The dirge of NeoOffice
Before we really get into this post on NeoOffice and OpenOffice.org, I need to give a little background, because:
- The following email exchange won't make any sense without it.
- There's been rumblings about what's been going on in bits of pieces around, but it's more confusing than clarifying for many.
This is something I've been sitting on, primarily because I wanted to see what was going to happen with it over the last week, and I've been trying to get something else out the door, but then I saw enough weird commentary floating around that it's worth getting out there.
I'm going to be working in reverse, which means we'll start out with the heavy stuff for those who are up to speed on what's going on, then break it down -- to the best of my knowledge -- for those who have just seen bits and pieces...
On October 22nd, I checked my mail at an airport terminal before hopping on a train, on my way away from the talk, and had the following email waiting for me. This was a reply to an email sent to Ed Peterlin of the NeoOffice/J project, telling them:
- Who has taken over the OOo x11 build, as well as co-leading the porting project.
- That in order to better promote the OOo 2.0 x11 build, they'd like to relocate mention of NeoOffice/J to a separate "derived projects" page. Currently, when you go to download, you see both the x11 build and the NeoOffice build side-by-side.
- Who would be taking over ownership of the pages.
I'm not including that original email for a variety of reasons, but here is a somewhat sanitized version of the reply, and who it went out to:
To: Louis Suarez-Potts < x@gmail.com >, "eric.bachard" < x@free.fr > Cc: Patrick Luby < x@planamesa.com >, Eric Hoch < x@gmx.de >, drunkenbatman < x@drunkenbatman.com >, Simon Phipps < x@Sun.COM >, jis xSun.COM, Tim Bray < x@Sun.COM >, David Pogue < x@pogueman.com >Subject: Re: updating mac os x x11 pages
The original goal we had last year was to devise a combined download page which presents the raw facts of both projects. The objective was to prevent people from "discovering" one or the other of these projects as a Macintosh cure-all (which was frequently the case on the various user and development lists). Unlike other OpenOffice.org supported platforms, a viable fork does exist for Mac OS X which is not directly supported by Sun Microsystems. Different users on the Mac platform may have different needs.
The existing download page provides a side-by-side comparison of the two major ports and is intended to allow users to choose which one better matches their requirements. The OpenOffice.org download page has always been set up to emphasize that the X11 port is more in conjunction with the mainline release on other platforms. For the online forums I host of my own pocket, X11 gets top billing as has been since I first put the server online. I do not see how work on the 2.0 X11 release should change this arrangement.
If you, Eric, CollabNet, and Sun Microsystems have a strong desire to relegate the NeoOffice project to a footnote and exorcise any mention of us from the OpenOffice.org site it is your prerogative as OpenOffice.org is but a corporate-sponsored front. No official entities entered into any legal agreement to provide concurrent mention of our separate products, so any unilateral modification is perfectly sound. I have no intention of using my community OOo commit privileges to instigate a "war" of differing viewpoints.
In the end it will be only end-users who suffer from this withholding of information. The OOo/Sun vaporware Cocoa port announcement has already dried up nearly all donation momentum from the NeoOffice project. This has severely hindered our ability to continue to have a full time volunteer engineer devoted to development. We do not even have funds to enter into arrangements to gain official access to a DTK.
This lack of funds precipitated directly by your actions restricts our ability to achieve compatibility with the Mac-on-Intel platform as well as all thoughts of future OOo 2.0 development. This may very well be the first nail in the coffin of our project.
I wish you and your corporate agenda all the best in your battle against us few from around the world who selflessly donate thousands of hours and thousands of dollars to make software freely available for all. Your cause is a noble one that I'm sure people will remember for years to come.
ed
If one had had a mind, one could backtrack chronologically through the smattering of blog posts out there and line them up with when this email went out, and if you did I wouldn't blame you if you were left with a raised eyebrow and some amusement.
Depressing stuff if you've followed what's been going on, and Ed is having to play a dangerous game to keep his project from getting crushed between interests that sometimes align with his work, yet sometimes threaten to diverge and cut off its oxygen. Now, if you haven't been following what's going on, it's worthing laying out some of the history -- as I know it -- so you're up to speed...
There used to be an office productivity suite known as StarOffice, which was a goofy name but made sense when you realized it was developed by a company named StarDivision.
Sun purchased the product in late 1999, and then open sourced most of the code about a year later in 2000, and dubbed it OpenOffice. Unfortunately, it turned out that someone owned the trademark to "OpenOffice", so its technically referred to as OpenOffice.org or OOo, but most people ignore that and just call it OpenOffice.
