INFO-VAX Sun, 09 Sep 2007 Volume 2007 : Issue 491 Contents: Re: I think I might understand N3 and the bullet Re: SOAP, WSIT, I'm LOST, sort of... Re: SOAP, WSIT, I'm LOST, sort of... RE: [INFO-VAX] I think I might understand N3 and the bullet ---------------------------------------------------------------------- Date: Sat, 08 Sep 2007 15:02:51 -0700 From: Doug Phillips Subject: Re: I think I might understand N3 and the bullet Message-ID: <1189288971.256674.291310@o80g2000hse.googlegroups.com> On Sep 8, 1:31 pm, "Paul Raulerson" wrote: > > -----Original Message----- > > From: INFO-VAX Discussion [mailto:INFO-...@LISTSERV.UGA.EDU] On Behalf > > Of Doug Phillips > > Sent: Saturday, September 08, 2007 12:06 PM > > To: INFO-...@LISTSERV.UGA.EDU > > Subject: Re: [INFO-VAX] I think I might understand N3 and the bullet > > > On Sep 7, 10:14 pm, Ron Johnson wrote: > > > Hmmm. I think just had an insight!!! > > > > As a (fast moving) gas molecule strikes the bullet, there is an N3 > > > reaction. Bullet goes one way, gas molecule goes the other. > > > Billions of such continuous tiny reactions: > > > > First overcome it's static inertia, then accelerate it down the > > > barrel, imparting more and more kinetic inertia as it zooms down the > > > barrel. > > > > As it zooms thru the fluid medium (air, usually), it strikes against > > > other molecules. These opposing reactions are what is, in > > > aggregate, known as friction. > > > > Am I still totally off-base? > > > Newton's laws are not just about "motion" even though they carry that > > name, but are about the relationship between forces acting upon > > objects. Other dead scientists expanded upon those ideas to describe > > the interaction of force fields and such. > > > An N3 reaction happens when any two (or more) masses or forces > > interact with each other. The explosion within the chamber (or the > > cartridge contained by the chamber) reacts with its entire > > surroundings. You describe what happens to the bullet, whose reaction > > is most obvious. > > > The relationship between energy, mass and velocity along with other > > factors, many you've already mentioned, will determine the outcome of > > the interactions. > > Your really talking about calculating linear response functions here, in > some kind of a time dependent domain, right? > I suppose we might be, but that's getting dangerously close to falling back into a QM discussion:-) I'd say the parent thread of this one was working around the edges of stochastic and probability theory, though, so it might come up there. Anyway, my practical exposure to LRF was limited to the electronics world and taking it into a gun chamber might force me to actually open a book (other than the stack of sci-fan that a friend loaned to me for the summer and which I'm determined to finish before the snow falls.) > I'm probably *way* out of date with this, but wasn't one of the classic > problems with doing this that you would up having to store really *huge* > state vectors - usually too large for main storage? > I haven't kept up on it either, but that's my recollection, too. > Seems to me there was a clever method that used polynomials to do it, but > even that was a nightmare. I remember having to dig back into Matrix Algebra > just to understand what it was doing. > > Point is, we actually looked at computing that on a PDP-11/45, but quickly > determined that didn't have enough > go-juice to do the job. I always wondered if it could be done on a VAX- I'm > sure Alpha and Itaniums could do the job. :) > I'd wonder if maybe someplace one or two aren't actually doing it. Looking back (which I was forced to do when I cleaned out 30 years of accumulated printouts, disks, tapes and manuals from the office) I'm amazed at what we could do with so little resource. One of my first jobs as a programmer/operator was running a bill-of- materials and production forecast through a mainframe -- using punch- cards -- drawers full of punch-cards. Ran most of the night and output drawers full of new cards. I think I only dropped a stack once (maybe twice) when I got a bit too daring. There was a halt about mid-way through the run and you had to dial in an instruction to jump back