Microsoft sues the Samba developers, alleging both trade secret violations and violation of the DCMA. (This would make them felons...)
Remember that the Digital Millenium Copyright Act has force of law in only one of the over two hundred countries in the world, and, as it happens, not all the Samba developers live there.
However, suppose fifty people were to download it without agreeing to the licence (I just have, from http://warot.com/freedom/kerbspec.pdf), and each quoted just one paragraph on their web site (copyright law allows for the fair-use quoting of small parts of a document), and then suppose someone else later came along independently and made a Web page which linked to all those single paragraph Web pages, what law would have been broken?
As an example of what I mean, I have posted an example paragraph here
It's a sad day when Slashdot feels it has to explain who Vint is. Vint is more than the guy who wrote TCP/IP; he's also one of the people who has lead and driven the Internet all these years, and worked to keep it open for all of us. He's also a really nice person to meet. He's one of the ubergeeks of the Net, alongside Jon Postel; one of the real greats.
Fortunaltely for us all, though, Vint is still very much with us!
Oh, and HURD doesn't thunder. It moves very slowly towards a point which recedes infinitely towards the HURD release event horizon.
Re:what i dont understand, please enlighten me
on
Geographic Screening
·
· Score: 2
In this particular case, iCrave was violating US laws in US jurisdiction, because their (re)broadcasts were accessable in the US. In other words, they were breaking US laws in the US.
Bollocks.. ICrave, as I understand it, were doing absolutely nothing at all in US jurisdiction. None of their equipment was in US jurisdiction. They had a server in Canada which responded to requests, but only to requests from people who said they too were in Canada. If someone from the US requested material which was copyright in the US, then that was midly naughty. If in order to do so they dishonestly pretended not to be in the US, then that was naughty too. But in neither case was ICrave in the wrong.
It's been pointed out before in this thread that US jurisdiction does not and cannot cover things people do outside the United States borders. The inability of some American contributers to this thread to understand this very simple fact merely confirms for those of us who live outwith the United States our already low opinion of the average intellectual capacity of those who don't.
In other words, we're not surprised you don't understand. We would like to be surprised, but we aren't holding our breath.
I normally read/. at with my filters set to '2 and above' so as to avoid this sort of flame fest, but this one attracted my attention and I backtracked it.
Some observations: it's notable that the people (or possibly just one person) attacking RMS here doesn't choose to reveal [his|her] identity.
Secondly, there's some extremely sloppy thinking going on. For example, the assertion that just because you paid for RedHat, anything you develop under RedHat you can automatically sell on. You can't. The C compiler RedHat distribute is the Gnu C compiler; the standard libraries RedHat distribute are the GNU libraries. If you use those tools, and if you believe software licences have any validity at all, you must abide by their licences. RedHat cannot issue those tools to you under any different licence than the GPL, and don't try.
Then there is the assertion that RMS, the Free Software Foundation, and Debian want to stop you making money out of software. This is just simply obviously untrue. RMS makes his living out of software; so did the guys at Cygnus who have done so much of the maintenance of the GNU toolkit over the past several years. What RMS wants to persuade you not to do is make money out of hiding the source of software, restricting the freedom of users of software.
Finally, there is the assertion that Marxism is the same as socialism, and that both are the same as standing up for the right to freedom of speach. Anyone who can believe this is either so politically naiive, or so indoctrinated, that their other opinions are invalidated.
Face it: RMS, for all his querkiness and his occasional displays of ego, is necessarily one of our heroes. Without him, there would be no GNU toolset; without the GNU toolset, there could be no Linux. There's no sense in the sort of argument you see frequently between supporters of different free software/open source heroes. All of them - RMS, ESR, Linus, Alan Cox, even Larry Wall - have contributed hugely to making the movement we find ourselves in. All of them are human and have their querks, but we all owe them all a huge debt of gratitude, and this sort of anonymous attack is simply childish and undignified.
Country codes are very much more political than you think. The current list is simply drawn from ISO 3166, which lists territories of all sorts, some of which are independent countries and some of which are not. Many countries which feel they should be independent but are not (such as my country, Scotland) do not have ISO 3166 listings, and so don't have TLDs; indeed we have the anomalous situation where a small island I can see from the top of the hill behind my house has a TLD (IM, the Isle of Man), but my country doesn't. Furthermore, all the obvious two letter abbreviations of Scotland are already taken.
