux represents might disappear.
Also, the result of bringing Windows-style computing to Linux may well be that Linux gets a lot of Windows-style users. Is that good? On the one hand, it means that a large number of users finally work on a Linux-based platform, a platform with decent APIs (POSIX, CORBA, etc.). On the other hand, it likely means that the ratio of users to open source developers gets even more skewed, and that will push the free community support models to their limits.
The second question of whether it matters that this is tied to a particular distribution doesn't seem that serious to me. In fact, I would find it kind of useful if RedHat became the Linux for ex-Windows-users and, say, Debian became the choice for the traditional Linux community. That would make it much easier to distinguish users when it comes to bug reports. RedHat may become the AOL of Linux distributions, and that's just fine with me.
The whole debate centers around the "rights to privacy of law abiding US citizens". It seems to me that if the US wants to promote democracy around the world, it should start respecting and protecting the privacy rights of citizens of democracies everywhere. That, however, is not a concern debated much by the politicians involved in this debate.
If the US ever wants to be a true "world leader", the well-being and rights of citizens of other nations has to enter into such debates. Until then, even US allies can't shake the suspicion that US aspirations to leading other nations are simply a result of economic self-interest and a certain degree of paranoia about US security interests.
In fact, programs like this are a symptom of national paranoia in the US. Other countries know they cannot get complete intelligence information or military supremacy (nor would they want to commit the economic resources for doing so), and they cannot have an absolute military advantage. Except for the US, other nations need to learn to develop trust and negotiate. Only the US is trying to maintain an intelligence and military advantage that is absolute, even in the absence of an identifiable foe.
Fortunately, this one has an easy solution: any country concerned about privacy rights of their citizens can simply unilaterally make/keep strong encryption legal. The suspicion that US intelligence agencies use their information for industrial espionage (whether true or not) should provide strong enough incentives. And the argument that a country needs to keep strong encryption illegal so that US agencies can spy on their citizens I suspect won't go over well in even the most friendly allies of the US.
That's the kind of logo I would expect for a commercial startup or 90's big company spinoff (in fact, if pressed, I would have guessed that the debian-log-with-vase is the Magic Cap logo).
Is a more corporate image where Debian wants to go? I also noted some rather corporate looking press releases. I like the old Debian logo because it is friendly, personal, and looked pretty non-commercial to me--a good match to Debian's goals I thought.
Perl is a general-purpose programming language whose process, file, and text manipulation facilities have made it the programming language of choice for tasks involving quick prototyping, system utilities, software tools, system management tasks, database access, graphical programming, and world wide web programming--just to name a few.
I don't think there is any such thing as a "general purpose programming language".
All programming languages involve tradeoffs. Some are easy to implement, some have lots of features, some can be compiled into very high performance code, some are backwards compatible with others, some are easy to understand for a particular community, some catch a lot of errors at compile time, some catch a lot of errors at runtime, some use lots of resources, some are well-suited to development projects with lots of developers, some are well suited to development projects with only a single developer, etc.
Somewhere in that space, Perl, Python, Java, Tcl, Eiffel, Fortran, Lisp, Scheme, C, C++, and all those other languages each have their place. If you pick the best language for each job, it is my experience that there is actually very little overlap between those different languages. Each of them has their communities, and few of them are in danger of going away because some other language supposedly superceded them.
My comment above may come across as too critical, so just for clarification. What I was saying is that I think Perl has some quirks and I don't like to see those elevated to language design principles.
But having said that, I still think among scripting languages, Perl is one of the most useful, and it gets a lot of things right, and I certainly appreciate all the effort that has gone into it.
They are adding fork(), improving performance, and adding internationalization. All of those are good. I'd be much more concerned if they added Win32:: to the UNIX version.
I doubt they'll keep the source proprietary--that would seem to be self defeating in this case.
Python doesn't force you to indent any particular amount, and it doesn't force you to indent the same amount everywhere. All it requires is that your indentation structure reflects your block structure. It appears that most people pay more attention to indentation than to syntax (consider the classic nested if/else-with-incorrect-indentation example), so arguably this is a good thing.