into the code. You got to where you could tell what the machine was doing just by listening to it. Woke up from some strange nightmares a few times back then. The next system I worked with actually had *disk* -- Wow!:-) Ah, the "good old days" (not!) ------------------------------ Date: Sat, 08 Sep 2007 14:18:54 -0500 From: Michael Anderson Subject: Re: SOAP, WSIT, I'm LOST, sort of... Message-ID: <6ss5e3h5koevc2r1ue9b6t0du6jfn6fqdc@4ax.com> On Sat, 08 Sep 2007 05:42:16 -0700, ja wrote: >Jan-Erik, > >-- gSOAP, which Bob Gezelter mentioned, is probably the way for you to >go as 1) it is a full-blown Web Service implementation offering both >client and server functionality, i.e., VMS acts both as client and >server whereby WSIT only acts as a server; 2) it is written entirely >in 'C', thus fitting nicely into your current environment; 3) it runs >on Alpha and Integrity. It would not be hard at all to integrate it >into OSU or, use the mod_gsoap module for Apache which we have just >finished porting. > >-- Windows Servers, or more correctly put, Microsoft, has plenty of >software to make the use of Web Services more or less transparent to >the user and even developer, be it from MS Office, a program, or >whatever. > >-- the Java environment offers many ways of processing Web Service >requests > >-- as do the 'P' (Perl, Python, PHP) languages > >-- the reason there is not support for the OBJ2IDL.EXE utility on >Alpha is, as Bob already pointed out, due to the fact that the format >of an object file on Alpha is not standard and therefore very hard to >use to obtain the interfaces of the compiled program. On Integrity the >format is well documented and it is therefore easy to parse the file >and obtain the interfaces. It should be pointed out that the result of >OBJ2IDL.EXE on Integrity can be used equally well on Alpha > >Cheers, John >PS: despite my Martinelli car parked in the HP car park, please feel >free to contact me directly as I do, occasionally, actually try and do >productive work:-) > >On Sep 7, 4:43 pm, Jan-Erik Söderholm >wrote: >> Jeffrey H. Coffield wrote: >> > Jan-Erik Söderholm wrote: >> >> >> To add the the picture, this "new" interface is ment >> >> to complement the current that is a simple mail based >> >> communication. The "other side" simply sends a mail >> >> to a specific user, using an agreed on subject with a >> >> datafile ZIP'ed and attachede using standrad MIME format. >> >> On the VMS system ("my" system) this mail is taken care >> >> of using DELIVER/MPACK/MUNPACK/UNZIP and the datafile >> >> is finely FTP'ed over to an IBM mainframe to be stored >> >> into a central DB2 database. The mainframe interface is >> >> not the target right now. >> >> >> Now, some users would like to have an alternative to >> >> this mail based communication. FTP has been discussed. >> >> WEB Services was also mentioned. And that's why I was >> >> asking. >> >> >> OK, I have to dig a little more into this. >> >> As you said, one of the problems with open source >> >> stuff is that it might be hard to find a consistant >> >> set of documentation... >> >> >> Jan-Erik. >> >> > If the "other side" is a person, why not put up a web page with a file >> > upload? If the "other side" is a program, then some change would have to >> > be made since it currently sends a e-mail. That can usually be switched >> > to cURL to do the post. The web page on your side is the same. >> >> > Jeff Coffield >> >> No, the other side are different client application. And they'd >> like to use the latest busniess buzz-words, of course. Whatever >> that looks good in the data sheets. :-) And it should be a fully >> automated setup. No manual intervention in any part. >> >> So, the *BEST* solution for *me*, is one where the client can use >> WEB services (or whatever "new" tools they like) and I can >> continue with the same hardware/VMS/OSU setup. >> >> I do not now what tools there are on Windows servers, but WEB Services >> was mentioned as one that they'd look at as interesting. >> >> Now, I guess I need some XML tool to receive and decode whetever >> the WEB Servicesa/SOAP tools at the client creates, right ? >> >> Jan-Erik.