With respect to the monolithicness of the file...
on
Apache Cookbook
·
· Score: 2, Informative
My favourite is simply
# Suck in configs for the various sites...
#
include/usr/local/etc/www-sites
And simply fill that directory with small little config files, one per site or logical function. Or often, symbolic links to the actual files in their project directory which are under cvs control. (Use something like.../*.conf to limit where needed).
Some care is needed with perms; apache will quite happily
such in what is there; and careless use of symlink or allowing the creation of such may cause security holes.
Some people question the need for this; just some background as to why we in Wireless Leiden need this patch:-)
The issue is that througout the city we have omni antenna's - where -anyone- can associate with - and directional antennas which provide the interlinks between nodes (although the network covers a medium sized city - we use no copper; all interlinks are wireless).
On these interlinks we only want node-to-node traffic.
As the network is totally open (no username, password or any thing) - we have no easy way to educate our users to use the right 'omni' antenna's, other than descriptive names. I.e. we do not catch them early enough.
So often people associate with the interlinks rather than the omni (if a beam passes over their house) - and then complain, or are surprized, that DHCP does not give them an address.
This problem is made worse by some windows userinterface tools which will automatically re-select networks based on some internal metric.
So what we wanted was to 'hide' the interlinks. So that clueless users are not accidentally ensnared. Rolands patch does exactly this.
And folks do grab your chance - put your money where your mouth is - and make that Donation (anonymous if you need to and are worried about google). 100$ is not that much - and 500x100$ = 50.000, which in Europe is quite a fortune when it comes to decent legal defence.
At Wireless Leiden you can read about a volunteer efford which has already build a complete and city wide WiFi network which is free and open (no logon, just open your laptop and hop on).
The 34+ nodes cover a medium sized city (120.000 inhabitants). They have been build by volunteers and rely on donated hardware, locations and the odd bit of electricty donated.
You can fetch the code for a cost of a download (see WiKi: NodeFactory) - all is open source; FreeBSD, OSPF, DHCP, SNMP and SSH are the key bits of technology.
Good to see that commercial effords are trying to follow suit.
This seems to only affect machines which did not come through an upgrade path from 10.1 or before; but had Panther instaleld on them cleanly.
The solution is documented in/etc/ttys, simply change the secure of the console to a insecure:
If the console is marked insecure, single-user requires the root password.
Since DirectoryServices is not running by the time we enter single-user mode, init will ask for the non-shadow crypt password stored for root in/etc/master.passwd. If no such password exists, it will not be possible to enter single-user mode from a console marked insecure.
I.e. The lines you want to edit is/are (with sudo vi/etc/ttys):
console "/usr/libexec/getty std.9600" vt100 on insecure
console "/System/Library/CoreServices/loginwindow.app/Cont ents/MacOS/loginwindow" vt100 on insecure onoption="/usr/libexec/getty std.9600"
Given that you propably still want to be able to log in if you have to - you propably also want to do:
netinfo or other default passwd:
sudo passwd root
default passwd file used during early boot stages
sudo passwd -i file root
Some people prefer the apache license - as it allows a wider range of bisness models to be used around the code. Which in their opion creates a healthier eco system around the code.
Secondly the ASF has always been very serious about the legal footwork and going through the Java certification processes.
Some year ago the ASF negotiated a change to the certification process of SUN which allowed any academic or open source project (apache, or non apache, BSD, Apache, GPL or otherwise licensed) to be certified by
taking the cost issue out of the equation.
Unfortunately none of the open source J2EE implementations have made use of that. This is a large driver for part of the community to work on a version which is properly certified.
Be aware that there is, and always may be, some overlap between the developers who work, or have worked on JBoss and are working on Geronimo.
And unless that developer fully signed over the (c)/ownership to either group (which is certainly not the default for both the apache world and the jboss world) then he or she is most certainly able to contribute his own code to both projects.
This is wht makes these things such a puzzle to sort out - and why doing it in the open really really helps to keep things fair.
There is actually in enteresting task for anyone who runs 5.x, the change to this new rcorder and/etc/rc.d/ layout means that most, if not all, of the dependency info this linux article talks about - are already encoding in the rc.d files.
As all these scripts are under control of rcorder - it should be very easy to change its guts to do things parallel.
