That’s most unfair.

He has no clue whatsoever in terms of the bytecode, but he might very well be a whizz when it comes to building his own Java compiler.

I mean, seriously, let’s get real here. This is not a blog where ad hominems are welcome.

The key purpose of mainframes is not to handle volumous batch processes – that’s the job of commodity machine clusters – but complex DBMS transactions:

http://www-01.ibm.com/software/htp/cics/tserver/

Of course, why use cheap, off-the-shelf machine clusters for batch processing documents when you can waste extra millions of dollars on big irons with no obvious benefits whatever? That doesn’t sound utterly absurd to anyone at all!

Of course if the documents you are produced on the mainframe cannot be opened by where they are being sent you have wasted time.

Given your sketchy knowledge in something as simple as generics, I am not surprised at all that you are displaying such feeble grasp in parallel computing.

You stitch together your cluster, feed your documents to the server node and then let the server node return the results back to you or send them off to where they are supposed to go. That’s just one way of getting this done without needing anything more than commodity Xeon-based machines.

Seriously, even an average CS student can easily figure out something this simple.

This your problem. There are quite a few that have no other option. Turns out Libreoffice UNO interface is the same as StarOffice 5.2. So is very well documented the interface as only been extended. So what you wrote for 5.2 works today.

oldman no everyone is using vCenter or even big systems that contain windows. Its call the next best thing. If you could not have MS Office and it was not an option what would you use Oldman. I think you would find yourself using Libreoffice like everyone else.

The reality not everyone has you option of even using MS Office to generate reports.

By the way oldman you only need two basic command line.
Network
soffice “–accept=socket,host=localhost,port=8100;urp;StarOffice.ServiceManager” -nologo -headless
Named pipe
soffice “–accept=pipe,name=foo;urp;StarOffice.ServiceManager” -nologo -headless

Fun part there is a –unaccept so you can close links 1 by 1.

Something like that. To tell it to start in service mode after that you talk to it and control it by UNO.

Those command lines is valid on StarOffice, Apache Open Office and Libreoffice.

Reality here oldman what you do by COM is replaced by UNO operations. UNO is network aware so you can have many different computers in a cluster send results directly back to a single instance exposed on the internal network or by encrypted tunnels.

I am sorry oldman when it comes doing some-things dcom with MS Office does out cut it.

oldman the serous fact is LibreOffice and what it descended from are server products. Able to cope with editing from many sources at the same time.

There are times that MS Office is trying to put a round peg into a square hole. Yes the square hole that libreoffice fits perfectly todo.

Even more scary is the fact that you can have the Libreoffice interface open while its receiving data by the network interface updating in real time. The headless flag is so it does not render the interface.

Oldman how do you debug yours when you have a error. I can simply remove a flag and watch mine.

You will find yours does not exactly scale. But you have to admit at times excel documents or ods documents or other office documents are highly useful.

Oldman the simple reality you are too stubborn to try out the other path. If you do you will find out there are some-things Libreoffice is particularly good for. As a desktop office suite still has a distance to go. As a server to generate documents is very powerful.

Going by powershell is possible by the CLI-UNO interface. Really since that is windows only I mostly would not bother with the powershell route.

vcenter python bindings plus python uno bindings is basically perfect. Of course in some cases throw in pyssh to allow port tunnelling.

oldman you reach for powershell and MS Office I reach for python and libreoffice. Both generate the same results in the end other than the fact I can run mine on cloud servers or where ever I need without worrying about if MS Office is there or not. Heck I don’t even have to worry about Libreoffice being there since I can tunnel over network if required to where ever libreoffice is.

Libreoffice as server is not that much different to a mysql server or some other server. 1 instance can service many servers at times.

Powershell tunnelling is not particularly fun.

There is some advantage to a network controllable office suite.

Oldman sometime you need to step past Libreoffice GUI and just control the backend for a bit and start to understand how the beast fits in.

There were 152 people responding to the survey, according to the cite, so 4.5 people said that at their company it was “supported”, apparently meaning that someone, somewhere was using it for business. I guess one person was not so sure or else there is a round off error.

A couple of things come to mind here, one being that Star Office is on a par with both Lotus and Google’s SaaS efforts in terms of results, with is a sort of shameful condition for Lotus and Google. Another is that this is a rather abysmal showing for the FLOSS entries, particularly with the realization that the people surveyed were IT managers who could be expected to be much more amenable to using open source products.