Centralist governments which are trying to resist nationalist movements (such as in Britain, Spain or Indonesia) strongly do not want to allow the aspirant nations which they control to have TLDs; independence movements within those nations strongly do.
This isn't an argument either to have country-code TLDs or not to have them, but it's worth being aware that there are very strong political undercurrents around this issue which will not make it easy to solve.
Finally, I suspect that governments particularly of smaller and less self-confident nations would strongly resist moves to do away with CCTLDs.
The thing that worries me most about this is an email I received this morning from my lawyers, which I'll quote:
------------------- begin quote -----------------
US SEIZES JURISDICTION OVER DOT.COM COMPANIES
Mere registration of top level domain sufficient
US companies able to seize worldwide registrations
US trademark owners able to have domain names of others expropriated
Over the last few years we have commented on Internet cases from around the world which we believe would be of interest to UK businesses and others. For some readers these cases may have been more relevant than others. Today however we report on a decision of the District Court for the Eastern District of Virginia which has perhaps the furthest reaching implications for worldwide internet governance.
In its Decision of Friday 5th March 2000, Caesars World, Inc -v- Caesars-Palace.Com and others (Civil Action No 99-550-A), Judge Albert Bryan effectively decreed that his court would be the arbiter of the property rights in respect of all the approx. 7,000,000.COM,.EDU, and.ORG top level domains registered. Although the Decision dealt solely with the motions of two of the defendants to dismiss an action for lack of jurisdiction, the effect of the Decision is to open the floodgate to litigation by holders of US trademarks against domain name proprietors based outwith the US (or at least outwith Virginia.
-------------------- end quote ------------------
Now, it may have escaped your attention in America, but the Internet is international; it has been international since Arpanet was linked to Janet in the early eighties. When we set up ICANN (I say 'we' - I was convenor of one of the Geneva sessions of the working party which led to the setting up of ICANN) it was precisely to prevent this sort of thing from happening.
The only possible consequence of this sort of nonsense is that the domain name service will collapse; courts and politicians in Europe (and, I imagine, in Japan) are not going to like being told that American trade-mark owners have first claim on an international resource. Either the Virginia court is persuaded to see sense, or we'd better get on with developing a replacement for the Domain Name System.
Yes, and I've only seen one ISP (UUNet) which actually uses that as their primary address. Many of the other big ISP's hold on to the.net TLD, but it's nothing more than a redirect to the.com address, which is by your definition another "ridiculous misuse of the namespace."
No, it's not, it's specifically what people were asked to do by IANA: to use.net addresses for all elements of the public network backbone, and.com addresses for everything else.
What, then, would you suggest Slashdot's URL be? "Slashdot.com" doesn't fit, because Slashdot isn't really a commercial venture (the ads notwithstanding). "Slashdot.net" doesn't work for the reasons you just said. "Slashdot.gov" and "slashdot.mil" are obvious problems as well.
It may have escaped your notice, but/. is a commercial venture of Andover.Net, inc. Both 'slashdot.org' (which isn't a non-profit) and 'andover.net' (which doesn't own any part of the network backbone) are gross abuses of the namespace.
I've been using talkback versions of Mozilla since about M8, I think - it doesn't always work. In my experience four out of five crashes go down so hard and so fast that talkback never gets a chance. Maving said that, M13 was in my opinion very nearly as stable as 4.7 - but having said *that* I've now ditched 4.7 in favour of 4.6, because I found 4.7 so fragile as to be more or less unusable.
Emacs (PSGML) is actually the best editor I know for XML files. It gives you context-sensitive pop-up menus of all the tags (or attributes) which are valid at any given point; makes it extremely easy to construct valid XML documents and actually quite hard ot construct invalid ones.
I would be very much in favour of a gradual conversion to XML-based configuration.
I'm surprised how many/. readers obviously still read print media. I guess I'm one of the older/. regulars, but I no longer do - I haven't bought a newspaper or a magazine for over a year.
Why?