- Hide quoted text - >> >> - Show quoted text - > I strongly recommend gSOAP which you can use from C or C++ cURL also has it uses on OpenVMS but for non SOAP related web stuff. SOAP messages when you start using things like WS-Security etc. becomes impossible to generate by hand you must use tool kits like gSOAP. With gSOAP you can do both client and server stuff with out a lot of pain. gSOAP also is really high performance much better than any java based stuff further you do not need to load up your system with the JVM etc. and try and get this running. If you need to distribute software to clients you only need send a single EXE and not require tons of layered products be loaded. In fact although gSOAP requires OpenSSL when using HTTPS you can statically link in OpenSSL and avoid loading OpenSSL on the customer system. ------------------------------ Date: Sat, 08 Sep 2007 21:36:27 GMT From: =?ISO-8859-1?Q?Jan-Erik_S=F6derholm?= Subject: Re: SOAP, WSIT, I'm LOST, sort of... Message-ID: Michael Anderson wrote: > On Sat, 08 Sep 2007 05:42:16 -0700, ja wrote: > >> Jan-Erik, >> >> -- gSOAP, which Bob Gezelter mentioned, is probably the way for you to >> go as 1) it is a full-blown Web Service implementation offering both >> client and server functionality, i.e., VMS acts both as client and >> server whereby WSIT only acts as a server; 2) it is written entirely >> in 'C', thus fitting nicely into your current environment; 3) it runs >> on Alpha and Integrity. It would not be hard at all to integrate it >> into OSU or, use the mod_gsoap module for Apache which we have just >> finished porting. >> >> -- Windows Servers, or more correctly put, Microsoft, has plenty of >> software to make the use of Web Services more or less transparent to >> the user and even developer, be it from MS Office, a program, or >> whatever. >> >> -- the Java environment offers many ways of processing Web Service >> requests >> >> -- as do the 'P' (Perl, Python, PHP) languages >> >> -- the reason there is not support for the OBJ2IDL.EXE utility on >> Alpha is, as Bob already pointed out, due to the fact that the format >> of an object file on Alpha is not standard and therefore very hard to >> use to obtain the interfaces of the compiled program. On Integrity the >> format is well documented and it is therefore easy to parse the file >> and obtain the interfaces. It should be pointed out that the result of >> OBJ2IDL.EXE on Integrity can be used equally well on Alpha >> >> Cheers, John >> PS: despite my Martinelli car parked in the HP car park, please feel >> free to contact me directly as I do, occasionally, actually try and do >> productive work:-) >> >> On Sep 7, 4:43 pm, Jan-Erik Söderholm >> wrote: >>> Jeffrey H. Coffield wrote: >>>> Jan-Erik Söderholm wrote: >>>>> To add the the picture, this "new" interface is ment >>>>> to complement the current that is a simple mail based >>>>> communication. The "other side" simply sends a mail >>>>> to a specific user, using an agreed on subject with a >>>>> datafile ZIP'ed and attachede using standrad MIME format. >>>>> On the VMS system ("my" system) this mail is taken care >>>>> of using DELIVER/MPACK/MUNPACK/UNZIP and the datafile >>>>> is finely FTP'ed over to an IBM mainframe to be stored >>>>> into a central DB2 database. The mainframe interface is >>>>> not the target right now. >>>>> Now, some users would like to have an alternative to >>>>> this mail based communication. FTP has been discussed. >>>>> WEB Services was also mentioned. And that's why I was >>>>> asking. >>>>> OK, I have to dig a little more into this. >>>>> As you said, one of the problems with open source >>>>> stuff is that it might be hard to find a consistant >>>>> set of documentation... >>>>> Jan-Erik. >>>> If the "other side" is a person, why not put up a web page with a file >>>> upload? If the "other side" is a program, then some change would have to >>>> be made since it currently sends a e-mail. That can usually be switched >>>> to cURL to do the post. The web page on your side is the same. >>>> Jeff Coffield >>> No, the other side are different client application. And they'd >>> like to use the latest busniess buzz-words, of course. Whatever >>> that looks good in the data sheets. :-) And it should be a fully >>> automated setup. No manual intervention in any part. >>> >>> So, the *BEST* solution for *me*, is one where the client can use >>> WEB services (or whatever "new" tools they like) and I can >>> continue with the same hardware/VMS/OSU setup. >>> >>> I do not now what tools there are on Windows servers, but WEB Services >>> was mentioned as one that they'd look at as interesting. >>> >>> Now, I guess I need some XML tool to receive and decode whetever >>> the WEB Servicesa/SOAP tools at the client creates, right ? >>> >>> Jan-Erik.