It is like your relying on W3Schools statistics wherein Linux shows marginally better than on other, less techie biased samples. You suggest that shows Linux is more common than the conventional thinking but I think that it only shows that it cannot show well even on its home ground.

Drop your delusion of grandeur – if any at all, it makes you look even more stupid than you already are.”

In fact Brillo this shows your lack of competence.

What will big iron be handling. Huge data sets at times. You do not wish to transfer these back to the desktop just to generate reports so you generate mainframe side.

Mainframes do a lot of batch processing mostly because it makes sense to. User interaction is highly lacking effectiveness.

The reality here is neither of you have worked with Mainframes so you think the idea of a Office suite on a Mainframe is stupid. A graphical Office suite you are correct. A office suite that can be controlled for batch production of results so avoiding having to send the dataset out of the mainframe has many advantages.

Of course if the documents you are produced on the mainframe cannot be opened by where they are being sent you have wasted time. This is why the means to have the same office suite on mainframe and desktop is so useful.

Not every mainframe workload is going to use this. You have spent big money to have a mainframe you want it todo as much as possible.

You need to produce documents people can open of results from processing on mainframes. Sending that data back to the client machines not a good option. Particularly when that data is quite a few G in size.

ODF and MS formats do have an advantage over html. They are single files even when containing images.

Office suite on mainframe that can batch process producing new documents and convert documents to multi format for sent out can be useful. Ie convert result to PDF and email auto to client send ODF source document internal to staff.

The mainframe does not have to send data back to desktop to email, im or place on web results.

Limiting office suite to desktop adds extra step in process.

Brillo the common mistake is we can run a windows machine todo this. Big datasets equals large amounts of ram. So now you end up having to use a middle ground format. Every extra step is a extra risk something will screw up.

Brillo and Clarence Moon you would have stayed well clear. Oldman knows this topic as well. He is not going to jump in here because he knows the reality. Since he does some of it. Reports have to be generated this is part of business. Don’t break the network moving around data to make reports.

Mr. O doesn’t know the meaning of “stupid”. Try “bogan” instead. According to Google, that would be in his claimed native language.

Nope. I use COM object automation and Powershell/PowerCLI to drive excel to create personal ad-hoc reports out of vCenter.

What The Hamster has described is a set of esoteric IMHO poorly documented command line switches that apparently allow libermanoffice to batch process files.

There are plenty of vCenter administrators who use my technique.

I’m willing to bet there are far fewer who leverage this geek crap.

Still a batch process. Still a waste of a big iron. I am sorry but you just aren’t fit for the adjective “competent”.

Drop your delusion of grandeur – if any at all, it makes you look even more stupid than you already are.

There is ZERO guarantee that this particular but of alpha code will go anywhere. Assuming the project survives infant mortality(more debugging), it then has to be deemed useful enough to be picked and fully supported by one of the major distributions(mode debugging), THEN it has to make it into one for the commercial distributions(more debugging), AND THEN it has to be vetted by individual business. SO what we have here Mr. Hamster, is another of your flights of idle speculation expressed as if it were a certainty.

Which it most certainly is not.

“In fact you have failed todo your homework on what the server mode of Libreoffice allows.”

It seems to me that there is not homework to do here Hamster. Once again you point to some bit of badly documented (it is referred to in one of the references as a “hidden feature”) geek esoterica. How many people use this feature sir? Few if any (beyond possibly yourself) I’ll bet!

Regardless, the presence of so called batch processing functions does not even begin to make up for the missing standard function and feature in LiebermanOffice, functions that people are more likely to use than they are this.

In fact using the UNO interface can do more than batch conversion. You can create complex documents from basically nothing as part of back-end.

In fact you have failed todo your homework on what the server mode of Libreoffice allows. Batch conversion is only one part. Creating of excel word powerpoint straight from data sets. The list goes on. So all the charting of Libreoffice can be exploited by UNO interface in fact. Yes you can have it drop a chart as a image.

Read what I said more carefully Brillo.
“documentation conversion and production”

Documentation conversion is batch conversion.

Production is creating a document from nothing other than template, data, scripts. You do not need a GUI to create documents. Basically Libreoffice can be used the same way Oldman does with MS Office to auto-create operational reports.

Having an office suite on the mainframe allows you it spit out conventional desktop formats. Even better have the same office suite on the desktop to read them.