Now, StarOffice didn't go away. People contribute code to the OpenOffice project (including Sun employees), and once its reached a certain level, Sun branches off the code and adds in some special proprietary seasonings to create the next version of StarOffice: fonts, templates, a database, various filters for importing other formats, a different spell-check system, etc. StarOffice then gets sold under more than one name, on more than one platform, sometimes via subscription model in tandem with their other offerings.
The whole deal is highly reminiscent of how AOL creates their Netscape suite, and gives Sun traction in areas they otherwise wouldn't if they weren't able to provide a credible answer to Microsoft's Office because, you know, people actually do need to do Office stuff. It's really hard to get a company to buy into say, Sun's linux distro or thin-client stuff if they can't open spreadsheets. Because it's OSS, all the other linux distros (or Windows, or FreeBSD, etc.) can include OOo (which isn't their only option), but the fact that OOo isn't an exclusive benefit Sun can offer is offset by:
- They're able to shift some of the development costs off their neck.
- Anyone not using Microsoft Office is a win for Sun, on any platform.
On the whole it works out well for Sun, or at least they think it does, and we need to be clear about the works out well for Sun part, because while there is nothing wrong with that, it's one of the primary reasons we're headed for a clusterfuck when it comes to the Mac.
The OpenOffice project has a lot going for it. It's free, which is always a big deal when it comes to adoption, and it's feature-rich. You can generally do anything you'd need to with it, and did I mention it's free?
It's not perfect -- it's got a lot of crud and such built up over time, and it's not going to win any usability awards, but it's still a big deal. I.E., I knew a couple in South America who was setting up a community computer lab, so people who couldn't afford computers in their home could go there and email relatives, do homework, receive training, etc. You're looking at the cost of the hardware and the operating system, and then the third-party software. In this case, unless they were able to find major discounts, a copy of Microsoft Office would cost the same or more than what they were paying for each computer.
There would have been other options if they'd been looking at a full-Linux solution (whatup, KOffice and Gnome Office), but for a variety of reasons (not the least of which was some donated educational software for kids and training issues) being able to throw OOo on the computers saved them a ton of cash and let them have six computers instead of two. OSS productivity suites are a big deal.
Unless, of course, you're using a Mac.
Something Mac users often don't take into account is that if buying a Mac is even on your radar, you're probably either doing alright in terms of disposable income or, if you aren't doing alright, have priorities way out of whack with those sharing your bracket.
It's a premium commodity -- a luxury premium item -- and you buy it because you can afford to buy it instead of a Dell for half the cost. Also, outside of some content creation areas -- especially film -- it's become heavily focused around the consumer, specifically consumers who place a high value on aesthetics. There's nothing wrong with it, but there are always tradeoffs, and it's always been really, really hard to make the case for a concerted effort for OOo on the Mac that they'd actually accept, yet the Mac base doesn't align well with Sun's traditional goals.
Windows was a no-brainer -- you want to be a player and present yourself as a viable option, you have to be on Windows. Linux and FreeBSD and such worked out fairly well too, via the x11 version. Sun is selling their own Linux solutions, as well as Solaris on some desktop-thingies, and due to the way Linux desktops work, with some tweaking a viable x11 version plugs into all of them well enough, and normally enough.
This isn't the case with Mac OS X, which doesn't build on x11 and instead uses its own graphical subsystem -- Quartz -- which you plugin into via Apple's APIs. This is a major stumbling block (and the Cocoa APIs in general, GNUStep notwithstanding) when getting Unix apps with a GUI onto the Mac, and vice versa. They do have an X11 environment ported to OS X, which Apple has tied to Quartz for a degree of acceleration, but it's native only in the sense of not being emulated.
An x11 app sticks out in a major way on OS X, although if you are coming to the Mac from looking to use unix scientific apps, it's doable and generally works well enough, but you really have to know what you want to be using and be willing to work within its limitations. This, unfortunately, excludes most consumers, who aren't even aware of the x11 environment, excepting a few here and there who downloaded a Gimp package, fired it up, and then forgot about it. They might get x11 OOo up and running, and be happy they can do some things, and then they might end up in font hell, among other things.
Productivity suites -- pound for pound -- are in the running for the most complex code bases on any platform. Microsoft's Mac Business Unit certainly doesn't get enough credit for getting their stuff over to OS X on day one, let alone when x86 Macs ship, and you have to keep in mind:
- While you may not need to make use of a lot of their features, a productivity suite does a hell of a lot of things, which comes from a hell of a lot of code. In OOo's case, we're talking a word processor, a spreadsheet app, a presentation app, a drawing app, a database app, an equation editor, and more depending on your platform.