I have used Perl nearly since its earliest public releases and have written a lot of code in it. I still use it as my main scripting language. I've written database management tools, pattern recognition tools, experimental design tools, and lots of CGI stuff in it (including a bug tracker).
Perl is a very useful tool, and when it came out it was a big improvement over awk/sed/sh, and it was free to boot. That's why it succeeded. Nowadays, most people simply use it because it's there (among other things, it's the only CGI language my ISP offers).
But far from being an advantage, Perl's haphazard error checking and "postmodern" syntax do cause problems in practice. For example, one of the big government projects Larry mentions lost a lot of data because Perl did not flag bad numerical input as an error but simply uses 0 by default (a design flaw that has since been partially corrected through addition of "-w"). And even the Perl system itself has problems with its syntax--there are many circumstances where Perl compiler or runtime error messages are way off.
The lesson I hope budding language designers will take away from Perl is that providing useful functionality in a timely manner and a free package is more important for success than clean design or robustness. But within the constraints of time to market, one should still strive for clean, robust designs, or one should at least aim to fix things up later. To Perl's credit, many of the initial design problems have been fixed.
And while Larry seems to think languages like Lisp and Python were designed by some CS types for thought control, reality is that, for example, Lisp evolved over nearly 40 years to meet the needs of its user community (and, in a twist of irony, Perl's syntax itself has followed a little bit of the same trajectory already; let's see where it ends up in another 30 years).
To me, Perl succeeded despite its quirks and problems, not because of them.
My "56k" modem consistently connects at 49k, which is pretty close to the 53k maximum that the FCC actually allows. And file transfer is proportionately faster compared to the 33k modems.
You have to be careful with how you initialize the modem. The initialization string I was using for a few months actually somehow disabled high speed connections. When I trimmed it down, things started working. And there are locations from where you can't connect--too noisy or too far from the switch. But my phone wires aren't perfect (water damage), so there seems to be at least some robustness.
My 56k modem came with my PC. I was quite sceptical about the 56k stuff and probably wouldn't have bought it, but I'm happy with it now. It will still be months until DSL becomes available, and ISDN is much more expensive.
RedHat 6 and Netscape 4.5/4.6 problems (SOLUTION)
on
Mozilla M6 released
·
· Score: 2
The problem with Netscape 4.5 and 4.6 under RedHat is due to missing fonts interacting with poor error checking in the Java implementation. By adding the non-scaled 75dpi fonts to your server's font path, you can fix the crashes. This is described in detail on RedHat's support site.
Netscape also has (and has had) a problem with lots of dialog boxes popping up about bad widget sizes. You can't get rid of the messages, but you can send them to stderr instead where they do no harm. Check DejaNews for how to configure that.
With those two fixes applied, 4.5/4.6 seems as good or better than 4.08.
It's called GNU Emacs. I know--you may not like it, but, then, the GNU people also usually don't like Windows-style IDEs. For example, if I wanted to use VC++, I'd just use it--I have a free MSDN subscription. But I actually uninstalled most of that stuff months ago.
I'm not convinced that it is entirely a good thing if Windows-style IDEs come to GNU or Linux. If Linux starts looking and behaving simply like a more robust version of Windows, where are all the UNIX people supposed to go?
Many Java IDEs aren't written in Java because vendors like Borland and IBM already had IDEs written in other languages that they could build a Java environment on. Also, they had existing user bases that already had a lot of experience with their IDEs, so from that point of view it also made sense to stick with the existing codebase for the IDE.
I tried JBuilder, VisualAge, Symantec Cafe, and Visual J++ and thought all of them were pretty awful, about as awful as the C++ and/or Smalltalk IDEs they were derived from. But, then, I'm sticking with the IDE I have been using for many years: Emacs, and it happens to be written in C and Lisp.
Pure Java IDEs are beginning to appear, however, and I think you are going to see more and more people moving in that direction over the next few years.
I never thought Microsoft could be kept from making a clean-room implementation in the long run.
What is actually much more important to me is whether Microsoft can call their implementation "Java". I believe the plethora of security problems, portability problems, and bugs Microsoft has introduced into their product and shipped under the trademarked name of "Java" are a classic case of trademark dilution.
As someone who has to "sell" (at least to management) Java projects, that matters a lot to me.
