Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?
The Internet

Public Beta For OpenDesk 82

Isaac-Lew wrote to us pointing out that has gone into public beta. From the looks of it, they are trying to implement an Office-like suite of features on the Web. Word Processor and such are still forthcoming, but they have got calendering tools, web mail and other "applications" in place. One of the most interesting aspects is that they claim to be open sourcing the project. I can't quite tell if it's a Sun-style commuity source licensem, or a GPL [?] style license.
This discussion has been archived. No new comments can be posted.

Public Beta For OpenDesk

Comments Filter:
  • Well, we don't really use Java either. Only Javascript for popups and the taskbar, etc., but we tried to use the least Javascript possible, due to problems with browsers! The apps are all done in Perl, so is the backend. The HTML is generated on the fly.
  • I not only RTFMed, I signed on with OD when I heard about it. I'm all for the OpenDesk concept. And I have no trouble with platform independence -- that's going to be a big thing now that computing is becoming ubiquitous. And as long as the developers are getting paid, free is good. Thus far, we're reading from the same page.

    However, the "interface nightmare" I was talking about was not the developer's, but the user's. I sympathize with developers (self-interest) but I must extend even more sympathy towards the users. When using a renderer based on something like TK (which I like), OD should work like a charm. You've got much better widgets in that case. It's when you're doing the rendering in HTML, which is the way you have to do it in a browser. As far as the end user is concerned, the shortcoming have everything to do with HTML.

    Based on what you've written so far, you sound like you know what you're talking about, so you'll have to agree that the interface widgets that HTML provides currently are very primitive compared to those in an OS GUI. Simple things like tabbing order, keyboard shortcuts, menu bars and drag-and-drop editing are quite difficult to kludge if not impossible to implement. A spreadsheet should (just) be possible to render in HTML, but a word processor that uses the controls available to an HTML renderer would probably be pretty terrible to most people, except perhaps for LaTeX and nroff-heads.

    I have nothing against the OD coders themselves and I think the idea's great. I just think that the current state of browsers is terrible for anything but reading and casual-use apps.

    (P.S. I'll quit with the cheap shots if you will. Call it the "SlashRage" syndrome.)

  • by Anonymous Coward
    If you just happen to like C better than Java, fine, but don't dress up your prejudices with false and ill-informed reasoning!

    Personally, I prefer to dress up my prejudices in ill-fitted tuxedos and invite them to dinner. It's usually quite pleasant as they tend to agree with me about everything. We also find the hate, ah the hate, keeps us warm on those cold winter mornings.
  • And get on the bandwagon of SmartWorker. For those who have no idea what it is, it's similar to Zope, except that it's written in Perl. It leverage the most popular web open source tech: mod-perl, apache and MySQL (Yeah, yeah, not really open source but close enough). It's close to an application server in functionality, but without much of the overhead and of the complexity. It allows to write simplify greatly complex web apps. Actually you end writing them in a way that looks like regular GUI apps.

    It takes care of all the persistence / session issues, you 'just' have to write the logic behind it ... and not the technicalities that you have to rewrite in eeeeevery CGI/module.

    Currently, the problem (for me!) is that all the presentation has to be hardcoded in Perl. You don't do any HTML, hopefully, you only manipulated high level objects. However, that's a problem for me who works closely with web page designers, and who is lazy enough to want them to deal with all the crappy HTML issues ...

    And since it's open source, well I decided to start working on this, so ASAP I will start implementing an XML interface to it, à la libglade, so that the HTML drone can take care of this aspect.

    Cool, heh?

  • And where the encryption/decryption will take place? If the answer is not "in your browser, dummy!" it is does not make much sense to implement the thing...
  • You mean like every ZD magazine host ad of Microsoft and have reviews of MS product? Come on. Everybody does that. The solution is let reader to tell the true so the rest of the reader won't be brainwashed, er, did I just describe Slashdot? You get my point?

  • I don't want to seem antagonistic or anything, and this isn't a flame, but I think you missed my point entirely. Read my post again.

    I *do* see the value in accessing my files/apps/data from more than one computer. I *made* the point that I *am* doing web applications for precisely the reason you mention. I am developing a web application and content management framework entitled Iaijutsu [] (in case you missed all the links. :) ). I *do* access the web applications I write using my Palm III, and am hoping to do so with my Sprint PCS phone. I wasn't talking about desktop applications. I am *for* web applications, but not in the way SmartWorker does it.

    The point I was making about another layer on the web is this: Why, specifically, another GUI toolkit? That's just another tight binding system between logic and presentation.

    If I understand correctly, SmartWorker can be made to present application differently with a set of 'rendering engines'. I noticed engines for WinCE, DHTML, etc...

    So, if I want to present the results of my application logic differently, I have to code another engine, right? Or, I have to reconstruct the code comprising my GUI layout logic, right? It requires someone with a knowledge of Perl to change the look & feel of an application in SmartWorker. In general, to change the presentation of my application in SmartWorker, I have to rework part of it's logic that has been devoted to manipulating the GUI toolkit. Is this correct?