Or in other words - -you'll find more of an ready made bed in freebsd - and applying the linux article should in fact be easier than on linux itself:-)
Very true - however there is an additional issue; if you believe the premisse that the ice melts than it is fair to assume that the (global or at least very large region) temperature has gone up a wee bit.
If that is true; then the sea water will get slightly warmer. And given that most sea water is well above 3celcius it will thus exand a little. This by itself causes a significant enough rize to cause issue.
Us, that is the folks at http://www.wirelessleiden.nl/ have done very much the same. Lets add the link to Subversion (cvs like source code mngt. system) with all the code - so that we can at least copy each others wheels.
Of course - at that time there was a fair range of budding programming languages - and one of the prime reasons that C took such a leading position was that Kernigham wrote a most excelent tutorial on 'programming in general' and in C specifically - and, in his own words, twisted 'Ritchie his arm into writing a book with him'.
You really want to read up on HIPAA - as it requires quite a quite a few very specific things for medical/hospital use which just have absolutely nothing to do with the buzzword 'military grade security'.
In fact - there are a number of requirements in HIPAA with respect to accountability and privacy which run rather counter to the more traditional requirement/compromizes made in military systems where both hierachy and the desire to do counter-intelligence are fundamentally different. And thus each need its own set of engineering compromises.
This is why just sprinkle some 'trusted unix' as pixy dust - and pretend you are HIPAA compliant is just not working:-)
But seriously - do read up on it; the HIPAA standards (http://www.hhs.gov/ocr/hipaa/) are surprizingly readable and actually very preceise with clear lists of requirements. Almost a checklist.
Yet, despite all those 22.000 laws - the US sees regular shootings at factories, schools and companies.
About 2 or 3 a year we read about in the newspaper here (and about double that when you follow US news papers closely).
At the same time, I have a hard time recalling any incident even remotely similar in Holland, Italy, France or the United Kingdom.
Perhaps just a single law; No guns for anyone, would be a tad bit more effictive.
And given the number of revolts or swapping of governements/those-in-power we've seen in those countries in the past centuries (while these gun laws where in place) - I have no worry about citizens not being able to rise against their governement should they feel like it.
Call it 'try by experience' - but that part seems to work fine and not require some amendment or constitution to allow them to bear arms. In fact the newbies in power seem to generally relish in rewriting the darn thing from scratch. (Nor does it seem to stop the criminals from getting guns either).
You may want to consult: http://www.domesdaybook.co.uk/ for the authoritative spelling. Which is indeed 'DomesDay'. For those who did not have history at school:
The King wanted to know what he had, and who held it. The Commissioners therefore listed lands in dispute, for Domesday Book was not only a tax-assessment. To The Kings grandson, Bishop Henry of Winchester, its purpose was that every "man should know his right and not usurp another's"; and because it was the final authoritative register of rightful possession "the natives called it Domesday Book, by analogy from the Day of Judgement"; that was why it was carefully arranged by Counties, and by landowners within Counties, "numbered consecutively... for easy reference".
Also note that in england at that time domesday was a regular, repeating, day on when judicial decisions were announced which essentially could not be appealed. Just like the book could not. So one can argue of the christian judgement referenece is all that accurate and if it was not the other way round; the christian references was named after the every day scheduled judgement day in normal life.
In germany (and in fact in most countries visited by Napoleon) the broad 'as-is' disclaimer generally is quite an issue.
In the US it is very normal that two 'grown up' parties agree to something fair reaching; such as waiving certain rights or liabilties with respect of each other.
In most of (continental) europe that is not quite so easy; and the contract or agreemnt which two parties may have with each other may simply be overclassed by national law or 'common sense' in that respect.
The national law dictates that there are certain minimal levels and that disowning it all is simply not an option.
So regardless of what the developer (dis)claims with respect to warranty; the court may well held him liable to a certain extend.
At the same time, there is also a bright side; those liabilities are generally much more limited and 'capped' than in the US; and hardly ever exceed a small multiple of the resonable sum/economic value of the good (and not what can be done with that good). And they also put very reasonable demands on the 'user'. Willy nilly risking 5 million of lost production on a bit of untested free software is not going to ring true with the judge. He expects (more) resonable caution than generally in the US.
Also note that the scope of damages is very propotional to the purchace/gains of the developer/transaction. Sor 'free' (as in gratis') software those amounts are obviously not going to be very large.
Except if there is a bit more blame; i.e. someone knowingly dropping the ball. And unlike the US, where that waiver is going to help you - it may do little or nothing in most of Europe. Whereas in the US you are fairly secure.
