Oh yes, I forgot to mention that SOAP currently uses HTTP as the underlying transport protocol. By using POST only, it limits the usefulness of those resources to the rest of the HTTP World Wide Web.
That's why SOAP 1.2 tries to allow the use of other HTTP methods where appropriate, so that ideally services using SOAP 1.2 over HTTP will be fully-fledged HTTP resources and will fit into the Web world.
Just how do you submit a complex transactional XML document using REST? Parsing a URL query string with 200 name-value pairs, then assembling that back into the XML document that the recipient expects does not seem like a good solution.
The above shows just how little jnana understands REST. Note that SOAP over HTTP uses POST and does not do any URL encoding of the requests. Just why can't a simple HTTP application do the same?
HTTP GET is to be used for retrieval only, and all URL encoding that may happen is only to identify the particular resource. In a well designed REST application no complex data structures need to go into the GET URL. For submissions to processing, there is the POST method.
The real difference between SOAP and REST is that currently the only protocol designed around REST is an application protocol for hypermedia transfer whereas SOAP is a general protocol upon which application protocols may be built, including RESTful ones. One of the protocols already built over SOAP is the RESTless RPC.
Disclosure: I'm a member of the W3C working group that's just now finishing SOAP 1.2.
In many systems, there is no concept of the almighty root, but there is the administrator. IIRC VMS has limited-rights administrator (dunno about Windows). This is a good thing in environments where you can't/mustn't fully trust even the administrators.
They should have chosen different names - they should have asked Adams and Eves. I don't know how the percentage would be affected, or the resulting numbers, but it would certainly be slightly more impressive. 8-)
Oh, and it would include women scientists, too. 8-)
I suggest you first try to get one of the relevant standards bodies (this could be ECMA, W3C, OASIS, IETF or others) to get this upon them - they would probably create a group, they already have processes to do that.
Should that fail, just look at their processes and learn from them when you're creating your group.
My experience is with W3C, where the process is basically the following (with some rewording):
Call for Participation (as public as possible, sent directly to those you think might be interested)
Charter (or should that be before the previous step?)
Requirements
Periodic public drafts
Last Call for Comments Draft (gathering as much consensus as possible)
Candidate Spec (gathering implementation evidence and comments)
Spec
W3C's Proposed Rec is mostly for approval by the Advisory Committee and the director, so this might be unnecessary.
Anyhow, you must consider the current open alternatives (I've notice XUL mentioned in other comments, I don't know the relevance myself) and decide if you want to improve or ratify one of them (based on requirements) or if you want to merge them into something new.
Only the last sentence that you quoted wasn't meant seriously.
Look, I know PCs are cheap and fast and usually do the job, but Talisman was talking about the quad-proc niche of cad/cam, video editing, 3d modeling, rendering and compiling crowd. In some of these, mostly just great CPU speeds are needed, that's where clustering helps. In other apps, it's not only the CPU speed but also the overall system throughput (see SCSI, switched IO etc.), and beowulf with the relatively slow LAN connections will be of limited help.
It's a niche, therefore it's going to be pricy. SUN, SGI are not dying because there is no need for the machines they build, they are dying because they (the companies) are too big for the niche.
I wonder why not consider some real hardware. See SUN Microsystems, Silicon Graphics etc.
I know a quad AMD would be considerably cheaper, but what about the failure rate? And, you know, CPU speed is not everything. Does nobody care any more? 8-)
did you mean perpendicular instead of parallel, then? 8-)
But in any case, while vibrations almost perpendicular to both the axes would not interfere, that's only a very small percentage of the vibrations, and how many vibrations do you get coming from the air above or from the center of the earth? Remember, you've got the installation basically on the ground and most vibrations travel along the ground.
The installations must have (as the article hints) excellent insulation from vibrations like trucks (there's no lorries in the USA 8-) ) and earthquakes. I wonder how exactly they are doing this - what kind of technology can be used to hold two things 4km away at precisely (give or take a few thousandths of the width of a proton) the same relative position all the time.
In Open Source, one usually has an issue with the software (like a bug or there is something I want it doesn't do) so one fixes it or logs it in a bugzilla.
In closed source (usually commercial sw) one thinks how to sell the next version, i.e. what to add to the features list. Features sell.
Conclusion: OSS developers don't have the drive to come up with features. No wonder OSS doesn't seem to innovate so much.
that's the best thought-out analysis of why we should move to base 12 I've ever seen. 8-)
Except that the number that's no longer nice is not 10 but Q, which never was a nice number anyhow.
Cool! Let's make our robots and AIs use base 12 (and convert in communication with us) so that when they take over they will be OK. 8-) People will never be able to go to base 12.