I hope Sun hasn't give up the right to limit Microsoft's use of the "Java" trademark as part of their contract, and I also hope Sun will start enforcing the "Java" trademark against Microsoft.
Who cares whether Microsoft ships a language called "J++" or anything else. The harm is done if Microsoft calls it "Java".
Microsoft signed a contract with Sun to obtain access to Sun's unpublished sources and to obtain the right to distribute products derived from Sun's sources. As part of that contract, they could have easily given away their right to ever write a cleanroom implementation of Java.
As far as I can tell, Microsoft's winning or losing this case has nothing to do with your rights or my rights to write software under the law. It has little to do with the right to reverse engineer or to make clean-room implementations. It mostly has to do with how smart Microsoft's lawyers were in their contract negotiations with Sun.
On balance, it's probably good that this source is becoming available (although the free Flash player for Linux works quite well already).
But the motivation for Macromedia is clear: the W3 is developing XML-based graphics standards and executable content in the form of Java (and ActiveX, if you must) also provides better and better graphics (Java 2 has antialiasing, better fonts, and a much better imaging model than earlier versions).
Macromedia has a big lead and advantage in authoring and server tools for Flash, and the longer they can keep Flash on the web, the more money they will make. When other formats will become more widespread, they will only be one among many vendors, with no particular lead or advantage.
I hope the release of Flash source won't preempt the adoption of the next generation of graphics formats. The XML-based formats are much easier to use for dynamic generation of graphical content from any scripting language (no need to buy expensive tools from Macromedia), and Java 2, once it is in browsers, will allow much better interactivity.
Free software is every bit subject to patent law as commercial products. Traditionally, distribution of sources would likely be considered "contributory infringement", although if the (sensible) notion prevails that software is protected speech, that might be a way out. However, it would still infringe to compile and execute that software.
There is also no exemption from patent law for research. A competitor can keep you from conducting research if it involves using their patents. This is a particular problem in the biotech industry, where patents on genes and gene products limit the kind of research other institutions can conduct.
Those patents can also be used by private companies to force publically researchers to agree to restrictive terms when they want to conduct publically funded research involving genes or products that are patented.
All of this is very recent, and I don't think the impact even has begun to make itself felt.
I have set up Linux machines to reboot in under ten seconds from past the BIOS check to login prompt.
Most importantly, the runlevel stuff (SysV initialization) is just abysmally slow. If you want fast reboots, put just what you need into/etc/rc and get rid of all the other stuff.
It also helps to make a custom kernel that includes only the modules and drivers you need.
If you have certain kinds of hardware or controllers in your system (Adaptec comes to mind), trying to achieve fast reboots is hopeless because they need lengthy initialization routines and microcote downloads.
This kind of setup isn't good for many day-to-day uses, and it will probably cause problems with distributions like RedHat or Caldera. But if you want fast reboots, you can get them. Last I did this was on an older distribution of Slackware.
One grass-roots solution to the patent problem vs. free software could be a kind of "Free Patent Foundation".
There are two things that can be usefully done:
Provide a forum where individuals can disclose ideas (possibly anonymously) that can serve as a repository of prior art and protect from future patent claims. Such a repository is pretty easy to set up (think of it as something like Slashdot, with each entry being a "disclosure" of an idea). Disclosure on USENET would be a possibility, of course, but USENET has become too messy to serve as an easily searchable repository of disclosures or prior art.
Actually make a volunteer effort to generate patents on some "donated" ideas and use them in bargaining with other companies to allow free use of their portfolios in free software.
The reason why so many obvious things get patented is because there is every economic incentive to doing so. In fact, in order to survive, any company (in particular small companies) need to be a patent portfolio for bargaining. Many of the people who submit such patents don't believe in their non-obviousness themselves.
When it comes to enforcing such patents, technical content makes little difference. What matters is mainly a tradeoff between management time spend on dealing with litigation vs. licensing cost. As long as the licensing cost is kept below the cost of management time, the patent will get licensed, valid or not.
This is an economic problem: patents are enormously valuable and impose risk on others; their cost compared to their value is tiny (although the cost of entry for making any money off them is fairly high, since they need to be prepared carefully and be backed up by a believable legal team in order to actually get licensing revenue).