On the other hand - any secondary damages issues are not nearly as much of a problem in europe, and virtually unheard of. Plus bear in mind that cost recovery and legal assistance is on an entirely different level in germany compared to the US. This making the issue of frivolous lawsuits by a megacorp which cripples a small developer virtually unheard of in most of the EU.
So in short; yes - you are bit more open to exposure in Europe - but as long as you behave resonably and are not vandalizing the hight of that exposure is very limited; and proportional to your fairly direct and clear cut gains from that software. And with open source / gratis - that is not going to make you go bust.
Note that similar rules already exist in most of europe for other types of non registered media. This just confirms that the internet becomes more normal once more.
And it once again sees removed a large chunk of friviolous, US inspired, type of lawsuits we can just do without, from the general 'speak your mind on the net' equasion. Not to mention the fact that it equals the playing field between the big guys and the small guys. So good for the small guys; and tough, but fair to the big guys.
As to verifying the authentity; that it is really up to the publisher if he or she wants to limit the amounth of counter argument it wants to publish.
You'd be surprized to learn how seriously some groups in the *BSD world and for example in the Apache Software Foundation take those things.
And indeed at the ASF we go through the legal paperwork to make sure that every commit is accounted for, every change monitored. Every code donation is matched with the right paperwork and grant documents.
And while it is certainly true that a single 'big company' may not be able to guarantee such checks and balances - a crowd or whole ecosystem does. In a very effective way.
Ultimately I think we will all find that this is what is needed to keep this "open source' world healty and viable.
It is just that perhaps the FreeBSD crowd found this out earlier, due to the AT&T/Berkeley lawsuit.
Dw
"Whats wrong with wires? Cheaper to implement, fix, replace"
Though that wire may be cheap - the actual connector and making sure the holes are in the right place of the casing; and making sure things are easy to assemble are not.
It is pretty common for 80% of the cost of something like temperature sensor with serial port, or a USB to FM radio, to be in the housing, assembly and the thoughdful steps to get the connectors mechanically sensible in place. And not with the electronics which does the sensible thing.
With bluetooth it is just some easy to place SMD stuff and a bit of firmware. Note of course that all is a compromize; because now you have a battery problem (and perhaps an edge connector for that) and you need some UI to pair the device to the right other endpoint.
But bluetooth is a useful element in the compromize game to make things cheaper.
Not so sure if all systems are that well prepared. As one cannot -predict- the UTC/TAI delta into the future.
Hence equipment needs to be told about it in real time. Or in other word; you cannot build something which will work for 10 years and 'does the right thing' without user intervention.
A lot of people use UTC as if it is TAI; i.e.. they take two timestamps in UTC, substract them and expect that the difference between them is the number of seconds elapsed between those two. Whereas in fact one would need to know if any additional TAI/UTC seconds where added or removed in that particular period.
It seems that a lot of the objection against UTC leap second is coming from people who use it as if it was a TAI time; i.e. a predictable monotomic increasing value. And UTC has been that for many years; as we needed few leap seconds.
Unix has the same issue; we use UTC in a lot of places; which is fine for crontab's which need to happen at time x or y - yet calculate differences/delays as (x-y) ; which really is wrong; we should use TAI for that.
So parhaps it is time to start using TAI and calculate UTC on the fly based on the offset (today that is 32 seconds). Like/usr/share/timezone this would require a table with the dates/times (in TAI) at which delta's where introduced with respect to UTC.
In most European countries it is very, very common for the 'attacker' to have to pay all court costs; including those of the person who had to defend him or herself.
This is an esspecially important, not to say effective, mechanism which prevents the sort of litigation-wear-down so commonly seen in the US. Which skews the
system towards those with the deepest pockets.
And more than once has a judge in Europe added something extra because the justitial apparatus wanted
to give the signal that it was seriously annoyed by such superfluous lawsuits with no merit.
Note that WMS demo is propably a better place to start. It uses the OpenGIS WMS (Web Mapping Server) protocol. There are some 10 databases behind this server; apache and freebsd committers/mirrors just being a small part of it.
Ideally one would use something like the WMSClient as mentioned on this page.
Are you sure ? (see http://www.benzedrine.cx/pf.html for a nice 'TODO' list for netfilter/ipchains).