    Now, I while I also appreciate that perhaps I can separate my application and a SmartWorker GUI into modules with an established interface between them, my question is: Given the fact that we have a machine readable user interface specification format that is much simpler than Perl (i.e. HTML), why circumvent it by layering another toolkit on top of it and adding complexity and tight coupling to the application? I'm not saying HTML is the best, but it's well known and there are plenty of tools to build it.

    The alternative I presented in my last post was this: A system like Zope [] or my Iaijutsu [] uses a template language to add flexibility to an HTML document to allow the presentation of the results of an application, mostly independent of the application itself. As long as the application generates the same well-defined data structure, the application logic is free to change, and the presentation is free to change-- independently of each other.

    More practically, what this means is:

    1) I'm the Perl guru, I write the backend code that churns through the database and returns some row data. 2) Upstairs is a graphic design geek who makes the pretty skins for my application. 3) The guy next to me is the HTML Production Artist who takes the designer's images and processes them into GIFs and HTML with Macromedia tools. He also takes my description of the template language and writes the simple looping tags that fill in table rows from my database results.

    With SmartWorker, I think it might go more like this:

    1) I write the Perl logic. 2) The designer hands the interface design image to the Production artist. 3) The production artist makes some GIFs, but no HTML, and hands that to me. 4) I write some Perl that lays out the tables and buttons and etc, as close as I can to the original Photoshop design. In this stage, we can't use anything like a nice Macromedia tool to generate a page because it's all done in Perl GUI toolkit calls.

    Now, the client asks for a radical change, but luckily enough, it's just cosmetic-- the application inner workings are still the same.

    With Zope [] or Iaijutsu [] the process goes like this:

    1) I don't have to do much at all, because my part (the application) has remained the same. 2) The graphic designer reworks the image. 3) The HTML Production Artist imports the image into a Macromedia tool, cuts out all the images again, and recreates the HTML template using the methods I gave him. Job done.

    With SmartWorker:

    1) The designer reworks the image. 2) The Production Artist gives me another pile of gifs. 3) I toss out the Perl module I wrote to layout the GUI, and start over writing more Perl code to try to match the new GUI.

    In summary, what's the problem here? For one, if any change is made to the presentation, we *all* have work to do because changing the SmartWorker GUI required Perl expertise. And for another, not only do we all have work to do, but we can't use all of the standard tools we have for HTML page layout.

    So, with SmartWorker, our process of delegating work to each other's specific skill sets is broken, and the tools that make 5 minutes' work out of making a 'skin' for the application in HTML are lost.

  • I can understand why you might be afraid of laying out some personal info on an external server, but unless you're the sysadmin of your Intranet, how can you make sure that that sysadmin isn't reading your mail or personal files? IMHO, any web based application is as secure (if there is encryption of course) as an Intranet. And I'm sure that a mailman somewhere opens mail before delivering it... We live in a strange world, don't forget that :)

    So forget about the security part, don't you think that such kind of app isn't what the Internet should be about? Accessing any info anywhere in the world? (and soon, from any device?) You might argue that with handheld devices, it is more and more easy to avoid those kind of apps. And I could agree with that. But can you easily share that information with someone else? I don't think so.

    PS: I am one of the developer of, so one might think that my opinion is affected by that. I don't think we are trying to force people to use our apps, some people will have a use for it, some don't.
  • Do you have something to say? I guess not, the sole point of your post was to ineptly flame me. Right back atcha, Skippy.

  • by ceeam ( 39911 )

    Actually, I added codepage switching to Opera and fixed a fullscreen-mode-with-a-taskbar-on-the-side bug..
    No recompilation - binary patching.

    Surely, reverse engineering is a useful thingie ;-)
  • The app is only as valuable as the data. This is the key to web services - tying services to data on a server.

    No, that's the key to internet services. The two are not synonymous. Mr. Carmack's product, Quake, understands the internet quite well. Playing Quake deathmatches with people hundreds or thousands of miles away uses internet connectivity.

    Web connectivity, on the other hand, implies a web-based application such as a java applet or HTML forms. Web applets are good for small, frequently changing or only briefly-used apps. Not for word processors, at least not at current bandwidths.

    To give a desktop application web connectivity you don't really even need to change the application itself much, at least with file systems like Linux and (yes) Windows. What you need to upgrade is the file system, to allow remotely stored data to look like local files, with read and write via the standard mechanisms. (Windows' namespacing allows this, and Linux's /proc directory is an example of non-standard filing for Linux.) Then all you need to do to the app is add some ability to find your data remotely without memorizing some long filepath, and you have all the functionality you want out of a web app.

    Both models have their pluses and minuses. I wouldn't want to have to download and install a problem to do many of the things web applets do. On the other hand, I'm not looking to play Quake in a Mozilla window.
  • I've got to admit, I was pretty surprised to see an advertisement for /. there. I'm still trying to figure out if it's meaningful or not:

    1) Should /. have "ethical" advertising policy (ie. only clear sites, only opensource software) - I'm not even sure if this is important or not

    2) /. linking to a site that contains their own ads. (Though this doesn't make much sense in being beneficial as they've have to pay more $$ for page views of people who already read /.).