And the solution, I believe, has to be economic, too: there should be real costs and risks to the patent holder if a patent is later invalidated because of obviousness or prior art. For example, people who paid licensing fees should be able to have them returned, legal fees should be reimbursed, companies who lost revenue because their management was kept busy through legal tactics should be able to recover expenses and damages, etc.
Changing the patent office itself won't help, and neither will increasing patent fees (that will only keep out the small inventor). The beauty of making patent holders liable if their patents are invalidated is that it is self regulating and puts the burden of careful evaluation on the patent submitter, where it should be.
I think it's great that Slackware keeps on going. Their install disks are very useful with any Linux distribution. And their low-level approach to installation works when others fail.
I have been using RedHat for a few releases, but with RedHat 6.0 I have had real problems. The PCMCIA install seems to have serious bugs, and for them to use glibc 2.1 seems to have been premature (it breaks Java and Netscape, among other things).
I'll probably install Slackware again on my other laptop; the combination of 2.2 kernel and pre-2.1 glibc could be quite good, and their install procedure has a better chance of working.
I hope that the market is big enough that three or four distributions with such different characteristics will be able to co-exist. It would be sad indeed if Linux only came as RedHat or Caldera. More hands-on distributions like Slackware and Debian are needed.
I'm sure this worries software companies, but in reality, it's hard for me to see what schools need beyond free software, bundled software, and Internet access.
In the past, there were big hopes for computer aided instruction (CAI), but those don't really seem to have come through.
What computers seem to be most useful for is to learn computer related things (programming, data analysis, etc.), as office tools (text processing, graphics, etc.), and for accessing the Internet. For programming and data analysis, free software is probably the best there is (why warp little minds with VC?). For office tools, most computers ship with some form of office suite anyway; if they don't, that's probably the cheapest purchase per student (unless it's MS Office, of course). And Internet access software is also included.
It's a simple fact: there isn't much need to spend much money on desktop software anymore. I'm religious about not pirating software, and still pretty much the only money I spend on software is on the occasional $30 game.
1600x1200 color LCD screens already essentially have paper-like resolution if you look at b/w images (you get about 1600x3600 or 4800x1200). Actual resolution is perhaps still a little lower than paper, but the fact that the LCDs let you control the intensity of each pixel more than makes up for that.
Note that an 800x600 display with paper resolution is not "digital paper"; at best, it's a "digital postage stamp display". The technologies used for making the 800x600 displays can't scale up to paper-sized displays right now.
The first company to produce LCD displays with paper-like resolutions was probably Xerox's dpiX. But their initial products were very expensive, and other LCD technologies seem to have caught up.
Of course, in addition to LCDs with paper-like resolutions, there is true "digital paper"--stuff where the display persists when the power is removed and that can be rolled (though not folded).
Slashdot Mirror
I think SmartMedia is actually a better design than CompactFlash. CompactFlash has pin-based connectors and carries a lot of electronics. SmartMedia cards have flat gold connectors (like smart cards) that don't get bent and don't trap dirt and are very easy to insert/remove. Te cards contain essentially just the storage and no extra electronics. SmartMedia cards are very thin and somewhat flexible. While individual SmartMedia cards are always a little behind CompactFlash in terms of capacity, their storage density seems higher, and it's easy to carry a whole bunch of them.
While both SmartMedia and CompactFlash are usable, I actually prefer cameras that use SmartMedia.
In the case of Borland, they probably want support for their Java "extensions" in JBuilder, plus more commitment by Borland to COM and MFC.
But despite some forays onto other platforms, Borland seemed largely a Windows company anyway. It's a shame because some of their products would have been ideal for a cross-platform strategy.
I wonder whether these "strategic investments" shouldn't be curtailed. While an investment does not mean the same thing as full ownership, it does guarantee a "seat at the table" and significant influence. It may also be easier to get past antitrust regulators for now.
Microsoft has sufficiently deep pockets to make those kinds of investments in just about any company that matters, and that bodes ill for any kind of real competition.