Partly because they're ill-informed, badly written and out-of-date by the time you get them;
partly because they're increasingly controlled by large media empires with narrow political interests, so the diversity of political opinion expressed is extremely narrow and there are no papers which reflect my own political opinions;
but mostly because it's just so much easier to sit down at my desk, and either before starting work or in the inevitable quiet bits of the working day, just scan through my favourite news sites (of which/. is one).
I know a lot of people say they find screens hard to read from. I don't: I read from a screen all day every day, it's my normal mode of working. By contrast, I don't find reading from broadsheet newspapers at all convenient - they're too big to handle comfortably in almost any normal reading situation.
And I strongly agree with Jon's comments both on the quality of technology news in newspapers and their understanding of what constitutes culture. I have never read a story about the Internet in a newspaper which didn't contain at least three fundamental misconceptions in the first paragraph; and I find the idea that 'culture' is represented only by the forms popularised by the Austro-Hungarian empire in the eighteenth and nineteenth centuries risible in the extreme.
And, in Scotland at least, the tabloid-sized newspapers contain nothing but salacious gossip about vacuous 'celebrities' in whom I have no interest.
I recently did a similar exercise on UK politicians Web sites (my local representative had just asked me to design one for him) and the results were equally revealing and amusing.
But what gets me is how many of them are in the commercial comain,.com Does that mean you can buy them? 'Hi, I'm Al Gore, and I'm
running this candidacy as a commercial operation...'
Furthermore, you would NEVER run a database on Linux because there are problems with synchronous IO in the kernel. Transactional Integrity is a 100% MUST HAVE.
Total, idiotic, bollocks.
I've run large, significant, commercially important databases on Linux since kernel 1.2. I know a lot of other people who do, too. The critical things you need for an on-line transaction processing system are up-time and reliability. If you put in a Linux server you know it is just going to run; none of my database servers has ever had a software failure in four years. I've had a dead hard disk and a sticking processor fan, but never a software failure.
As far as transactional integrity goes, any serious database system has a begin/commit mechanism; it's not an operating system issue.
Now, sure, if I ran my databases on A/S400 I would get at least as good reliability; I might on AIX or Solaris. But Windows NT does not even begin to cut it as a database server.
The only problem I've had so far is when I get moderator access on/., half of the moderation options are rendered incorrectly or are out of position! Despite being a better browser, it's not good enough for Slahsdot:)
Well, if you think IE5 is bad, try reading Slashdot with Mozilla M13. You often get black text on a black background: cool, groovy, far out... but not readable. The problem here isn't IE5. I'm extremely hostile to Microsoft generally, but apart from a number of annoying trying-to-be-too-clever gimmicks IE5 is not a bad browser. Mozilla M13 isn't a bad browser either.
The problem is that Slashdot's HTML is atrocious. It's pretty much tuned to Netscape 4, and consequently looks reasonably alright in Netscape 4; but it's so bad that any reasonable HTML parser is going to have severe trouble with it. Run it through the validator if you don't believe me
You know what's really terrifying?
on
Textmode Quake
·
· Score: 3
On the AA Project home page, there's a section on 'Software known to support AA'; and there, second on the list, is The GIMP. textmode GIMP! Waaahhh, that is so perverse. I greatly fear I shall have to play with this...
I drew a different conclusion from this article. I saw WinNT roughly the same, Linux sharply up, Novell sharply down, and the conclusion I drew is that (most) of the Windows crowd are staying where they are, but a significant portion of the Novell market is moving to Linux. This is not to say that no-one is preferring Linux over WinNT.
Still as of this moment I don't believe Microsoft is suffering that much from the Linux advance (I believe it will); I think the people who are suffering are Novell.
Some people are really too clever. Among last year's entries try out banks.c - it is truly elegant... Never mind the obfuscation, to produce code that tight is just awsome.
But the one which had me falling off my seat giggling was this!
Efficient, cheap, pervasive...Re:Not a patch on /.
on
Is Usenet Dying?
·
· Score: 4
Uhhhmmmm... what are you comparing?
If everyone in the world visits the same Website to pick up their technical advice and chat, then that Website needs stupendous bandwidth, which costs a lot; which means that if it is to be provided in a capitalist environment it has to be covered with flashy adverts. Furthermore, that website is a single point of failure, and a single point of control - AKA censorship.
