Slashdot Mirror
IBM To Publish Java Office Suite
prostoalex writes "The Big Blue will bundle J2EE-based word processor, spreadsheet, and presentation graphics applications in its WebSphere portal. What's more interesting is that the package is server-side, with functionality of the application being delivered to the user over the network. Both CRN (linked above) and The Register considered that a major move against MSFT."
Subject speaks of the comment.
http://saveie6.com/
Good job, IBM! It's always nice to see an industry leader promoting competition in the software arena. It forces all candidates to develop a better product. Perhaps this will spark some ideas with M$? (Not that they'll be any good, but still... ;-) )
Free, 100% Free, when you buy our $580,000 WebSphere Portal software/server combo!
-
http://fink.sourceforge.net/
How does making the application server side help in office applications? These are inherently GUI based applications with the focus on WYSIWYG. Why J2EE, all J2EE would do is save the files?
For all practical purposes when I'm at work and the network is down..so am I. That's where the work is stored, that's where we access important information, that's how we communicate.
Even with desktop apps, when the network goes down - we're doing nothing.
It was slow, and depending on which JAVA VM you used depended on how long it would go before crashing.
It will be remembered as one of the many wasted efforts from the java-craze years.
Never overestimate the end user. -jeramy b. smith
I think that people are going to notice that there are other office suites out there... Especially if it comes bundled with the server. I know plenty of people who are tired of the way you have to pay $230+ for MS Office per machine. This definitely is a threat to MS.
Also, I wonder if they will adopt an existing file format, or if they are just going to go on thier own. I would think that people would like it much better, and would be less hesitant to switch to it, if they didn't have to hassle with thier file formats...
I dont think this is quite the same thing. IBM have invested heavily in their Websphere app server technology and this just looks like an attempt to squeeze Weblogic market share. Its more evidence that IBM see Webspere as a platform in and of itself.
It has the convenient side effect of putting the Lotus code base to work instead of sitting around doing nothing and if its well received new markets open on the desktop for IBM.
Personally I dislike Websphere but I think this is very clever idea which will go down well in the corporate sphere.
Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
What truth?
There is no dupe
If anything, it bodes well for Microsoft, because it gives them another solution to point to as a competitor, dispelling claims about MSOffice being a monopoly.
File formats? Compatible with Office? I doubt it. That means this thing is boat anchor. Either that, or it will join those thousands of boxes of the old Lotus Suite gathering dust in cabinets that companies got for buying Notes.
If there was an award for software distributed that never got used, nobody would ever beat IBM.
It's called the HTML editing scriptlet and I've personally (with a couple of lines of ASP code and some javascript) developed a fully customized web based word processor, with Word-like toolbars and icons and such. IE has this hidden feature that is basically an integrated HTML editor; the object just needs to be called. You then put the contents of the document into a POST operation to save (natively in HTML format).
I hear the Mozilla crew has finally been thinking about integrating this kick-ass feature. All in all, its integrated with the browser, so no shitty Java code to run in your VM.
IBM has basically released the source of their java applications. Java is *trivial* to decompile. Names of variables, function calls etc are kept intact - you only lose comments.
Don't believe me? Search google for "DJ's Java Decompiler"
If this is tightly integrated with WebSphere, I can see how it would be a benefit to those who have already deployed or decided to deploy WebSphere. But without said tight integration, there's really not much benefit over using OpenOffice or some other freely available office suite. Having the apps served via the network may make it easier to deploy updates, but I still don't believe the suite is going to induce more people to buy Websphere unless it's tightly integrated and truly exceptional relative to other free alternatives.
Java has come a long way since the early days of Core's suite. Performance isn't an issue anymore, *especially* with IBM's SWT toolkit, a blazing alternative to Swing. IMO, one problem Sun has had is not offering enough slick, Java desktop apps. Perhaps the slickest one of all is InstallAnywhere -- something everyone uses and appreciates the slickness of, but doesn't realize that slickness is in no small part due to being written in Java. Maybe this will help get the ball rolling.
Because when the network is down you are down, regardless.
Many business, like manufacturing, are so connected to computers that if the computer network goes down then business grinds to a halt.
I've worked in this type of place. It isn't as tough as you think, considering they only worked 2 shifts, leaving 8 hours for maintenance, etc. It isn't 24x7 but more like 16x5.