    Weird stuff, just a weird coincidence I guess
  • >>No-one's stuff works with anyone else's
    >Come on. That's a teeny bit exaggerated. Yes
    >there are problems, but these are small compared
    >to using say C or C++ to develop the average
    >serious cross-platform app.

    Ah, but when I mail NetBeans about the fact that the NetBeans IDE doesn't work with kaffe, they tell me, "Use the Blackdown JVM, it's the only one we support." But the speed on my 450a is not acceptable with Blackdown's interpreter. NetBeans doesn't work right with IBM's JITC either. The NetBeans people themselves told me "We recommend our customers turn off JITC."

    C'mon, that's sad.

    Sure, porting an app between compilers sucks. But for God's sake, when it's done, it works. I can give you a Windows or Linux binary and you can expect it to work. Period. I don't consider dumping this problem off on the user a solution; nor do I expect companies to continually assure compliance with every JVM that comes out. It's a problem you should have to solve once, rather than continually reinventing the wheel.


    I don't deny, Java has some worthwhile concepts in it. But it doesn't live up to the promise. It's not fast enough to use, and it's not even delivering cross-platform support, due to the difficulty of adhering bug-for-bug exactly to the reference implementation (even worse when each app developer supports a different reference JVM).

    Maybe that's why Java has had such slow uptake, according to that recent programmer poll. It's just not that good at getting solutions done yet. Maybe that will change, but I have yet to see the cross-platform major application done successfully in Java. The only use I've seen for it so far is as a toy. C/C++ is a proven solution for a reason.

    BTW I know Gtk+ has major problems in its Win32 support that makes it a poor choice for that platform. But it has excellent support (as in perfect) across all Unices, which does cover a lot of ground. I agree that there's still no good way to write completely cross-platform WIMP apps yet.

    It's a serious problem.

    But I don't see slowing my machine down to the speed of a 486 as a solution. We could also all just buy and use 486s, but no one's suggesting that, either.


    Coding in assembler can make you a cool hacker. But big assembly projects usually end up as messes -- there's too much detail, and it's too hard to abstract. That's why OSs stopped being written in assembler 25 years ago.

    On the other hand, being cross-platform in a large C/C++ project forces some things that are very good for large object models:

    - It forces you to abstract design blocks that are not consistent across platforms. This can be a very useful thing since those non-standardized blocks are the ones most likely to change in the future, in which case the abstraction can pay off.

    - It forces you to avoid clever hacks and tricks and compiler quirks that really don't belong in any significant project, since they're much harder to understand and maintain later. It'll also make your life much easier when your compiler changes major versions.

    this is way offtopic. woohoo!
  • I don't think that's paranoid at all.

    I've got an account with them, I admit, I bit on a slashdot banner because it looked interesting, and I think that the idea is fabulous. But no, I wouldn't put all of my information up on their machines.

    Not that I don't trust them, I'm sure that they're honest, but if I was to put my personal information up on opendesk, then the security of my information would depend on the security of my box plus the security of their box, i.e. what was keeping my info from other people would be the weaker of the two computers security schemes.

    I don't think some of this Active wallet stuff for online shopping is such a great idea for similar reasons. (You know, click this button, fill out the form, we'll keep your credit card and all your marketing info in our secret folder and automagically fill out online forms for you) If you think about it, it can't possibly be a good security idea since you're widening the scope of people who have access to the data.
  • I'm not getting on the bandwagon, and here's why...

    For those who have no idea what it is, it's similar to Zope, except that it's written in Perl.

    Nope, it's not like Zope. Check out my project, Iaijutsu []. THAT is like Zope, only written in Perl.

    Currently, the problem (for me!) is that all the presentation has to be hardcoded in Perl. You don't do any HTML, hopefully, you only manipulated high level objects. However, that's a problem for me who works closely with web page designers, and who is lazy enough to want them to deal with all the crappy HTML issues ...

    Zope (and Iaijutsu []) is an object-oriented content management tool and application framework.
    SmartWorker is more akin to GTK and QT, as it's an abstract windowing toolkit on top of the web.

    With a system like Zope or Iaijutsu [], logic is separated from presentation by an HTML-based template language. Both Zope and Iaijutsu have object persistence databases which allow storage, search, and retrieval of content and application instances.

    And in the case of Iaijutsu [], I just recently finished a Skins support system that allows one to create sets of templates to completely change the look & feel of an app or set of content from a user preference, for example.

    While SmartWorker does seem to have some persistence and session data facilities, from what I've seen so far, it's nothing like the object persistence database features of Zope or Iaijutsu []. On the SmartWorker site, they claim 'This is not a system for generating web content from databases', and thus I think they'd agree. From what I've seen in SmartWorker so far, the data storage is nowhere near as sophisticated as Zope or Iaijutsu []. I may be mistaken, so I'm going to keep investigating SmartWorker.