Usenet works (and will continue to work in the long term) because it distributes the cost of transmission of news across a very large number of nodes. Its store-and-forward method of propagation means that it is invulnerable to intermittencies in service. Each individual connection is relatively short distance, limiting congestion at network choke-points. It's easy to be selective of the groups you will cover on your own node, and, if you want groups which aren't available on your upstream node, it's easy to peer with another to feed them.
Furthermore, it's a profoundly subversive technology. It worked originally over dialup networks using UUCP; although we probably wouldn't use UUCP again, if the corporates succeed in getting control of the network backbone and deciding what we can transmit over it, Usenet would be incredibly easy to get working over other links.
Usenet will evolve of course, and ultimately will be supplanted by something else (RDF hints at some possibilities). But sites like/. cannot even begin to carry the wealth and breadth of expertise that washes over Usenet every day. The reason they can't compete is the sheer volume of material. If you think the/. effect is bad, just imagine the consequence of every current Usenet user hitting the same site at the same time. Any replacement for Usenet must be distributed.
I'm not really at all disappointed that new Net users aren't adopting Usenet in droves. Usenet has always had a problem with acculturating large numbers of newcomers at the same time (it's always September...), and, as the people who are joining the Net are increasingly drawn from a broad, non-technical, non-academic public, the proportion who really have something to contribute (and who are able to express that something clearly and eloquently) declines. If Usenet were to dwindle into the hangout of the hardcore techies of the Net, that wouldn't worry me at all.
This may happen. But what won't happen is that Usenet will just die. Enough of the people who use it and care about it have the ability and resources to maintain it that you can guarantee it's survival for a long time. I first used it fifteen years ago; I will still be using it fifteen years into the future - not as my only information source, but as one of my information sources.
I've been reading through what's already been written, and much of it's good advice. There are some things I would add:
Go with people you trust
I've done this three times, and twice come unstuck because of failures of trust between the technical people and the finance/sales people.
In the first case we were in a long development cycle for a product. The finance director kept giving us reports which said, yes, things were tight, but we were solvent and could do it. If he hadn't, we'd have diverted effort into consulting, which would have made a bit of immediate income.
In the end, just as we were about to launch the product, we ran out of money very hard. The finance director (who was experienced, and a personal friend) had been keeping two sets of books, and showing the rest of the team doctored figures. His excuse? 'If you'd known how bad things really were, you have gone off and done other things and we'd never have got the product finished'. Well meant, but not helpful.
In the second, the marketing director, a very experienced businessman introduced to us by the local business development agency, tride to bribe a local government representative to get a us contract. After we'd launched product, as sales were taking off, when we were already nicely profitable. Well meant, but very, very stupid.
If you're east of the Atlantic, don't waste time with VCs
I could not possibly count the hours I've put in to trying to raise venture capital. With excellent business plans - I had a business plan for Web auctions fully fleshed out and costed back in 1996, for example - competent key people, all that was needed (except finance) to make the project fly. It's just a complete waste of time, here in the UK. UK Venture Capitalist houses are extremely cautious and do not like either technical projects or geeks. If you invested that same amount of time in development that you waste hunting VC, you'd get your product to market.
I know only three groups who've successfully gone down the VC route, two from here in Scotland, one from France. All three moved their whole companies to California, first.
Don't take it too seriously
But finally, remember that it's a gamble, a chancy game. I wish you all the best: but, if you fail, remember that it's a set-back, not a disaster. Don't put yourself in a position where the company going down would personally bankrupt you. If it does happen, and you're reasonably sure the reason it happened wasn't due to your incompetence of some deep personal failing, brush yourself off, sort the mess out, and start again.
Actually, I first learned to program on an 8k PET with the funny little snap-on-top keys with the symbols written on paper under the snap-on tops. Yes, it was a crap machine - but it was one of the very first computers which ordinary people could actually buy. I've actually still got a PET - a later, 32k model with a proper keyboard and twin disk drives. And yes, it still works.