The biggest issues were desktop apps having problems. Amazing how much that stuff freezes and crashes when people *insist* on having Outlook, IE, Word, Excel, an SNA client (TN 3270) and possibly a CAD viewer (java applet) and maybe an MS Access database or two running all at once.
Believe it or not, Sun has the right idea. Build the network so that it is so reliable it makes the phone and power companies look like slackers. Then move 90+% of the apps back upstream to a professionally-managed & maintained server.
Learning HOW to think is more important than learning WHAT to think.
I suspect IBM will want to be able to import MS Office files into their system... perhaps they'll share some code with the OpenOffice gang.
The filters in OpenOffice are pretty good... but there's always going to be room for improvement. (plus, those MS file formats are a moving target...) It would be a nice bonus if IBM would open source their filters, or better yet, use the OpenOffice filters and contribute patches. I have no personal experience with recent Lotus packages, but I'm going to guess that OpenOffice filters are more advanced than the Lotus ones by now anyway.
Personally, I'd like to see some basic VBA compatibility... say what you want about VB, but I find it very handy for little custom functions in Excel -- and no, I don't want to rewrite them just to use oO.
Interesting co-opetition if this did happen. IBM working with a group largely supported by Sun, both trying to take a big bite out of MS.
"First it was dumb terminals then network computers and now this. Its dead give it up."
Do you have ANY idea how useful something like this could be to large environments? Where I work, we have 35,000 computers on the supported list. Two or three different platforms worth, PCs, Macs, and some occasional Linux machines. It would be kick ass if we could deploy one version of one productivity suite across the whole network, especially if we could do it with site based central servers rather than having to work on each and every PC on the fucking network.
If this supports server-side file storage, it's even better, since then we don't have to worry about user data any longer. We'll gladly build fault-tolerant servers if we only have to do it for about a hundred machines, and suddenly we can also roll out upgrades to the products with only a few days' work, not months like we currently have to.
The days of dumb terminals rocked. If one broke, we brought another one out, and swapped. If the server broke, we dropped everything and fixed it. Regardless, the user wasn't without a connection or machine for days at a time like which happens in the Windows world. If Microsoft hadn't managed to con everyone into believing that their dumbass standalone workstation idea was the best, we'd probably be using X-Terms now, and have even better centralization of critical data, rather than every user having to know how to copy their data to the network attached storage (and most of them are not interested in learning).
Just because a computing model is old doesn't mean that it's outdated.
Do not look into laser with remaining eye.
Think of servers as a fixed cost and clients as a variable cost. With thin client models you only have to support a browser on the client machines. It is the end to a management headache: all those apps configured on all those clients.
Think about all those companies that are paying big bucks for all of those client OS's and Apps. Now they can get, for less than $200, loaded PC's (1.1 GHz PC w/Linux installed, no monitor).
Walmarts $199.98 PC.
Open source development is my way of competing with the low-cost programmers in India...
Sure. They could have chosen to write it in Whitespace instead of Java.
We can neither love nor pity nor forgive. If you make a slip in handling us you die!
"Microsoft doesnt care."
But they should. Imagine if IBM does what IBM typically does well, which is deliver high-end computing in large-scale environments, with this product for users...
Large companies, school districts, government organizations, anywhere that has had computers longer than Microsoft has been in full force will be able to appreciate this. It's a support thing. If you can have a platform independent system that is centrally installed and highly available, you'll make it in evironments that have experience with IBM successes like AS/400's, System 370/390's, and RS/6000's, since these groups already trust IBM. In fact, companies that don't have IBM, because they purchased a cheaper competitor's computing platform, like an HP system or a Hitachi might be inclined to add this to their computing systems as well. They don't then have to go out to each PC when some dumbass library breaks, spending the significant amount of time necessary to fix Microsoft Office, they may have to go out and upgrade a web browser or java engine at the most. Then, they can do all of the product support and updates for the productivity suite in a localized area, and NOT have to pay Redmond the blood-money that they currently do for network-wide site licenses.
I'd SO go for it...
Do not look into laser with remaining eye.
It could be argued there are plenty of office suites already, especially word processors and spreadsheets. However, what this huge steaming pile of free and open source officeware lacks is a real alternative to MS Access. There simply is nothing, except for some half-assed iimitators that only run on Windows themselves.