    However the reason I'm not jumping on the SmartWorker bandwagon and will continue Iaijutsu [] development, is this:

    SmartWorker is a GUI toolkit for the web. I don't need this. In fact, I'm doing web applications to escape the OS and GUI toolkits and tightly coupled logic and presentation.

    I understand SmartWorker has different 'rendering engines' for different 'platforms', but it just doesn't seem flexible enough to me. Beyond just 'platforms', what about the simple fact that I have to collaborate with a) A graphic designer, b) An HTML production artist, c) The Money-providing Client?

    I don't want to touch any Perl code when all we want to do is rearrange the form on the screen. We're not changing the data that flows in and out of the application, we're just changing the 'skin'. With Zope and Iaijutsu [] that's a task I can delegate off to an HTML slave. With SmartWorker, I have to rebuild the code that makes the GUI.

    SmartWorker just seems like a misapplied metaphor to me. Because, we have an OS. On the OS is a windowing system. On that windowing system is a windowing toolkit. On that windowing toolkit is a web browser. And so with SmartWorker, we try to make the browser into a windowing system and build another layer of windowing toolkit? Just seems like a mistake to me that wipes out most of the flexibility the web browser provides for logic/presentation separation.

    My $0.02. Oh, and check out Iaijutsu []. =)

  • Gtk looks nice, but I'm not sure most Windows users would want to or would bother installing Gtk to run an app.

    "This app requires vbrun100/200/300/400/500.dll in order to execute. You can download it from..."

    I know plenty of windows users that have d/l'd vb runtimes and what have you, I can't see ti as being much more difficult to set up a nice little installshield deal that will stab gtk libraries into the ever-tangled pile-o'-spaghetti that is c:\windows\system. Or is the GTK/win32 install more complicated than that could handle?

  • That's unfortunate that you're anti apps on the web, because that's where it's all headed. Some neat, neat stuff, in my opinion.
  • OK, deal, no more cheap shots. I guess I just get irritated with people who whine about things that other people have put a lot of effort into and are giving away for free solely because it's cool technology. I think that's beyond criticism, so I never complain about stuff I get for free.

    OK, I understand what you're saying about HTML widgets. The entire point of SW is to do a "best fit" UI depending on what it's rendering to. So, if it's rendering to HTML3 browsers, it will dumb down the interface but (in theory) retain functionality. All of that is transparent to the programmer and the user. You're right: to implement some things in HTML or DHTML is like pulling teeth, but with SW, it only needs to be done once.

    Moreover, the whole reason for making the source available is to have people like you extend the functionality: if you want a Tk renderer, write one! It'd not be a difficult problem. There is already a project in the works to run SW apps using Perl/Gtk, and describing the UI using Glade, making it trivial to write and deploy apps that are truly distributed.

    I like the SW way of doing things because it totally avoids HTML. As an app developer, you never need to worry about TD tags or forms or hidden fields. And, it's automatically deployable on many different platforms, remotely or locally.

    The entire purpose of SW and OD is to avoid HTML and templates. This is utterly fundamental to the project. I guess the theory goes that a few years down the road, we'll have a much more sophisticated interface language running in web browsers, at which time SW will really come into its own. For now, limited to DHTML and HTML, I agree with you that the UI is limited. But that's not really the point, no?

    So, I encourage you to take a stand and get yourself involved in SW. It would help the project, and if you wrote a Tk renderer, it would really bring things along.


    PS. Ok, I'll come clean, I helped out with the original SW project, hence my defence of it.

  • Well, before you get your ethical panties in a bunch, you should note that OpenDesk has advertising banners on SlashDot. Looks like a banner exchange to me.

  • by hadron ( 139 )
    Ok, fine. I was really only objecting to the poor journalism that typifies slashdot nowadays. :)
  • You might not like to run a word processor on a web browser, but I would wager in three to five years, that's about the only place you'll be able to run a word processor. It's all going distributed. There will come a day when computers will ship with no software installed on them, because all the software you need will be running of the net or someone else's machine.
  • An open-source office project is definitely good. I know there's Applix, StarOffice, and WordPerfect, but they seem to be lacking. Besides, most of them are big and bloated anyways. Hopefully OpenDesk will gain some popularity.


  • Besides the one javascript error I get when I bookmark sites this thing seems really cool. And handy for us who are always in front of a different computer!
  • their web site wasn't descriptive enough with examples of what you could do when I clicked on the /. banner ad. Hmm... are they getting some promotional consideration here ;)

    Hope it is really Open/Free "Open Source." Sean

  • Is it just me, or do chartruese and light yellow look really really bad on a white background?
  • It took me some time to figure out what they were doing. It is not really an office suite, but a set of applications running on a web site.
  • by Ed Avis ( 5917 ) <> on Friday October 29, 1999 @07:06AM (#1577099) Homepage
    'When Licences Collide'

    Hopefully Corel's recent experience with licence clashes will convince them to go with an existing choice (preferably GPL, LGPL or XFree) rather than inventing YAL. I have argued previously on Slashdot that the whole QPL 'you give us rights to do what we like with your modificiations' thing is unnecessary, since companies require copyright assignment for legal security in any case.