This is an important part of the history of our culture, and a very fleeting one. Those early machines were in use for only half-a-dozen years between the emergence of the first micro-processors in the mid seventies and the arrival of IBM's horrible Intel based kludge in 1982; within that time there was incredibly rapid evolution, and many interesting designs were tried. The 8K PET deserves an honourable mention in this process because it was one of the first machines to come all in one box, ready to buy off the shelf; most of it's contemporary competition was kits, often without provision for a keyboard or a video display.
It is also the direct and very obvious ancestor of the Commodore 64K and the VIC20, both of which were common home/games machines in the early eighties.
The primary goal of XHTML is to allow you to extend the core set of tags with your own tag sets so that you may add markup functionality without breaking the standard (as has been done in the past).
No! This is a total misunderstanding. XHTML 1.0 is simply a recasting of HTML 4.01 into XML compliant syntax. You cannot extend XHTML as such by adding your own tags. You can produce hybrid documents by combining XHTML with other XML dialects, but the result would not be XHTML. You could even combine XHTML with XML dialects you create yourself. But you would be very foolish to do so.
XML dialects are only useful if they serve a significant community who have tools which understand the dialect and can do useful things with them. If you just make it up yourself as you go along, then the only thing you can really do with it is use XSL to translate it back into standard XHTML, so you've gained nothing.
I run an ASUS P2B-DS with two PII/300s and two IBM ultrawide SCSI disks. The machine is bombproof and very fast. This is my desktop machine and is used largely for Java development - none of our servers have or need anything like this performance.
However, I wouldn't build or buy a new machine just now. I'm waiting to see whether anyone is going to put together a four-way or eight-way Crusoe motherboard and if so what it will cost. The Crusoe is a much smaller die than the Pentium and consequently is very much cheaper to produce. This doesn't mean it will be sold cheaper, but it could be; I'm prepared to bet the real production cost of eight Crusoes is significantly less than two PIIIs
Slashdot Mirror
Remember that the Digital Millenium Copyright Act has force of law in only one of the over two hundred countries in the world, and, as it happens, not all the Samba developers live there.
However, suppose fifty people were to download it without agreeing to the licence (I just have, from http://warot.com/freedom/kerbspec.pdf), and each quoted just one paragraph on their web site (copyright law allows for the fair-use quoting of small parts of a document), and then suppose someone else later came along independently and made a Web page which linked to all those single paragraph Web pages, what law would have been broken?
As an example of what I mean, I have posted an example paragraph here
It's a sad day when Slashdot feels it has to explain who Vint is. Vint is more than the guy who wrote TCP/IP; he's also one of the people who has lead and driven the Internet all these years, and worked to keep it open for all of us. He's also a really nice person to meet. He's one of the ubergeeks of the Net, alongside Jon Postel; one of the real greats.
Fortunaltely for us all, though, Vint is still very much with us!
You mislepped 'HURD'. Hope this helps.
Oh, and HURD doesn't thunder. It moves very slowly towards a point which recedes infinitely towards the HURD release event horizon.
Bollocks.. ICrave, as I understand it, were doing absolutely nothing at all in US jurisdiction. None of their equipment was in US jurisdiction. They had a server in Canada which responded to requests, but only to requests from people who said they too were in Canada. If someone from the US requested material which was copyright in the US, then that was midly naughty. If in order to do so they dishonestly pretended not to be in the US, then that was naughty too. But in neither case was ICrave in the wrong.
It's been pointed out before in this thread that US jurisdiction does not and cannot cover things people do outside the United States borders. The inability of some American contributers to this thread to understand this very simple fact merely confirms for those of us who live outwith the United States our already low opinion of the average intellectual capacity of those who don't.
In other words, we're not surprised you don't understand. We would like to be surprised, but we aren't holding our breath.
Oh, and: Moderation -1, Flamebait
That's how it goes with open source.
You want it, you get off your backside and write it. If you can't be bothered to write it, you don't complain.
I normally read /. at with my filters set to '2 and above' so as to avoid this sort of flame fest, but this one attracted my attention and I backtracked it.
Some observations: it's notable that the people (or possibly just one person) attacking RMS here doesn't choose to reveal [his|her] identity.
Secondly, there's some extremely sloppy thinking going on. For example, the assertion that just because you paid for RedHat, anything you develop under RedHat you can automatically sell on. You can't. The C compiler RedHat distribute is the Gnu C compiler; the standard libraries RedHat distribute are the GNU libraries. If you use those tools, and if you believe software licences have any validity at all, you must abide by their licences. RedHat cannot issue those tools to you under any different licence than the GPL, and don't try.