What I'd like to see is something programmed in Java, using an embedded Java RDBMS engine such as McKoi, but also able to be used as a front end to any SQL database -- just like Access. The problem with Acces is, of course, that it only runs on Windows. Wouldn't it be groovy to have a cross-platform, true alternative?
Microsoft was a victim of sun's harh contracts
IIRC, the virtual machine debate was the fault of Microsoft for not upholding its end of a bargain.
Good troll.
We can neither love nor pity nor forgive. If you make a slip in handling us you die!
I am sure you're trolling, but anyway.
.Net runtime is a Microsoft product that only runs on Microsoft Windows and not on any of IBM's systems?
.Net runtime is slower than current Java runtimes. In fact, on non-Windows platforms the .Net runtime doesn't even exist!
Why did they choose to use java and not the faster and more modern C# ?
Because IBM are heavily into Java (and have VMs for all their platforms)? Because C# and the related
(Microsoft was a victim of sun's harh contracts)
I am sure IBM aren't so stupid they think it's OK to violate a contract, or to sign one they don't intend to honour.
as well as it's lackluster performance
Newsflash: the
From the parent comment: "If there was an award for software distributed that never got used, nobody would ever beat IBM."
Exactly. Remember TopView? It was a way of running multiple programs under DOS. That was the beginning, I guess, of the present software incompetence of IBM. Their failures seem to be a political problem with management.
IBM killed SmartSuite so efficiently! One month they bought it, and the next month it was dead! Awesome!
I remember news reports saying that IBM had lost $1,000,000,000 on OS/2. They gave it a name that means "deform so much as to make unusable": Warp. There were later reports that IBM lost $2 billion on OS/2. It was a better OS than Windows at the time, but IBM wouldn't support it with drivers. I guess IBM management took a hard look at OS/2, and decided losing $1 billion wasn't enough.
Recently, there were those stupid-looking spacemen selling IBM web software.
IBM Linux is one of IBM's few software successes. I suppose that's because IBM management is not able to ruin it.
I agree with the comment above the parent to this one: "The bloat of any even semi-featureful office suite, combined with the horrible, piggish slowness and ugliness of a Java app."
Didn't anyone in IBM management notice that Java does not have good GUI support?
>Do you have ANY idea how useful something like this could be to large environments? Where I work, we have 35,000 computers on the supported list. Two or three different platforms worth, PCs, Macs, and some occasional Linux machines. It would be kick ass if we could deploy one version of one productivity suite across the whole network, especially if we could do it with site based central servers rather than having to work on each and every PC on the fucking network.
In that case you should sit back, and take some time to study the centralized management features of MS Office. There are books about that, e.g. the "Office xxx resource kit" book for the version you run.
You will find that you can deploy office around your network (at least on the PC) from a central server without having to work on each and every PC on the fucking network.
The same thing probably exists for the Mac. If not, convince management that they should ditch it.
IBM solved the Java GUI support problem. Take a look at Eclipse, based on IBM's SWT (Swing replacement).
- Tal Cohen
Good
1. Finally it would be easier to work on WSAD and a word document opened together. Anyone who has worked with these mammoth applications opened together would know what i am talking about. By making individual PCs dummy terminals, this could free up PReciOus processor power
2. Easier to maintain / Upgrade. The guys at the IT should definitely NOT be happy about this one. They will probably get laid off now that it is easier to upgrade due to the centralization.
Bads
1. The only good thing about a monopoly is the standard that it establishes. The article talks says the J2EE suite has
"80 percent of the Office functionality most people use".
There would now be a possibility for a doc file developed in MS Office to look different on these IBM systems. Imagine your resume getting rejected because of that !
2. Centralization could suck with Network breakdown. Switching PCs will not work !
Siggy Say, Siggy Do
As Ethernet bandwidth increases, the argument for putting the power back in the server farm gets stronger. The server farm is in a controlled environment, it's easier to manage. If you assume in a few years many corporates will have gigabit Ethernet to the desk, and simple, cheap thin clients running XPE or Embedded Linux, the IBM approach makes sense. It is also going to be cheaper for developing countries to do this from the start than to put big, expensive, rapidly obsoleting boxes on every desktop.
To a certain extent too, it leverages the Linux strength in the server versus its perceived weakness in the desktop.