  • by davie ( 191 ) on Friday October 29, 1999 @07:01AM (#1577100) Journal e.html []

    In a nutshell, they haven't decided on their final license yet. They state that whatever license they decide on, it will be OSI-friendly.

  • I wonder if OpenDesk runs on Microsoft's ActiveDesktop? Maybe ActiveDesktop was just ahead of its time.

    Another scary thought. If OpenDesk becomes popular, when will the first ActiveDesktop-clone be available for x-windows?
  • by Telcontar ( 819 ) on Friday October 29, 1999 @07:03AM (#1577102) Homepage
    There are actually two licenses: []

    " public license (ODPL), and the SmartWorker public license (SWPL). The former covers applications written and hosted within The latter covers our open source application framework, SmartWorker.

    The two licenses differ in the following ways:

    1. SWPL applies to the core API that is built on. However, it can also be used for software and, therefore, must be flexible to accommodate other uses. The SWPL will
    be similar in intent and structure to the GNU Library General Public License (LGPL, available at This means that it will allow linking with non Open Source code while always remaining free and Open Source itself.

    2. ODPL is more strict than SWPL insofar as it does not allow linking to non-Open Source code. Anyone
    who uses ODPL covered code in either a new distribution or a new server application must release the source code to their software under ODPL. While the GNU General Public License (GPL, available at was considered, we are finalizing a license closer to the Mozilla Public License (MPL, available at as it is clearer about uses of source code in server applications, as well as including the aforementioned patent and copyright infringement clauses."

  • Please note that if they are going to make it open source, then they cannot be considering making it available under a SCSL-style license, because that is not an open source license.
  • Nevermind, I am sofa king wee tod id. And our T1 is running really slow.
  • by BadlandZ ( 1725 ) on Friday October 29, 1999 @07:20AM (#1577106) Journal
    Uh... People are talking about this like it's a DE for X or something. AFAIK, this isn't a DE, and it's not a UI for X at all.

    From the looks of it, it's a thing where you set up "your desktop" on thier web server, and thus can access "your desktop" through a browser from anywhere....

    Maybe I'm wrong here, but this isn't the next KDE/Gnome... It's a weird spin on client/server "work environments." And, I'll add to that, I'd much rather have a nice big server on a local LAN and do "X -query server" ;-) for a more traditional way of keeping all my work in one place and accessing it from anywhere (locally, bandwidth allowing ;-)

  • by c+era ( 102193 ) on Friday October 29, 1999 @07:04AM (#1577107)
    While this is a great idea, who is going to prevent the company from viewing your workspace. Now some of you will think I'm just paranoid, but look at how many ISP were reading e-mail. With the current laws in place it is legal for them to "check" your workspace to "make sure" no laws are being broken.

    They also make it seem that it is imposible to get your password because they use challenge-responce. Let's get real, look at NT!

  • Unless that 'open-source' license is the GPL, I'll just stick with GNOME, thank you.

    Beware non-free licenses.
  • This isn't the only site pushing this model. [] is also pushing the site-as-desktop metaphor.

    Frankly, it doesn't make sense to try to emualte the old user metaphor onto the new one. Most unix users never became tied to the desktop metaphor (as unix traditionally never had a strong desktop), so I see these services as training wheels for mac and win98 users.

    The problem I have with these services is that you can already get email, calendaring, message boards, etc. at a number of sites (yahoo, excite, lycos, aol) in a much more lightweight fashion - straight html with forms.

    While I found to be quite cute, it wasn't nearly speedy enough to be really useable...and since I was never a Mac or Windows user, I was never tied very strongly to the desktop metaphor. As a unix user, client server has always felt natural to me.

  • If I ran something like this, I would want it be on a local server. For some reason this reminds me of Domino, only OSS and web-based.

    I have been doing alot of reading about DHTML and HTML 4.0 recently and I can't help but think that most of the technologies exist to implement something like this: XML data files, Javascript/DHTML window messaging, CSS desktop preferences, and your choice of web server backend. The weakest area seems to be in graphics, where one is limited to gif and java applets.

    The problem is not the web, it's the browsers - the only way I could see something really coming together is if a really good, but extremely lightweight open source browser would come into being. Where the hell is mozilla and is it DHTML/JScript/CSS capable now?
  • by Anonymous Coward
    ok, so having all info exposed out on the internet is not so usefull. but throw this on the company intranet and you have a real cost saver. most files and emails are kept on intranet servers anyway so why not go with something like this instead of msoffice? it's gotta be a cost saver. no more hellish experience after IT upgrades you to the lastest version of office. no more upgrading hardware just so it runs ms's latest and greatest. as long as it runs a current browser who cares what it is.

    it's platform independent. the more you think about it something like this could hurt ms real bad. not that my goal in life is to see ms go down but some competition would be nice.
  • by aheitner ( 3273 ) on Friday October 29, 1999 @07:28AM (#1577112)
    I don't want to run apps that suck.

    This includes, dismissively (evidence the the contrary lacking completely, in my experience)

    - all wordprocessors, IDEs, and anything else written in Java. Too many compatibility problems with JITC (no one's stuff works with anyone else's), and otherwise way too slow.
    Forget it.