Specific things which are 'killer apps' to me are the ease by which you can configure things like blocks/groups of ports, services, machines - and then re-use those in your cluster config or wider rules.
This has safed us many hours of debugging downtime; and allows a much wider new range of 'new sysadmins to be' to deal with the simpler requests; such as adding a machine, service or port,
As it is fairly safe to modify the right table/macros/lists - without having to have a complete understanding of every detail. And yet, when designed by an experienced hand, robust enough not to be bitten by an overly broad side effect.
Have a look at http://www.openbsd.org/cgi-bin/cvsweb/~checkout~/s rc/etc/pf.conf?rev=1.19&content-type=text/plai n - to see good examples.
Besides, most people don't need all the bells and wistles Apache offers.
How about simply stripping apache down; i.e. just run those two or three modules you really need. On for example a static images only server you could get away with just a logging module and the mmap module quite nicely. (Though realistically alias/rewrite is usually needed (or lots of symlinks) when your friendly marketing staff barges in with yet-another-campain which breaks all historic links.
Its modular - go play with it - and have lots of fun. (And yes - you can actually run apache completely and validly with just 2 lines of config.
Slashdot Mirror
Some care is needed with perms; apache will quite happily such in what is there; and careless use of symlink or allowing the creation of such may cause security holes.
Dw.
The issue is that througout the city we have omni antenna's - where -anyone- can associate with - and directional antennas which provide the interlinks between nodes (although the network covers a medium sized city - we use no copper; all interlinks are wireless).
On these interlinks we only want node-to-node traffic.
As the network is totally open (no username, password or any thing) - we have no easy way to educate our users to use the right 'omni' antenna's, other than descriptive names. I.e. we do not catch them early enough.
So often people associate with the interlinks rather than the omni (if a beam passes over their house) - and then complain, or are surprized, that DHCP does not give them an address.
This problem is made worse by some windows userinterface tools which will automatically re-select networks based on some internal metric.
So what we wanted was to 'hide' the interlinks. So that clueless users are not accidentally ensnared. Rolands patch does exactly this.
Dw
Head to http://www.choicepc.com/ and get yourself that t-Shirt you cannot buy at the gap.
Dw
The 34+ nodes cover a medium sized city (120.000 inhabitants). They have been build by volunteers and rely on donated hardware, locations and the odd bit of electricty donated.
You can fetch the code for a cost of a download (see WiKi: NodeFactory) - all is open source; FreeBSD, OSPF, DHCP, SNMP and SSH are the key bits of technology.
Good to see that commercial effords are trying to follow suit.
Dw
The solution is documented in /etc/ttys, simply change the secure of the console to a insecure:
I.e. The lines you want to edit is/are (with sudo viconsole "/usr/libexec/getty std.9600" vt100 on insecure
console "/System/Library/CoreServices/loginwindow.app/Cont ents/MacOS/loginwindow" vt100 on insecure onoption="/usr/libexec/getty std.9600"
Given that you propably still want to be able to log in if you have to - you propably also want to do:
netinfo or other default passwd: sudo passwd root
default passwd file used during early boot stages sudo passwd -i file root
Note that in most cases you want to change both.
Dw
Secondly the ASF has always been very serious about the legal footwork and going through the Java certification processes.
Some year ago the ASF negotiated a change to the certification process of SUN which allowed any academic or open source project (apache, or non apache, BSD, Apache, GPL or otherwise licensed) to be certified by taking the cost issue out of the equation.
Unfortunately none of the open source J2EE implementations have made use of that. This is a large driver for part of the community to work on a version which is properly certified.
Dw.
And unless that developer fully signed over the (c)/ownership to either group (which is certainly not the default for both the apache world and the jboss world) then he or she is most certainly able to contribute his own code to both projects.
This is wht makes these things such a puzzle to sort out - and why doing it in the open really really helps to keep things fair.
Dw.
See see the REQUIRED et.al. in /etc/rc.d/diskless for an example.
As all these scripts are under control of rcorder - it should be very easy to change its guts to do things parallel.
Or in other words - -you'll find more of an ready made bed in freebsd - and applying the linux article should in fact be easier than on linux itself :-)
Dw
If that is true; then the sea water will get slightly warmer. And given that most sea water is well above 3celcius it will thus exand a little. This by itself causes a significant enough rize to cause issue.
Assuming that the premissie holds of course.