Corporate IT should be about delivering the necessary, usable functionality to end users. Geeks often lose sight of that. Microsoft might lose sight of it. But it's IBM core business.
Panurge has posted for the last time. Thanks for the positive moderations.
The editor is called Midas...looks pretty cool.
concrete5: a cms made for marketing, but strong enough for geeks.
I must say I gave up mostly on the alleged compatibility of open-source software with MS Word. There was always something not right in the presentation. Most of the MS Word stuff that I receive is forms from management and outside partners; those people apparently don't know how to make PDF forms.
We have a solution: we use rdesktop to access a single Windows 2000 machine from our Unix desktops, and we run MS Office and Acrobat on them.
only works for x86 boxes. Many of us use PowerPC or Sparc chipsets, and the java version runs a hell of a lot faster than the wine version in bochs!
You can't judge a book by the way it wears its hair.
Just let the user connect with ssh and use vi.
When the network is down, other things might grind to a halt but there is no reason my word processing should. Also, what if your server is up but /.'ed? Do I have to wait 15 minutes for my file to save when I have a perfectly good hard drive to hold it? Not to mention that I might want to unplug my notebook and use the same apps on a flight.
The only way it makes sense is if you can also install a local "server" on your PC and synchronize your documents with the real server when network is up. We have a project like that (webtogo, which is part of Oracle Lite) to run servlets a local PC, with access to replicated data in a local database. The same approach could be used for office applications.
I mean, really. Who the hell would want that? "Look at this, its just like a regular word processor, but extra laggy!!!"
It would definetly be a lot more laggy then a pure-java word processor, thats for sure.
autopr0n is like, down and stuff.
We do not live in the 21st century. We live in the 20 second century.
It was called project Hat-trick (4 apps). It never saw the light of day.
On a separate note, Larry Ellison likes to make lots of predictions - has he ever been right?
And
Now maybe Mono could open up the CLR a bit but I suspect a lot of people consider the project still work in development as well as having the Sword of Damocles hanging over it in the form of MS lawsuits.
I am all for the ASP model, and I really think that something like this has great potential. Esp, if I don't have to fire up Office every time I want to make a change to my .doc documents.
My sketicism is driven by the comment
It's a well known fact, that most people only use something like 10% of Words features. It's also well know that marketers like to exaggerate.
No, Lotus failed in most software they made all on their own. The CEO of Lotus had his title changed to General Manager after a few years of not producing results. I have a hope that Lotus has absolutely nothing to do with this software.
The company itself is strange, a friend of mine did some consulting for them, they have ganja themed commercials on closed circuit tv.
Sleep is for the weak.
Writing yet another windowing system based on the same concepts as the original is not the answer to the problem, taking the time to optimise the existing code is the answer to the problem.
1. It came out recently. It will get faster.
2. It's faster than swing.
3. It's NOT aweful on OSX.
And once someone writes a plugin for the OS, that's it. Nothing else. Look how many times gtk was ported.. or qt? It's not inherently platform itself, but your code itself is since it (swt) uses a fascade pattern to make all the api's common.
--
"I'm not bright. Big words confuse me. But Wanda loves me and that should be enough for you." - Cosmo
True, IBM has experience killing off retail software, like SmartSuite.
But they also have experience dealing well with server software, like Websphere.
This is not competition for today's bloated Microsoft Office running on your desktop. This is competition for tomorrow's subscription Microsoft Office running on your company's big iron server.
Bloat is a not that much of an issue there (and at the Websphere price scale), and I don't expect it to be that bloated, memory-wise. It's likely to have less graphic candy, wizards, and certainly less "covert OS upgrade components" than MS Office.
GUI support is almost certainly a non-issue too. This is Websphere we're talking about: thin-clients, J2EE, Servlets, EJB and Web Services... that kind of stuff. If IBM chooses Applets for their GUI they should be beaten to a pulp literally, and probably will metaphorically. But that is doubtful, unless SWT is much better than it looks right now.
They'll likely use a big, complex Web interface and just require all users to use IE or Mozilla 18.whatever (probably the later for flexibility's sake), which is certainly less than a requirement to install some other custom client OR an Office suite.
I can already hear the complaints: "What? They force me to install a particular browser instead of a 1GB Office Suite? Oh no!". I'm just speculating, but that sounds to me like the sensible solution.