Then there is the assertion that RMS, the Free Software Foundation, and Debian want to stop you making money out of software. This is just simply obviously untrue. RMS makes his living out of software; so did the guys at Cygnus who have done so much of the maintenance of the GNU toolkit over the past several years. What RMS wants to persuade you not to do is make money out of hiding the source of software, restricting the freedom of users of software.
Finally, there is the assertion that Marxism is the same as socialism, and that both are the same as standing up for the right to freedom of speach. Anyone who can believe this is either so politically naiive, or so indoctrinated, that their other opinions are invalidated.
Face it: RMS, for all his querkiness and his occasional displays of ego, is necessarily one of our heroes. Without him, there would be no GNU toolset; without the GNU toolset, there could be no Linux. There's no sense in the sort of argument you see frequently between supporters of different free software/open source heroes. All of them - RMS, ESR, Linus, Alan Cox, even Larry Wall - have contributed hugely to making the movement we find ourselves in. All of them are human and have their querks, but we all owe them all a huge debt of gratitude, and this sort of anonymous attack is simply childish and undignified.
Country codes are very much more political than you think. The current list is simply drawn from ISO 3166, which lists territories of all sorts, some of which are independent countries and some of which are not. Many countries which feel they should be independent but are not (such as my country, Scotland) do not have ISO 3166 listings, and so don't have TLDs; indeed we have the anomalous situation where a small island I can see from the top of the hill behind my house has a TLD (IM, the Isle of Man), but my country doesn't. Furthermore, all the obvious two letter abbreviations of Scotland are already taken.
Centralist governments which are trying to resist nationalist movements (such as in Britain, Spain or Indonesia) strongly do not want to allow the aspirant nations which they control to have TLDs; independence movements within those nations strongly do.
This isn't an argument either to have country-code TLDs or not to have them, but it's worth being aware that there are very strong political undercurrents around this issue which will not make it easy to solve.
Finally, I suspect that governments particularly of smaller and less self-confident nations would strongly resist moves to do away with CCTLDs.
The thing that worries me most about this is an email I received this morning from my lawyers, which I'll quote:
.Com .COM, .EDU, and .ORG top level domains registered. Although the Decision
------------------- begin quote -----------------
US SEIZES JURISDICTION OVER DOT.COM COMPANIES
Mere registration of top level domain sufficient
US companies able to seize worldwide registrations
US trademark owners able to have domain names of others expropriated
Over the last few years we have commented on Internet cases from around the world which
we believe would be of interest to UK businesses and others. For some readers these
cases may have been more relevant than others. Today however we report on a decision of
the District Court for the Eastern District of Virginia which has perhaps the furthest
reaching implications for worldwide internet governance.
In its Decision of Friday 5th March 2000, Caesars World, Inc -v- Caesars-Palace
and others (Civil Action No 99-550-A), Judge Albert Bryan effectively decreed that his
court would be the arbiter of the property rights in respect of all the approx.
7,000,000
dealt solely with the motions of two of the defendants to dismiss an action for lack of
jurisdiction, the effect of the Decision is to open the floodgate to litigation by
holders of US trademarks against domain name proprietors based outwith the US (or at
least outwith Virginia.
-------------------- end quote ------------------
Now, it may have escaped your attention in America, but the Internet is international; it has been international since Arpanet was linked to Janet in the early eighties. When we set up ICANN (I say 'we' - I was convenor of one of the Geneva sessions of the working party which led to the setting up of ICANN) it was precisely to prevent this sort of thing from happening.
The only possible consequence of this sort of nonsense is that the domain name service will collapse; courts and politicians in Europe (and, I imagine, in Japan) are not going to like being told that American trade-mark owners have first claim on an international resource. Either the Virginia court is persuaded to see sense, or we'd better get on with developing a replacement for the Domain Name System.
No, it's not, it's specifically what people were asked to do by IANA: to use .net addresses for all elements of the public network backbone, and .com addresses for everything else.