Dw
Background: http://www.wirelessleiden.nl/wcl/cgi-bin/moin.cgi/ NodeFactory
Code (in public subversion):g /factory/trunk/install/install.sh
g /factory/trunk/
http://wleiden.webweaving.org:8080/svn/node-confi
http://wleiden.webweaving.org:8080/svn/node-confi
Machines using the code: http://www.wirelessleiden.nl/wcl/cgi-bin/moin.cgi/ NodeMap
Though this one is a bit more fully fledged; as it also includes SNMP management and OSPF routing.
Dw.
Of course - at that time there was a fair range of budding programming languages - and one of the prime reasons that C took such a leading position was that Kernigham wrote a most excelent tutorial on 'programming in general' and in C specifically - and, in his own words, twisted 'Ritchie his arm into writing a book with him'.
Dw
In fact - there are a number of requirements in HIPAA with respect to accountability and privacy which run rather counter to the more traditional requirement/compromizes made in military systems where both hierachy and the desire to do counter-intelligence are fundamentally different. And thus each need its own set of engineering compromises.
This is why just sprinkle some 'trusted unix' as pixy dust - and pretend you are HIPAA compliant is just not working :-)
But seriously - do read up on it; the HIPAA standards (http://www.hhs.gov/ocr/hipaa/) are surprizingly readable and actually very preceise with clear lists of requirements. Almost a checklist.
Dw
Yet, despite all those 22.000 laws - the US sees regular shootings at factories, schools and companies.
About 2 or 3 a year we read about in the newspaper here (and about double that when you follow US news papers closely).
At the same time, I have a hard time recalling any incident even remotely similar in Holland, Italy, France or the United Kingdom.
Perhaps just a single law; No guns for anyone, would be a tad bit more effictive.
And given the number of revolts or swapping of governements/those-in-power we've seen in those countries in the past centuries (while these gun laws where in place) - I have no worry about citizens not being able to rise against their governement should they feel like it.
Call it 'try by experience' - but that part seems to work fine and not require some amendment or constitution to allow them to bear arms. In fact the newbies in power seem to generally relish in rewriting the darn thing from scratch. (Nor does it seem to stop the criminals from getting guns either).
Dw.
Also note that in england at that time domesday was a regular, repeating, day on when judicial decisions were announced which essentially could not be appealed. Just like the book could not. So one can argue of the christian judgement referenece is all that accurate and if it was not the other way round; the christian references was named after the every day scheduled judgement day in normal life.
Perhaps.
Dw
In germany (and in fact in most countries visited by Napoleon) the broad 'as-is' disclaimer generally is quite an issue.
In the US it is very normal that two 'grown up' parties agree to something fair reaching; such as waiving certain rights or liabilties with respect of each other.
In most of (continental) europe that is not quite so easy; and the contract or agreemnt which two parties may have with each other may simply be overclassed by national law or 'common sense' in that respect.
The national law dictates that there are certain minimal levels and that disowning it all is simply not an option.
So regardless of what the developer (dis)claims with respect to warranty; the court may well held him liable to a certain extend.
At the same time, there is also a bright side; those liabilities are generally much more limited and 'capped' than in the US; and hardly ever exceed a small multiple of the resonable sum/economic value of the good (and not what can be done with that good). And they also put very reasonable demands on the 'user'. Willy nilly risking 5 million of lost production on a bit of untested free software is not going to ring true with the judge. He expects (more) resonable caution than generally in the US.
Also note that the scope of damages is very propotional to the purchace/gains of the developer/transaction. Sor 'free' (as in gratis') software those amounts are obviously not going to be very large.
Except if there is a bit more blame; i.e. someone knowingly dropping the ball. And unlike the US, where that waiver is going to help you - it may do little or nothing in most of Europe. Whereas in the US you are fairly secure.
On the other hand - any secondary damages issues are not nearly as much of a problem in europe, and virtually unheard of. Plus bear in mind that cost recovery and legal assistance is on an entirely different level in germany compared to the US. This making the issue of frivolous lawsuits by a megacorp which cripples a small developer virtually unheard of in most of the EU.
So in short; yes - you are bit more open to exposure in Europe - but as long as you behave resonably and are not vandalizing the hight of that exposure is very limited; and proportional to your fairly direct and clear cut gains from that software. And with open source / gratis - that is not going to make you go bust.