    - any app run through a web browser. Come on, the web browser is the single flakiest app on my machine. Slow, bloated, crashy. IE5 is better than Netscape but not so much so that I'd want my productivity to depend on a web based desktop.
    Also forget it.

    A call to application developers:

    There is, to paraphrase Carmack, a self-imposed discipline that comes from supporting multiple platforms. Get yourself a nice cross-platform widget library, Gtk+ (1.0 works in win32, but they need to spend the time to get 1.2 to win32 perfectly), wxWindows, Qt and pay for it, I don't care.

    No Motif. Period.

    No MFC. Period.

    Anything you write should run beautifully on my SPARC 5 85MHz. Period. Nothing useful may ever require more power than that.


    The OpenDesktop song is pretty good tho :)
  • But what about 7 years from now after the hostile takeover?
  • Well.. the same folk who need a portable computer with a modem might need this. And it might be more secure (harder to steal, but more access? with the right security it might be a good choice). And it will let them access their latest work from the airport hotel. But not for me.
    This year.
  • The desktop metaphor is problematic, but not for the reasons you're citing. The most telling phrase in your statement is
    and since I was never a Mac or Windows user, I was never tied very strongly to the desktop metaphor. As a unix user, client server has always felt natural to me.
    The Win/Mac approach is very different from the Unix approach. Unix is by programmers for programmers. It's designed with toolbuilders in mind -- to use Unix, you have to have a mental model that is much closer to what the computer is really doing than in consumer OSs. It's natural for programmers to have service spread all over the place and to cobble them together as needed using either single command lines with or | or scripting. It is unnatural (and cruel) to expect this for most other users. We should not force them to learn computer science anymore than we ourselves are forced to learn internal combustion just in order to drive a car.

    The desktop metaphor is a pretty good concept for the average person. This should not be read as "stupid", but rather as people whose primary field is not computer science or straight-out unix use.

  • Web pages are interface nightmares. The ability to use them on just about any platform is an advantage that is far outweighed by their many disadvantages, the least of which is that their interface controls are still pretty primitive in comparsion to those on just about any GUI system, be it Windoze, MacOS, or any of the Unix-based ones such as X, KDE or Gnome.

    I suggest that the developers of this odious office-suite-on-a-browser be forced to program using a web-based suite of development tools. If ever there was a time that someone had to "eat their own dog food," this is it!

  • Open Source aside, there are a few other free Desktops on the web, with varying degrees of functionality.

    There was an article in Computer World [] about them.

    I used two of them, Netscape Netcenter [], and Visto's Briefcase []. Each has its strong points, but neither is perfect (Visto fails to sync the Pilot flawlessly, and NetCenter doesn't upload all bookmarks and Address books).

    The idea of syncing my Pilot to the web, having all the web bookmarks, E-Mail anywhere in the world when I travel, at home, at client sites, or at the office is very appealing.


  • What a bizzare perspective you have. Nobody is FORCING you to do anything at all, so why do you suggest that the developers of this "office-suite-on-a-browser" be FORCED to program using a web-based suite of development tools?

    I find it amusing that you complain when you are given something for free. Some people work hard to provide something cool and free and you can only bitch about it.

    Save your cynicism for the companies that deserve it.

    Incidentally, I don't know if you use "cvsweb" but it's a PERFECTLY usable web-based development tool. I believe the OD boyz use it, too. So fit that punishment to your crime, fool.

  • by KrAphtd1nN3r ( 33859 ) on Friday October 29, 1999 @07:30AM (#1577122)
    Hey, I'm one of the core developers of the system, and I'm currently in charge of the Crypto modules. In one of the following versions, you will be able to encrypt your data on the server, using an algorithm of your choice (3DES, Blowfish, etc.).

    I'm sorry, just didn't see the email you sent.
  • Hey,

    we're all a bunch of security freaks here, don't worry, and we'll have an SSL key (128-bit global id) within a week. We've been delayed due to NSI bureaucracies!!

    Besides, I don't think anyone around here has an interest in what's on the users' workspaces...we are not Microsoft, repeat, we are not Microsoft!
  • While I realize that this is still in development and we should offer the benefit of a doubt, let's keep in mind the consequences of the 'office suite' mentality.

    There was a time when many products flourished in the traditional categories of productivity software (processing words, spreading sheets, basing data) and customers had a choice between competing 'best of breed' products. The monolithic all-in-one (and preloaded) office suite changed that, and I believe for the worse.

    It was now possible for inferior products to 'inherit' market share through bundling with more competitive products, after all who would spend extra money and effort to use a better DBMS if the one included was (almost, if buggily and painfully) not too bad? Perversely, this bundling did nothing to reduce feature bloat (table management in word processors, sorting in spreadsheets, etc) since each product was originally developed separately, and in fact was still required to be used stand-alone.

    Thus we landed in a market where vendors who essentially did one thing and did it well (WordPerfect, for instance), and vendors who did several things well at good prices (Borland), were marginalized into near oblivion. The characteristics of the current market leader in desktop office suites are well documented here and elsewhere, how much of those problems are inherent in the concept of the one-stop-shopping software experience?