It may have escaped your notice, but /. is a commercial venture of Andover.Net, inc. Both 'slashdot.org' (which isn't a non-profit) and 'andover.net' (which doesn't own any part of the network backbone) are gross abuses of the namespace.
I've been using talkback versions of Mozilla since about M8, I think - it doesn't always work. In my experience four out of five crashes go down so hard and so fast that talkback never gets a chance. Maving said that, M13 was in my opinion very nearly as stable as 4.7 - but having said *that* I've now ditched 4.7 in favour of 4.6, because I found 4.7 so fragile as to be more or less unusable.
I would be very much in favour of a gradual conversion to XML-based configuration.
Why?
I know a lot of people say they find screens hard to read from. I don't: I read from a screen all day every day, it's my normal mode of working. By contrast, I don't find reading from broadsheet newspapers at all convenient - they're too big to handle comfortably in almost any normal reading situation.
And I strongly agree with Jon's comments both on the quality of technology news in newspapers and their understanding of what constitutes culture. I have never read a story about the Internet in a newspaper which didn't contain at least three fundamental misconceptions in the first paragraph; and I find the idea that 'culture' is represented only by the forms popularised by the Austro-Hungarian empire in the eighteenth and nineteenth centuries risible in the extreme.
And, in Scotland at least, the tabloid-sized newspapers contain nothing but salacious gossip about vacuous 'celebrities' in whom I have no interest.
But what gets me is how many of them are in the commercial comain, .com Does that mean you can buy them? 'Hi, I'm Al Gore, and I'm
running this candidacy as a commercial operation...'Total, idiotic, bollocks.
I've run large, significant, commercially important databases on Linux since kernel 1.2. I know a lot of other people who do, too. The critical things you need for an on-line transaction processing system are up-time and reliability. If you put in a Linux server you know it is just going to run; none of my database servers has ever had a software failure in four years. I've had a dead hard disk and a sticking processor fan, but never a software failure.
As far as transactional integrity goes, any serious database system has a begin/commit mechanism; it's not an operating system issue.
Now, sure, if I ran my databases on A/S400 I would get at least as good reliability; I might on AIX or Solaris. But Windows NT does not even begin to cut it as a database server.
Well, if you think IE5 is bad, try reading Slashdot with Mozilla M13. You often get black text on a black background: cool, groovy, far out... but not readable. The problem here isn't IE5. I'm extremely hostile to Microsoft generally, but apart from a number of annoying trying-to-be-too-clever gimmicks IE5 is not a bad browser. Mozilla M13 isn't a bad browser either.
The problem is that Slashdot's HTML is atrocious. It's pretty much tuned to Netscape 4, and consequently looks reasonably alright in Netscape 4; but it's so bad that any reasonable HTML parser is going to have severe trouble with it. Run it through the validator if you don't believe me
On the AA Project home page, there's a section on 'Software known to support AA'; and there, second on the list, is The GIMP. textmode GIMP! Waaahhh, that is so perverse. I greatly fear I shall have to play with this...
Still as of this moment I don't believe Microsoft is suffering that much from the Linux advance (I believe it will); I think the people who are suffering are Novell.
But the one which had me falling off my seat giggling was this!
Uhhhmmmm... what are you comparing?
/. cannot even begin to carry the wealth and breadth of expertise that washes over Usenet every day. The reason they can't compete is the sheer volume of material. If you think the /. effect is bad, just imagine the consequence of every current Usenet user hitting the same site at the same time. Any replacement for Usenet must be distributed.
If everyone in the world visits the same Website to pick up their technical advice and chat, then that Website needs stupendous bandwidth, which costs a lot; which means that if it is to be provided in a capitalist environment it has to be covered with flashy adverts. Furthermore, that website is a single point of failure, and a single point of control - AKA censorship.
Usenet works (and will continue to work in the long term) because it distributes the cost of transmission of news across a very large number of nodes. Its store-and-forward method of propagation means that it is invulnerable to intermittencies in service. Each individual connection is relatively short distance, limiting congestion at network choke-points. It's easy to be selective of the groups you will cover on your own node, and, if you want groups which aren't available on your upstream node, it's easy to peer with another to feed them.