A company like Microsoft wants to get their products onto as many desktops as possible and to extract as much money from any customer as possible. So, at the one end, they have full price versions that just install, then they have "upgrades" with all sorts of cumbersome legal or software requirements (home users may bother, corporate users often don't), and they have "low cost" versions for universities ("hey, buddy, the first one is on me" comes to mind).
Tolerating some degree of piracy among people who couldn't otherwise afford their software would actually make sense for them. Of course, piracy needs to be curtailed among people who could otherwise pay for their products.
Another common example of "piracy" involving Microsoft products is MSDN. Even though the $2500/year MSDN subscription includes Office and a lot of other applications, you are not (in principle) permitted to use that software for anything other than development. If you actually want to write, you need to shell out more money. Of course, many (most?) developers who couldn't afford separate full versions seem to ignore this requirement, while big corporations dutifully license Office and all the other software.
I very much hope software companies will crack down more on piracy. That will make the true cost of their products much clearer to people who right now are getting the impression that something like Windows is "cheap".
I stick religiously to software licensing terms, and I personally simply avoid most of this unpleasant business by just using free software whenever I can. I recommend you do the same...
Bringing Windows-style computing to Linux has risks. Most people who use Linux right now probably use it because they like the way it works, not because they don't have access to Windows. But what if people start writing more and more "free" Linux software that relies on some components on the CodeWarrier IDE? What if people stop developing egcs because CodeWarrior Professional is cheap enough and most users are happy enough with it? Commercial software may displace some free software, and in the process, some of the alternative computing paradigms Linux represents might disappear.
Also, the result of bringing Windows-style computing to Linux may well be that Linux gets a lot of Windows-style users. Is that good? On the one hand, it means that a large number of users finally work on a Linux-based platform, a platform with decent APIs (POSIX, CORBA, etc.). On the other hand, it likely means that the ratio of users to open source developers gets even more skewed, and that will push the free community support models to their limits.
The second question of whether it matters that this is tied to a particular distribution doesn't seem that serious to me. In fact, I would find it kind of useful if RedHat became the Linux for ex-Windows-users and, say, Debian became the choice for the traditional Linux community. That would make it much easier to distinguish users when it comes to bug reports. RedHat may become the AOL of Linux distributions, and that's just fine with me.
If the US ever wants to be a true "world leader", the well-being and rights of citizens of other nations has to enter into such debates. Until then, even US allies can't shake the suspicion that US aspirations to leading other nations are simply a result of economic self-interest and a certain degree of paranoia about US security interests.
In fact, programs like this are a symptom of national paranoia in the US. Other countries know they cannot get complete intelligence information or military supremacy (nor would they want to commit the economic resources for doing so), and they cannot have an absolute military advantage. Except for the US, other nations need to learn to develop trust and negotiate. Only the US is trying to maintain an intelligence and military advantage that is absolute, even in the absence of an identifiable foe.
Fortunately, this one has an easy solution: any country concerned about privacy rights of their citizens can simply unilaterally make/keep strong encryption legal. The suspicion that US intelligence agencies use their information for industrial espionage (whether true or not) should provide strong enough incentives. And the argument that a country needs to keep strong encryption illegal so that US agencies can spy on their citizens I suspect won't go over well in even the most friendly allies of the US.
Is a more corporate image where Debian wants to go? I also noted some rather corporate looking press releases. I like the old Debian logo because it is friendly, personal, and looked pretty non-commercial to me--a good match to Debian's goals I thought.
I don't think there is any such thing as a "general purpose programming language".
All programming languages involve tradeoffs. Some are easy to implement, some have lots of features, some can be compiled into very high performance code, some are backwards compatible with others, some are easy to understand for a particular community, some catch a lot of errors at compile time, some catch a lot of errors at runtime, some use lots of resources, some are well-suited to development projects with lots of developers, some are well suited to development projects with only a single developer, etc.
Somewhere in that space, Perl, Python, Java, Tcl, Eiffel, Fortran, Lisp, Scheme, C, C++, and all those other languages each have their place. If you pick the best language for each job, it is my experience that there is actually very little overlap between those different languages. Each of them has their communities, and few of them are in danger of going away because some other language supposedly superceded them.
But having said that, I still think among scripting languages, Perl is one of the most useful, and it gets a lot of things right, and I certainly appreciate all the effort that has gone into it.