- The need for productivity software (word processing, spreadsheets, databases, presentations) has been around forever, and you don't build a productivity suite overnight. These code bases go back a long, long way.
StarOffice was originally cross-platform through a proprietary C++ library, and it does a hell of a lot of things internally rather than using platform-specific APIs. Strategies for how to handle cross-platform development always seem to snap back between different poles over time, but at the time it was created, the idea was to do as much as you possibly could within the app itself so you weren't bogged down by a bunch of platform-specific junk. As times goes on, this can be problematic, because what you're competing against may be doing everything it can to leverage platform-specific features.
I.E., you may have a built-in spellchecker that works on all the platforms, but Mac users want it to use the system-wide spellchecker so it picks up the words they've already added. They knocked over a lot of pins with the x11 version, but that still left Windows and Mac OS X. Whatever Sun did, it had to have a native version for Windows that wasn't all weird and whacked out, and they did put a ton of effort into making that happen. It's not perfect, and you're not going to sing along while using it, but it's native and it works.
And then there's the Mac.
Sun actually looked at porting StarOffice to the Mac, which meant OOo on the Mac, and met with some fairly disastrous results. For all intents and purposes this was a horror show.
At the time, Sun was really trying to get some traction for StarOffice, and was getting none. Anywhere. When they announced they were all hot on OS X, I don't think they were able to get one major PC maker to jump on board and bundle it. They really liked the idea of Apple bundling a native-ish version (using Java) with their machines.
Imagine if you would, getting Office X to OS X without the Carbon APIs, and you can have an idea. The OOo code base expected a hell of a lot of things to work a certain way, and included a hell of a lot of things, and matching those up to how OS X expected things to work was a scenario leading to a rubber room. This isn't an accurate analogy, but it gets you mentally close to what they were having to deal with.
It died, and it died hard.
The x11 port stayed around, but even that was a struggle because no matter what some Unix guys say at first glance, OS X likes to do some things differently. Take a handful of guys and throw them several hundred megabytes of code on a System where you can't change a few makefiles and recompile, and you are going to feel the burn. It also didn't get much love, as you had the distinct impression Sun wasn't interested in throwing many resources towards its development, and was hoping Apple would be interested enough to pony up some cash to help it happen -- and this all was excacerbated by the fact that they needed people who could deal (read -- make sense of) with OOo's code base as knowing the ins-and-outs of OS X.
Those people don't grow on trees, and throughout it all, Sun burned about as much karma as you can get by, you know, saying things were coming and being worked on, and then several years later having it come out that they hadn't been touching it much at all.
Two guys associated with the OS X port of StarOffice and OpenOffice, Edward Peterlin and Patrick Luby, took what they'd learned and decided to roll their own, and dubbed it Neolithic Office. Back when a native port was in the works, there were two strategies for getting it up and running:
- NeoOffice/J
The idea here was to internally map what OOo was doing to what OS X expected via Java. This was "supposed" to be sort of a stop-gap while the real port took place. - NeoOffice/C
The idea here was to do what NeoOffice/J was doing, but using Cocoa. It'd be faster, more efficient, and feel native.
The Cocoa version of NeoOffice/C was DOA, and couldn't be run for more than a few minutes before crashing, and that was if they were able to get it launched at all. Ed and Patrick decided to focus their efforts on the Java version (which, just to be clear, only used Java for a very small amount of the code) and ended up getting NeoOffice/J out the door.
This is the only working and viable native port of OOo for the Mac, and it was -- for all intents and purposes -- done by two guys over several thousand hours. It's not perfect, and is a continual work in progress. Just launching it will show you this is an app with a non-Mac pedigree -- it's obvious where it comes from and what is is. However, it uses native OS X printing, the System's fonts, and it has a menu bar.
While it's not perfect, and arguably at a stage where only a mother could love it, it is still a remarkable effort. Adjectives can get thrown around to make things interesting to read, but remarkable is really warranted here, and it's why I first came into contact with Patrick. These guys have dropped huge amounts of time into a project that may not be acceptable by Mac standards, but is remarkable for existing at all -- and it's in serious, serious trouble.
NeoOffice/J exists outside of OpenOffice.org -- it isn't counted among their projects, but it is entirely dependent upon it, sharing 98-99% of its code base with the Mac OS X x11 port of OpenOffice. Because of this, it was always going to lag behind the x11 port a bit, and all the others on other platforms.