Furthermore, it's a profoundly subversive technology. It worked originally over dialup networks using UUCP; although we probably wouldn't use UUCP again, if the corporates succeed in getting control of the network backbone and deciding what we can transmit over it, Usenet would be incredibly easy to get working over other links.
Usenet will evolve of course, and ultimately will be supplanted by something else (RDF hints at some possibilities). But sites like
I'm not really at all disappointed that new Net users aren't adopting Usenet in droves. Usenet has always had a problem with acculturating large numbers of newcomers at the same time (it's always September...), and, as the people who are joining the Net are increasingly drawn from a broad, non-technical, non-academic public, the proportion who really have something to contribute (and who are able to express that something clearly and eloquently) declines. If Usenet were to dwindle into the hangout of the hardcore techies of the Net, that wouldn't worry me at all.
This may happen. But what won't happen is that Usenet will just die. Enough of the people who use it and care about it have the ability and resources to maintain it that you can guarantee it's survival for a long time. I first used it fifteen years ago; I will still be using it fifteen years into the future - not as my only information source, but as one of my information sources.
Simon, aka control@scot.news-admin.org
Go with people you trust
I've done this three times, and twice come unstuck because of failures of trust between the technical people and the finance/sales people.
In the first case we were in a long development cycle for a product. The finance director kept giving us reports which said, yes, things were tight, but we were solvent and could do it. If he hadn't, we'd have diverted effort into consulting, which would have made a bit of immediate income.
In the end, just as we were about to launch the product, we ran out of money very hard. The finance director (who was experienced, and a personal friend) had been keeping two sets of books, and showing the rest of the team doctored figures. His excuse? 'If you'd known how bad things really were, you have gone off and done other things and we'd never have got the product finished'. Well meant, but not helpful.
In the second, the marketing director, a very experienced businessman introduced to us by the local business development agency, tride to bribe a local government representative to get a us contract. After we'd launched product, as sales were taking off, when we were already nicely profitable. Well meant, but very, very stupid.
If you're east of the Atlantic, don't waste time with VCs
I could not possibly count the hours I've put in to trying to raise venture capital. With excellent business plans - I had a business plan for Web auctions fully fleshed out and costed back in 1996, for example - competent key people, all that was needed (except finance) to make the project fly. It's just a complete waste of time, here in the UK. UK Venture Capitalist houses are extremely cautious and do not like either technical projects or geeks. If you invested that same amount of time in development that you waste hunting VC, you'd get your product to market.
I know only three groups who've successfully gone down the VC route, two from here in Scotland, one from France. All three moved their whole companies to California, first.
Don't take it too seriously
But finally, remember that it's a gamble, a chancy game. I wish you all the best: but, if you fail, remember that it's a set-back, not a disaster. Don't put yourself in a position where the company going down would personally bankrupt you. If it does happen, and you're reasonably sure the reason it happened wasn't due to your incompetence of some deep personal failing, brush yourself off, sort the mess out, and start again.
This is an important part of the history of our culture, and a very fleeting one. Those early machines were in use for only half-a-dozen years between the emergence of the first micro-processors in the mid seventies and the arrival of IBM's horrible Intel based kludge in 1982; within that time there was incredibly rapid evolution, and many interesting designs were tried. The 8K PET deserves an honourable mention in this process because it was one of the first machines to come all in one box, ready to buy off the shelf; most of it's contemporary competition was kits, often without provision for a keyboard or a video display.
It is also the direct and very obvious ancestor of the Commodore 64K and the VIC20, both of which were common home/games machines in the early eighties.
Oh yes oh yes oh yes! Not just one and two but three as well.
Settlers and Braben's original Elite and I'd be happy...
Simon, already has CivCTP.
No! This is a total misunderstanding. XHTML 1.0 is simply a recasting of HTML 4.01 into XML compliant syntax. You cannot extend XHTML as such by adding your own tags. You can produce hybrid documents by combining XHTML with other XML dialects, but the result would not be XHTML. You could even combine XHTML with XML dialects you create yourself. But you would be very foolish to do so.
XML dialects are only useful if they serve a significant community who have tools which understand the dialect and can do useful things with them. If you just make it up yourself as you go along, then the only thing you can really do with it is use XSL to translate it back into standard XHTML, so you've gained nothing.