I doubt they'll keep the source proprietary--that would seem to be self defeating in this case.
Python doesn't force you to indent any particular amount, and it doesn't force you to indent the same amount everywhere. All it requires is that your indentation structure reflects your block structure. It appears that most people pay more attention to indentation than to syntax (consider the classic nested if/else-with-incorrect-indentation example), so arguably this is a good thing.
Perl is a very useful tool, and when it came out it was a big improvement over awk/sed/sh, and it was free to boot. That's why it succeeded. Nowadays, most people simply use it because it's there (among other things, it's the only CGI language my ISP offers).
But far from being an advantage, Perl's haphazard error checking and "postmodern" syntax do cause problems in practice. For example, one of the big government projects Larry mentions lost a lot of data because Perl did not flag bad numerical input as an error but simply uses 0 by default (a design flaw that has since been partially corrected through addition of "-w"). And even the Perl system itself has problems with its syntax--there are many circumstances where Perl compiler or runtime error messages are way off.
The lesson I hope budding language designers will take away from Perl is that providing useful functionality in a timely manner and a free package is more important for success than clean design or robustness. But within the constraints of time to market, one should still strive for clean, robust designs, or one should at least aim to fix things up later. To Perl's credit, many of the initial design problems have been fixed.
And while Larry seems to think languages like Lisp and Python were designed by some CS types for thought control, reality is that, for example, Lisp evolved over nearly 40 years to meet the needs of its user community (and, in a twist of irony, Perl's syntax itself has followed a little bit of the same trajectory already; let's see where it ends up in another 30 years).
To me, Perl succeeded despite its quirks and problems, not because of them.
You have to be careful with how you initialize the modem. The initialization string I was using for a few months actually somehow disabled high speed connections. When I trimmed it down, things started working. And there are locations from where you can't connect--too noisy or too far from the switch. But my phone wires aren't perfect (water damage), so there seems to be at least some robustness.
My 56k modem came with my PC. I was quite sceptical about the 56k stuff and probably wouldn't have bought it, but I'm happy with it now. It will still be months until DSL becomes available, and ISDN is much more expensive.
Netscape also has (and has had) a problem with lots of dialog boxes popping up about bad widget sizes. You can't get rid of the messages, but you can send them to stderr instead where they do no harm. Check DejaNews for how to configure that.
With those two fixes applied, 4.5/4.6 seems as good or better than 4.08.
I'm not convinced that it is entirely a good thing if Windows-style IDEs come to GNU or Linux. If Linux starts looking and behaving simply like a more robust version of Windows, where are all the UNIX people supposed to go?
I tried JBuilder, VisualAge, Symantec Cafe, and Visual J++ and thought all of them were pretty awful, about as awful as the C++ and/or Smalltalk IDEs they were derived from. But, then, I'm sticking with the IDE I have been using for many years: Emacs, and it happens to be written in C and Lisp.
Pure Java IDEs are beginning to appear, however, and I think you are going to see more and more people moving in that direction over the next few years.
What is actually much more important to me is whether Microsoft can call their implementation "Java". I believe the plethora of security problems, portability problems, and bugs Microsoft has introduced into their product and shipped under the trademarked name of "Java" are a classic case of trademark dilution.
As someone who has to "sell" (at least to management) Java projects, that matters a lot to me.
I hope Sun hasn't give up the right to limit Microsoft's use of the "Java" trademark as part of their contract, and I also hope Sun will start enforcing the "Java" trademark against Microsoft.
Who cares whether Microsoft ships a language called "J++" or anything else. The harm is done if Microsoft calls it "Java".
As far as I can tell, Microsoft's winning or losing this case has nothing to do with your rights or my rights to write software under the law. It has little to do with the right to reverse engineer or to make clean-room implementations. It mostly has to do with how smart Microsoft's lawyers were in their contract negotiations with Sun.
But the motivation for Macromedia is clear: the W3 is developing XML-based graphics standards and executable content in the form of Java (and ActiveX, if you must) also provides better and better graphics (Java 2 has antialiasing, better fonts, and a much better imaging model than earlier versions).