I've heard the Boston Central Artery (aka Big Dig) compared to a heart surgery performed on a marathon runner while he is running marathon. I wonder if the robot could be used in the Dig, too. 8-)
Slashdot Mirror
Has nobody really thought of Natalie Portman in Strip Poker? 8-)
That's why SOAP 1.2 tries to allow the use of other HTTP methods where appropriate, so that ideally services using SOAP 1.2 over HTTP will be fully-fledged HTTP resources and will fit into the Web world.
The above shows just how little jnana understands REST. Note that SOAP over HTTP uses POST and does not do any URL encoding of the requests. Just why can't a simple HTTP application do the same?
HTTP GET is to be used for retrieval only, and all URL encoding that may happen is only to identify the particular resource. In a well designed REST application no complex data structures need to go into the GET URL. For submissions to processing, there is the POST method.
The real difference between SOAP and REST is that currently the only protocol designed around REST is an application protocol for hypermedia transfer whereas SOAP is a general protocol upon which application protocols may be built, including RESTful ones. One of the protocols already built over SOAP is the RESTless RPC.
Disclosure: I'm a member of the W3C working group that's just now finishing SOAP 1.2.
I find my foreign music in my local music shop. That is, foreign to y'all insensitive clods! 8-)
Really, root is one of the disadvantages of UNIX.
Within a UNIX system, root is the God. What piece of this concept requires explaining? 8-)
did you mean "it removes the PCI barrelneck but doesn't remove the HDD bottleneck"? 8-)
Oh, and it would include women scientists, too. 8-)
Should that fail, just look at their processes and learn from them when you're creating your group.
My experience is with W3C, where the process is basically the following (with some rewording):
W3C's Proposed Rec is mostly for approval by the Advisory Committee and the director, so this might be unnecessary.
Anyhow, you must consider the current open alternatives (I've notice XUL mentioned in other comments, I don't know the relevance myself) and decide if you want to improve or ratify one of them (based on requirements) or if you want to merge them into something new.
Look, I know PCs are cheap and fast and usually do the job, but Talisman was talking about the quad-proc niche of cad/cam, video editing, 3d modeling, rendering and compiling crowd. In some of these, mostly just great CPU speeds are needed, that's where clustering helps. In other apps, it's not only the CPU speed but also the overall system throughput (see SCSI, switched IO etc.), and beowulf with the relatively slow LAN connections will be of limited help.
It's a niche, therefore it's going to be pricy. SUN, SGI are not dying because there is no need for the machines they build, they are dying because they (the companies) are too big for the niche.
I know a quad AMD would be considerably cheaper, but what about the failure rate? And, you know, CPU speed is not everything. Does nobody care any more? 8-)
But in any case, while vibrations almost perpendicular to both the axes would not interfere, that's only a very small percentage of the vibrations, and how many vibrations do you get coming from the air above or from the center of the earth? Remember, you've got the installation basically on the ground and most vibrations travel along the ground.
And how would one go about making a 4km-long thing rigid, even if only in one axis?
Any reader here who can explain?
Weeerl. In my endless idealism I assumed that the submitters read what they submit, even if I have long ago stopped believing the editors read it. 8-)
I thought the summary in the /. posting might say at least a word about that. 8-)
Anyhow, nice metallic cars could also eliminate auto painting. Who needs a painted car anyway? 8-)
Note that SuperBlack is a bit on the bluish-pinkish side.
Nice number. But... exactly how many bodily functions are there? 8-)
In Open Source, one usually has an issue with the software (like a bug or there is something I want it doesn't do) so one fixes it or logs it in a bugzilla.
In closed source (usually commercial sw) one thinks how to sell the next version, i.e. what to add to the features list. Features sell.
Conclusion: OSS developers don't have the drive to come up with features. No wonder OSS doesn't seem to innovate so much.
Except that the number that's no longer nice is not 10 but Q, which never was a nice number anyhow.
Cool! Let's make our robots and AIs use base 12 (and convert in communication with us) so that when they take over they will be OK. 8-) People will never be able to go to base 12.
Because XML is the fashionable acronym now.
true.<and false/> />
<thing foo=5/>.<equal <thing foo=5/>
object.<get key lookup=false if_missing='return' default=false/>
I've seen a language that works with XML as a first-class data structure, but it had a normal C-like syntax, say
x = <foo bar="3"/>;
if (x.bar == 3) {...};
IMHO the latter makes much more sense - you can work with XML and you don't have to learn a whole new LISPy hybrid syntax.
I've heard the Boston Central Artery (aka Big Dig) compared to a heart surgery performed on a marathon runner while he is running marathon. I wonder if the robot could be used in the Dig, too. 8-)