  • "I'd much rather have a nice big
    server on a local LAN and do "X -query server""

    I agree but I can think of situations where a clone of this system running on the local LAN might be handy.
  • Along with the trust required of the host of your information, the whole thing runs without and encryption (SSL).

  • Ok, i guess i;m a bit slow with posting, distregard.

  • How often (due to system load, bugs in the software, network congestion, or an errant backhoe) does slashdot, or another popular Web site, become inaccessable to a user? How often does a critical server at the average high-tech company go down, taking some server-based applications (if not the company's whole intranet) with it? Do you really want mundane office applications that you (and the non-geek office workers in your company) use all the time to depend on an Internet connection to one group of servers?

    To make matters worse: A system like this will probably be most heavily stressed at a time when the people who use it need it the most. Just imagine people all over the United States using a spreadsheet or accounting application through, and then all of them trying to run it simultaneously on April 14th. (Note to non-US readers: that's the day before US Federal taxes are due.) If this thing is installed on one company's server and run over an intranet, it makes a little more sense, but if the people who maintain that server aren't competent, the server and the intranet become a point of failure for all the company's office applications.

    Desktop machines and desktop applications have plenty of administrative problems, but at least if one of them goes down, only one person's work is hosed.

  • I wasn't saying it bothered *my* ethics (I'm as unholy and unscrupulous as they come ;) ) but ask yourself how you'd feel if M$FT & /. worked up a banner exchange? I'd imagine there'd be quite a few /.ers chomping at the bit. Along the same lines, ho would you feel if /. did a banner exchange with a site that was deliberately FUDding, or trying to capitalize on the OpenSource movement, while not really "getting" it (I'm looking at you, Sun).

    I guess it's the same argument for any business or organization: supporters of Greenpeace would go ballistic if they advertised on Exxon's site, or vice versa.

    The question is, do you think this _in any way_ compromises /.? (Before anyone gets excited, let me first say that I would guess this wasn't even realized or intended at all by our faithful /. staff, and I don't feel that it compromises anyone - I was just curious as to what everyone else thought).
  • by Anonymous Coward
    Obviously this application of web technology is weak. Carmack though, simply doesn't understand the web if he prefers GTK apps to web apps.

    The app is only as valuable as the data. This is the key to web services - tying services to data on a server.

    Compare GNUCash to On the Wells Farg owebsite, I can balance my checkbook (as with GNUCash), view statements (as with GNUCash), and pay bills, transfer funds (can't do with GNUCash). Anything remotely interesting I want to do with my money cannot be done with desktop apps.

    Now for calendaring. I can use Yahoo calendar or I can use the Gnome or K calendar. THen one day I get on a plane and go to Europe...I need to check my calendar. Gnome and KDE aren't too much help here. With a web calendaring service, I can use a public terminal to check my calendar.

    If you people can't see the use of this model, then you're still stuck in the 80's, competing on the "desktop" with Microsoft. Its a dead metaphor kids.

  • Well, for your information, the apps on the server have been almost completely written by the core developers on the system. Unlike most of the projects, we try to adapt the backend for speed and ease of use, and we're the ones using it right now, so we did "eat our own dog food"...and we know what it feels like, and frankly, it's not that bad!!
  • by ceeam ( 39911 )
    It doesn't work with Opera.
    NN or IE only.

    Speaking about 'Open'.....
  • "Free" I have no problem with. Many good things come for free. Many bad things too (disease, for one, is free, guessing by the number of poor people who have 'em). My concern is whether or not a product serves the users well, the people for whom we write code.

    It's my opinion that HTML forms are a user interface nightmare and have really lowered the bar on user interface design. Simple things such as enabling and disabling controls require crazy workarounds, and even then, they don't quite work as well as the interface controls on GUIs for OSs.

    Besides, did you think I was being literal about forcing anyone to do anything? When they get around to providing lives and clues on a web page, I strongly suggest you bookmark that one.

  • > Why do you think window managers are so popular?

    don't judge a tool's popularity based on where linux developers are working.

    linux people are still trying to build a good word processor ten years after microsoft locked up the market. as far as the desktop is concerned, linux people are chasing the tail lights of apple and MS. this isn't an insult or a troll, just the way it is.

  • Re:Let the punishment fit the crime (Score:) by Master of Kode Fu on Friday October 29, @05:40PM EDT (#) (User Info) "Free" I have no problem with. Many good things come for free. Many bad things too (disease, for one, is free, guessing by the number of poor people who have 'em). My concern is whether or not a product serves the users well, the people for whom we write code. It's my opinion that HTML forms are a user interface nightmare and have really lowered the bar on user interface design. Simple things such as enabling and disabling controls require crazy workarounds, and even then, they don't quite work as well as the interface controls on GUIs for OSs.

    dude, you are missing the point. OD is built using the SmartWorker framework, which attempts to take the "interface nightmare" away from the developer. If you look at the way a smartworker app is structured, it has no HTML in it whatsoever: SW handles all the translation into whatever GUI it's rendering to, be that IE5, Opera (yes, eventually), a Palm Pilot, or a cell phone. The OD team is building renderers for all these things... you could even run an SW app locally using a Perl/GTK renderer if you want. It has nothing to do with HTML. RTFM.