There's a broad market of options for Web-based interfaces that work quite well if you don't have to deal with compatibility issues, your application logic is not the issue, and you have the resources to debug them properly as an application (as opposed to as 'just a website').
This passes the GUI requirements to the browser support of whatever you're using for GUI: Javascript and DHTML works fine. Or maybe they could go for one of those new fancy XML-based 'web-app GUI' projects that one keeps hearing about in Slashdot. Or they can go the plug-in way.
Whatever they find works best for their Websphere market, which is what matters to them here.
Freedom is the freedom to say 2+2=4, everything else follows...
Have you actually tried SWT on anything but Windows? It's awful! SWT is inherently not cross platform so it solves the Java GUI support problem on one platform only.
Actually, yes we are using SWT on a stand-alone application project right now. And no, we haven't had any cross-platform problems with it using the same SWT code for Windows 2000, Mac OS X and Linux (GTK, Red Hat 8) concurrently.
As well I found that Eclipse 2.1 for Mac OS X is just as GUI-sluggish as the rest of the OS X apps, so no big difference there. True, Windows beats it hands down for speed, but that's not SWT's fault - it's Mac OS X's fault.
The Linux GTK version of Eclipse 2.1 performed quite well on my AMD 1.47Ghz -- better than Mac OS X's performance and about 80%-90% as fast as Windows.
SWT was designed to be a "thin" abstraction layer. True, the other platform versions of SWT are a bit behind Windows SWT in terms of features (view dragging in Eclipse comes to mind) and speed but I think they are satisfactory. I'm really looking forward to further SWT developments from IBM.
----- rL
Well, first off there is no IBM Linux. IBM doesn't have its own Linux distribution.
As to IBM's few software successes, they include:
I'd guess you're just a bitter OS/2 fan. Get over it.
GCHQ Quantum Insert installed. If only our tongues were made of glass, how much more careful we would be when we speak
>> IBM Software and its Lotus Software Group have built J2EE-based spreadsheet, document and presentation graphics "applications" that will be bundled for free with the company's WebSphere portal
At first I thought "What has a desktop Office suite to do with Java2EE, which is a server side technology?"
In fact the slashdot story text is misleading. In the IBM announcement is used a little bit different term - "spreadsheet and word processing 'applications'", where applications is quoted. It comes from Lotus and Java2EE is involved, so it looks like a collection of collaboration tools. Most probably the documents live on the server and the office "applications" are Java thin clients that can show and edit them.
Really nice application for Java2EE, though. What is not nice is that they have bundled the suite with the WebSphere portal, which is a beast of extreme size, both financially and technically. It may be a nice solution for "IBM only" shops, but to little use to other people.
We can only hope that the software is not tightly coupled with WebSphere, but is generally Java2EE compatible, so it can be used with any J2EE server.
This is also interesting for another reason. Back when people shared CPU cycles, there was actually some interest in writing efficient software. You waste tons of cycles, Johnny down the hall runs slower, and people are going to be looking at the CPU statistics with the evil eye.
When workstations became popular, things changed. If you don't use cycles, they're simply wasted. So you might as well suck down most of the cycles on the machine. Efficient software stops becoming worthwhile.
Moving to server-based software again might mean that programmers actually have to write decent software again.
May we never see th
However, the great majority of workers are only mobile within site, where wireless networking is going to be continuous. And to do productive work they usually need to get resources off the network...they should not be relying on possibly obsolete versions of docs while mobile.
In terms of data integrity, it could be highly advantageous to many corporates if many workers could not do certain things without being connected to current data feeds.
So, I understand where you are coming from, but no MBA 101 for you.
Panurge has posted for the last time. Thanks for the positive moderations.
I think that where this is a bona-fide threat to MS is in changing the paradigm.
change may take place in two ways. The first is as mentioned by some other posters, you would get centralized management of the app and be able to reduce your TCO by not having to install/upgrade on every machine.
the second possible shift is where the real potential is. People don't just buy websphere and drop it in, they customize it to do something for them... so, now that there's going to be an office suite in websphere, companies that make customizations to websphere and have custom apps running on it can count on a standardized, cross application office suite being there that they can wrap their application around.
I think that's where the most potential for this is to truly change things.