Macromedia has a big lead and advantage in authoring and server tools for Flash, and the longer they can keep Flash on the web, the more money they will make. When other formats will become more widespread, they will only be one among many vendors, with no particular lead or advantage.
I hope the release of Flash source won't preempt the adoption of the next generation of graphics formats. The XML-based formats are much easier to use for dynamic generation of graphical content from any scripting language (no need to buy expensive tools from Macromedia), and Java 2, once it is in browsers, will allow much better interactivity.
There is also no exemption from patent law for research. A competitor can keep you from conducting research if it involves using their patents. This is a particular problem in the biotech industry, where patents on genes and gene products limit the kind of research other institutions can conduct.
Those patents can also be used by private companies to force publically researchers to agree to restrictive terms when they want to conduct publically funded research involving genes or products that are patented.
All of this is very recent, and I don't think the impact even has begun to make itself felt.
Most importantly, the runlevel stuff (SysV initialization) is just abysmally slow. If you want fast reboots, put just what you need into /etc/rc and get rid of all the other stuff.
It also helps to make a custom kernel that includes only the modules and drivers you need.
If you have certain kinds of hardware or controllers in your system (Adaptec comes to mind), trying to achieve fast reboots is hopeless because they need lengthy initialization routines and microcote downloads.
This kind of setup isn't good for many day-to-day uses, and it will probably cause problems with distributions like RedHat or Caldera. But if you want fast reboots, you can get them. Last I did this was on an older distribution of Slackware.
There are two things that can be usefully done:
When it comes to enforcing such patents, technical content makes little difference. What matters is mainly a tradeoff between management time spend on dealing with litigation vs. licensing cost. As long as the licensing cost is kept below the cost of management time, the patent will get licensed, valid or not.
This is an economic problem: patents are enormously valuable and impose risk on others; their cost compared to their value is tiny (although the cost of entry for making any money off them is fairly high, since they need to be prepared carefully and be backed up by a believable legal team in order to actually get licensing revenue).
And the solution, I believe, has to be economic, too: there should be real costs and risks to the patent holder if a patent is later invalidated because of obviousness or prior art. For example, people who paid licensing fees should be able to have them returned, legal fees should be reimbursed, companies who lost revenue because their management was kept busy through legal tactics should be able to recover expenses and damages, etc.
Changing the patent office itself won't help, and neither will increasing patent fees (that will only keep out the small inventor). The beauty of making patent holders liable if their patents are invalidated is that it is self regulating and puts the burden of careful evaluation on the patent submitter, where it should be.
I have been using RedHat for a few releases, but with RedHat 6.0 I have had real problems. The PCMCIA install seems to have serious bugs, and for them to use glibc 2.1 seems to have been premature (it breaks Java and Netscape, among other things).
I'll probably install Slackware again on my other laptop; the combination of 2.2 kernel and pre-2.1 glibc could be quite good, and their install procedure has a better chance of working.
I hope that the market is big enough that three or four distributions with such different characteristics will be able to co-exist. It would be sad indeed if Linux only came as RedHat or Caldera. More hands-on distributions like Slackware and Debian are needed.
In the past, there were big hopes for computer aided instruction (CAI), but those don't really seem to have come through.
What computers seem to be most useful for is to learn computer related things (programming, data analysis, etc.), as office tools (text processing, graphics, etc.), and for accessing the Internet. For programming and data analysis, free software is probably the best there is (why warp little minds with VC?). For office tools, most computers ship with some form of office suite anyway; if they don't, that's probably the cheapest purchase per student (unless it's MS Office, of course). And Internet access software is also included.
It's a simple fact: there isn't much need to spend much money on desktop software anymore. I'm religious about not pirating software, and still pretty much the only money I spend on software is on the occasional $30 game.
Note that an 800x600 display with paper resolution is not "digital paper"; at best, it's a "digital postage stamp display". The technologies used for making the 800x600 displays can't scale up to paper-sized displays right now.
The first company to produce LCD displays with paper-like resolutions was probably Xerox's dpiX. But their initial products were very expensive, and other LCD technologies seem to have caught up.
Of course, in addition to LCDs with paper-like resolutions, there is true "digital paper"--stuff where the display persists when the power is removed and that can be rolled (though not folded).