Dw
And it once again sees removed a large chunk of friviolous, US inspired, type of lawsuits we can just do without, from the general 'speak your mind on the net' equasion. Not to mention the fact that it equals the playing field between the big guys and the small guys. So good for the small guys; and tough, but fair to the big guys.
As to verifying the authentity; that it is really up to the publisher if he or she wants to limit the amounth of counter argument it wants to publish.
Dw
You'd be surprized to learn how seriously some groups in the *BSD world and for example in the Apache Software Foundation take those things. And indeed at the ASF we go through the legal paperwork to make sure that every commit is accounted for, every change monitored. Every code donation is matched with the right paperwork and grant documents. And while it is certainly true that a single 'big company' may not be able to guarantee such checks and balances - a crowd or whole ecosystem does. In a very effective way. Ultimately I think we will all find that this is what is needed to keep this "open source' world healty and viable. It is just that perhaps the FreeBSD crowd found this out earlier, due to the AT&T/Berkeley lawsuit. Dw
Actually - those elevend days are for the US - quite a few other countries, take Italy for example, have to wait for a year and a months longer.
Though that wire may be cheap - the actual connector and making sure the holes are in the right place of the casing; and making sure things are easy to assemble are not.
It is pretty common for 80% of the cost of something like temperature sensor with serial port, or a USB to FM radio, to be in the housing, assembly and the thoughdful steps to get the connectors mechanically sensible in place. And not with the electronics which does the sensible thing.
With bluetooth it is just some easy to place SMD stuff and a bit of firmware. Note of course that all is a compromize; because now you have a battery problem (and perhaps an edge connector for that) and you need some UI to pair the device to the right other endpoint.
But bluetooth is a useful element in the compromize game to make things cheaper.
Dw.
Hence equipment needs to be told about it in real time. Or in other word; you cannot build something which will work for 10 years and 'does the right thing' without user intervention.
A lot of people use UTC as if it is TAI; i.e.. they take two timestamps in UTC, substract them and expect that the difference between them is the number of seconds elapsed between those two. Whereas in fact one would need to know if any additional TAI/UTC seconds where added or removed in that particular period.
It seems that a lot of the objection against UTC leap second is coming from people who use it as if it was a TAI time; i.e. a predictable monotomic increasing value. And UTC has been that for many years; as we needed few leap seconds.
Unix has the same issue; we use UTC in a lot of places; which is fine for crontab's which need to happen at time x or y - yet calculate differences/delays as (x-y) ; which really is wrong; we should use TAI for that.
So parhaps it is time to start using TAI and calculate UTC on the fly based on the offset (today that is 32 seconds). Like /usr/share/timezone this would require a table with the dates/times (in TAI) at which delta's where introduced with respect to UTC.
Dw.
This is an esspecially important, not to say effective, mechanism which prevents the sort of litigation-wear-down so commonly seen in the US. Which skews the system towards those with the deepest pockets.
And more than once has a judge in Europe added something extra because the justitial apparatus wanted to give the signal that it was seriously annoyed by such superfluous lawsuits with no merit.
Dw.
Ideally one would use something like the WMSClient as mentioned on this page.
Dw
Specific things which are 'killer apps' to me are the ease by which you can configure things like blocks/groups of ports, services, machines - and then re-use those in your cluster config or wider rules.
This has safed us many hours of debugging downtime; and allows a much wider new range of 'new sysadmins to be' to deal with the simpler requests; such as adding a machine, service or port,
As it is fairly safe to modify the right table/macros/lists - without having to have a complete understanding of every detail. And yet, when designed by an experienced hand, robust enough not to be bitten by an overly broad side effect.
Have a look at http://www.openbsd.org/cgi-bin/cvsweb/~checkout~/s rc/etc/pf.conf?rev=1.19&content-type=text/plai n - to see good examples.
Dw.
How about simply stripping apache down; i.e. just run those two or three modules you really need. On for example a static images only server you could get away with just a logging module and the mmap module quite nicely. (Though realistically alias/rewrite is usually needed (or lots of symlinks) when your friendly marketing staff barges in with yet-another-campain which breaks all historic links.
Its modular - go play with it - and have lots of fun. (And yes - you can actually run apache completely and validly with just 2 lines of config.
Dw.
Hey - toss me a bone here :-) this is open source, tell me/us what annotations should be added to httpd.conf.. and we'll make apache a beter product.
Action >> Reaction ;-)
Dw.