To make matters more complicated, licensing issues presented themselves. Originally, Sun muddied the waters in a big way by having two licenses for the projects: Sun Industry Standards Source License (SISSL), and the LGPL. Around a month ago, Sun decided to retire the SISSL for OOo and make it entirely LGPL, but to put it bluntly they'd made enough weird decisions and sidesteps that they were wigging people out, and strange things started squeezing out in the sides.
A directly applicable example is that NeoOffice is GPL, yet dependent upon OOo, which is LGPL. Search the site for LGPL for a better explanation, but suffice to say the OOo guys can't just go and use NeoOffice code in their project even though they're directly related. This in many ways has complicated the relationship between the teams, but it's understandable -- you have a few guys who felt abandoned, then put their blood and sweat into pulling off a hat trick, and don't want a company swooping in once much of the heavy lifting has been done and profiting.
The above licensing situation hasn't always made the relationship between the two projects hunky-dory, but we'll get into why the above is a problem in a moment, as there are other things chafing NeoOffice in a big way and they all interrelate:
- We're really talking about two guys here, with ancillary help along the way, putting in thousands upon thousands of hours of work to get NeoOffice to the state it's in. They're also about the only two guys who have a clue as to what's going on.
While it's OSS, due to the size and scope of the project, it might as well not be. This isn't the type of project where you can notice a problem, drop into the source (several hundred megs I believe) and submit a patch -- you are looking at devoting a hell of a lot of time to even figure out what's going on before you're doing anything worthwhile. The type of programmers they need don't grow on trees, and the ones who could help simply can't devote the time to get up to speed.
- NeoOffice was originally going to be a test-bed, a sort of proof-of-concept to figure out what did and didn't work before something real got started. Unfortunately, for all the talk, nothing real ever got started and NeoOffice/J ended up being the only viable offering. Sun had stated they weren't going to be doing it, the OOo group wasn't going to be doing it, so that left NeoOffice/J to do it, and they set about trying to do it.
- Two guys working in their spare time can only get so much done, and they were actually starting to build some interest with their last releases, which meant getting donations to hopefully devote more time to the project as well as trying to get a DTK (developer transition kit) from Apple in order to get the thing working in x86 land, which costs $1,000.
Lots of things come up in a project like this, like having to move to the newer 1.4 Java Virtual Machine in OS X and such before the switch, or adding SpotLight support via NeoLight, instead of further work on improving its integration.
They're already strapped, but they were building some momentum. And then in a presentation at OOoCon, OOo announced that they were going to be dumping x11 and doing their own native, Cocoa-fied version of OpenOffice for the Mac, and NeoOffice was back to square one.
Overnight, that comment went all over the place, and NeoOffice's momentum was on life support if not dead, which brings us back to the email I referenced. It's worth noting:
- Who was CC'd, as a few addresses stick out.
- That I know Ed and Patrick from awhile back when they made my radar, because they're the kind of people I think need more light on them. I was looking at doing one of my chewable interviews with them, and the above is just what I picked up researching, not from their mouths.
Now, if you do the math on who sticks out in the list of recipients -- using the only real stick they have to not have it relegated to a far-off page on the site (fyi, the page was updated with more info, but the NeoOffice part wasn't removed) -- you can get an idea of just how hard they're fighting to try to keep the project alive and just how much trouble it it's in. They're starting to play a very dangerous game, and smart people don't do that unless they are out of options.
I'm not going to go so far as to say it's obvious Sun isn't down with having NeoOffice gain traction anymore, but I will say NeoOffice was worthwhile to Sun only as providing as an excercise in mindshare. Since it's GPL and not LGPL, Sun can't producticize it as they can OpenOffice.org, and there is a reason why most of those within OOo are on Sun's payroll in some way. Not all, as there is a community and people do contribute some code (the original OS X port was volunteer-led, with some resources provided by Sun), and some other corporations like Novell and RedHat do cycle in some dedicated developer time.
However, when Sun starts talking tens upon tens of thousands of people involved in OpenOffice.org, it can be a little misleading, as while there are volunteers for the most part they aren't contributing much to core development of the project -- that is funded by Sun.
Either way, NeoOffice is in a real pickle:
- Their project is heavily based on the X11 OS X port, and with that shutting down in favor of another Cocoa-fied port, NeoOffice is going to have that much more work as new versions come out.
- When it comes to funding, it's awfully hard to get people to sign on and contribute funds to a project when the Big OOo Hope™ is just around the corner yet again. They've done a lot of work gearing up to get it to x86 Macs, but effort doesn't put a Developer Translation Kit in their laps to actually test it.
- When it comes to getting people interested in learning the code in the hopes of bringing on other contributors, this again becomes a non-starter because no one wants to latch their horse to NeoOffice when the big brother might be coming next year and relegating it to a sideshow.