    I re-iterate: It's free.

    When they get around to providing lives and clues on a web page, I strongly suggest you bookmark that one.

    Oh, rest assured, I will bookmark it using my OD bookmark manager. And when you're using someone else's computer and you can't remember where you bookmarked that "life" of yours, I'll be laughing at you.

  • I hardly see how it relates to open desk, except that it might look cool.
    The ActiveDesktop simply a the new desktop/background - what's new is they made it an OLE container...meaning it could contain activex/javabean componenets. Since IE3/4/5 are ActiveX controls, you can any web page you like (and many of them) on the desktop. So the idea of OpenDesk on the ActiveDesktop is a good one, but you make it seem like the ActiveDesktop is some obscure technology's not. It's OLE and ActiveX - and Ofcourse you can stick any website on the ActiveDesktop - or any program providing you make it an OLE provider.

    Before it comes to any X-Window manager, they need to get inplace a component model, something like KOM or Bonobo (no i don't like Gnome anymore).

  • well those people would be the immature ones with ego problems.
  • by greenrd ( 47933 ) on Friday October 29, 1999 @08:27AM (#1577142) Homepage
    anything else written in Java. Too many compatibility problems with JITC,

    Just in time compilation, if that's what you mean by JITC, doesn't cause any compatibility problems AFAIK. It's VM implementors (including Sun) not implementing the APIs correctly (usually by accident). The core Java 2 Platform APIs will be staying pretty much fixed for some time now, so chances are compatibility problems will be seen less and less.

    No-one's stuff works with anyone else's

    Come on. That's a teeny bit exaggerated. Yes there are problems, but these are small compared to using say C or C++ to develop the average serious cross-platform app.

    Gtk+ (1.0 works in win32, but they need to spend the time to get 1.2 to win32 perfectly)

    Hmmm... I smell double standards. "Java has temporary problems? Bin it! GTK has temporary problems? Don't worry, it'll be sorted out soon." Even if GTK manages to be a cross-platform widget library, there's still other issues like code generation, filesystem, networking, etc. Java gives you a complete VM, GTK doesn't.

    So, you're not using Java because it has small problems - but you recommend GTK even though it has much larger holes in its cross-platform abilities?? That doesn't make sense to me.

    There is... a self-imposed discipline that comes from supporting multiple platforms.

    There's also much discipline to be learnt from coding 80186 assembler - I know, believe me. That's not a good reason for using it. (Yes, before anyone flames, there really was an 80186 chip - the RM Nimbus was one example of a machine that made us of it.)

    If you just happen to like C better than Java, fine, but don't dress up your prejudices with false and ill-informed reasoning!

  • by Anonymous Coward on Friday October 29, 1999 @08:31AM (#1577143)
    Many of the posts I see in here (which are moderated up), seem to extoll the desktop model, while doubting the efficacy of networked applications.

    This is odd considering that as a group, you already make extensive use of a web message board technology by virtue of using this site.

    While I don't agree with the "web site as desktop" metaphor, there is obvious power in website as applications.

    I've never seen "desktop applications" as terribly useful - I get around, and I need data to be accessible from wherever I am. Web email, file storage, and messaging apps are crucial to me.

    I can't even remember the hassles of having to haul around my own laptop in order to see my data.

    I understand there are security concerns. Use sites that emply ssl, pick your password intelligently - you'll be fine.

    When I see people devoting their time to making calendar apps for KDE or GTK, I can't help but think that these folks just don't understand. Even though it sounds pithy, Sun is correct - "the network is the computer".

  • Surely, since you're a proponent of open software, you could just throw a debugger on Opera, modify a little code, and recompile it to make it work?

  • by Anonymous Coward
    But that's too hard. Microsoft goes to the trouble of continualy giving me tools they say I need to include in my apps, I've done my best to appease them. I've even forgotten how to write those 'linkered lists' or whatever they're called.

    Anyhow I just can't port my apps 'cause I use the MSXML and WININET controls and at this point my brain has atrophied enough that I can't escape this hole I stepped into so willingly. I've found it's better to let others fix the bugs and I can just worry about rubberbanding this shit together. I mean, I have to spnd most of my time filling out those papers for that 'net IPO.

    I've created this cool clone of notepad, but the really cool thing is that it runs over the internet. It's amazing!!!! It's the first step in out taking over the non-existant ASP market. My next app is a calculator THAT RUNS OVER THE INTERNET.

    I though these ideas were ingenious, but when the marketing people told me what I had actually created... I am a geeenious. When the company goes public I will buy and sell you all!!!!

    Does anyone have a text control I could have the source to? I was just told this thing needs to work on other platforms and I can't find the right MFC class to do that...

Outside of a dog, a book is man's best friend. Inside of a dog, it is too dark to read.