- Hide quoted text - >>> >>> - Show quoted text - > > I strongly recommend gSOAP which you can use from C or C++ cURL also > has it uses on OpenVMS but for non SOAP related web stuff. > > SOAP messages when you start using things like WS-Security etc. > becomes impossible to generate by hand you must use tool kits like > gSOAP. With gSOAP you can do both client and server stuff with out a > lot of pain. > > gSOAP also is really high performance much better than any java based > stuff further you do not need to load up your system with the JVM etc. > and try and get this running. If you need to distribute software to > clients you only need send a single EXE and not require tons of > layered products be loaded. In fact although gSOAP requires OpenSSL > when using HTTPS you can statically link in OpenSSL and avoid loading > OpenSSL on the customer system. Right, everything sounds just perfect. Just missing a good link to a kit for VMS now. Yes, I have googled around, but found nothing specificaly speaking about VMS. Jan-Erik. ------------------------------ Date: Sat, 8 Sep 2007 13:31:41 -0500 From: "Paul Raulerson" Subject: RE: [INFO-VAX] I think I might understand N3 and the bullet Message-ID: <001301c7f246$80cfe810$826fb830$@com> > -----Original Message----- > From: INFO-VAX Discussion [mailto:INFO-VAX@LISTSERV.UGA.EDU] On Behalf > Of Doug Phillips > Sent: Saturday, September 08, 2007 12:06 PM > To: INFO-VAX@LISTSERV.UGA.EDU > Subject: Re: [INFO-VAX] I think I might understand N3 and the bullet > > On Sep 7, 10:14 pm, Ron Johnson wrote: > > Hmmm. I think just had an insight!!! > > > > As a (fast moving) gas molecule strikes the bullet, there is an N3 > > reaction. Bullet goes one way, gas molecule goes the other. > > Billions of such continuous tiny reactions: > > > > First overcome it's static inertia, then accelerate it down the > > barrel, imparting more and more kinetic inertia as it zooms down the > > barrel. > > > > As it zooms thru the fluid medium (air, usually), it strikes against > > other molecules. These opposing reactions are what is, in > > aggregate, known as friction. > > > > Am I still totally off-base? > > > > Newton's laws are not just about "motion" even though they carry that > name, but are about the relationship between forces acting upon > objects. Other dead scientists expanded upon those ideas to describe > the interaction of force fields and such. > > An N3 reaction happens when any two (or more) masses or forces > interact with each other. The explosion within the chamber (or the > cartridge contained by the chamber) reacts with its entire > surroundings. You describe what happens to the bullet, whose reaction > is most obvious. > > The relationship between energy, mass and velocity along with other > factors, many you've already mentioned, will determine the outcome of > the interactions. Your really talking about calculating linear response functions here, in some kind of a time dependent domain, right? I'm probably *way* out of date with this, but wasn't one of the classic problems with doing this that you would up having to store really *huge* state vectors - usually too large for main storage? Seems to me there was a clever method that used polynomials to do it, but even that was a nightmare. I remember having to dig back into Matrix Algebra just to understand what it was doing. Point is, we actually looked at computing that on a PDP-11/45, but quickly determined that didn't have enough go-juice to do the job. I always wondered if it could be done on a VAX- I'm sure Alpha and Itaniums could do the job. :) -Paul ------------------------------ End of INFO-VAX 2007.491 ************************