- When it comes to picking up user interest, which is often intertwined with the preceeding two issues, it's awfully hard when as far as the public is concerned, the real native OpenOffice version is on its way, and there is just some weird experimental project floating about.
To be fair, NeoOffice isn't entitled to exist, despite the amazing efforts of its founders, and if OpenOffice.org wants to shut down their X11 port and create a Cocoa-fied port that's all fine and good. A Windows-quality port would certainly be sweet. The problem is that a lot of this doesn't make a whole lot of sense:
- There are serious technical limitations inherent to the OpenOffice code base, and Cocoa, that make them extremely unsuited for each other, which the original porters hit hard. No one has said why this time would be different, as nothing has changed in the code base that would change things, and nothing has changed much on the OS X side of things that would change the equation.
- Even assuming there was a will and a way, there can't be that much will and there can't be that much way. At best, we're looking at a year until we're seeing anything real, and at worst, several. This is non-trivial stuff, and when I asked around about who would be doing this Cocoa-fied version of OOo, I wasn't put at ease. We're again talking a very small handful of people at best, who are solid contributors but not necessarily who you'd flag for a crack Cocoa-porting team.
However, the real problem is that Sun -- and OpenOffice.org -- burned through their karma long ago when it comes to a native OS X port, having several 'announcements' bandied about that turned out to be vaporware, and haven't come close to saying why this time would be different.
Considering they're doing a bang-up job of asphyxiating a project which actually works now with their announcements, I think it's perfectly fair to ask why we should think this time will be different, and if we aren't given that, whether these types of announcements -- while within their right -- are actually right.
Posted by drunkenbatmanOctober 31, 2005, at 11:37 AM
Comments (55)
Posted by: Romain at October 31, 2005 12:26 PM
Given the importance of MS Office/Mac for Apple I doubt the OOo team gets any help from Cupertino.
Or maybe there's someone at Apple who knows how long it will take MS to come out with MS Office 13 and thinks it is possible to build a working Cocoa OpenOffice in the meantime.
Anyway. Neo/J is mighty good for what it is.
Posted by: stinksoup at October 31, 2005 01:00 PM
Hmmm... do you think that the smell of Google Dollars has anything to do with this?
Posted by: hbdesiato at October 31, 2005 01:17 PM
Yeah... I was a little over a 1/3 of the way into the blog when I realized what at luxury it was to be spending any time at all reading it. I'm so disdainful of anything that can be characterized (even backhandedly) as a premium commodity - a luxury item, that I could read no further and I must stop and basically insult anyone for wasting their time reading the blog when they could be reading some absolute drivel some place else.
Posted by: Thomas Brewer at October 31, 2005 01:35 PM
Yeah... I was a little over a 1/3 of the way into the blog when I realized what at luxury it was to be spending any time at all reading it. I'm so disdainful of anything that can be characterized (even backhandedly) as a premium commodity - a luxury item
As a long-time Mac user with a background in economics, he is right on the money and I take no offense. The Macintosh is the BWM of computers, a luxury item. Unless you don't think computers are a commodity as yet?
He's also right that the traditional Mac consumer will not be happy with an X11 port (never tried it myself and no plans to), and Sun targets the Enterprise and business rollouts. Perhaps with Java Desktop this will change?
Posted by: Hal at October 31, 2005 01:52 PM
hbdesiato, you're why people Mac fun of Mac users.
To the post, with Office X and iWork and Mesa and Nisus and other options, why should we care about Open Office or NeoOffice at all?
Posted by: Adam Rice at October 31, 2005 01:59 PM
Thanks for this article--it cleared up some issues that I was vaguely aware of, but didn't really understand.
Question: given Sun's history, and the low likelihood of a CocoaOOo coming out anytime soon, why might Sun be pushing this story?
Note: you consistently use "i.e." when you mean "e.g."
Posted by: aaron at October 31, 2005 02:39 PM
I guess I'm missing something in the arguement.
1) Sun dropped Mac support years ago and hasn't shown an interest outside of marketing FUD. NeoOffice has stepped in to fill a void that is negligable (what kind of adoption does NeoOffice have compares to MS Office or even Pages?)
2) You (DB) believe that they have no intention or possability of supporting OOo in the future.
3) Sun is announcing revived support in the future. This is effectivly killing NeoOffice.
Why? It's not like Sun OOo is competing in the Apple market. No company does something like this without motivation.
I appreciate the facts and background but what's your opinion of the facts?
Posted by: hawaiiano at October 31, 2005 02:59 PM
db,
Great write-up about pluby & ed. And it brings a lot of light on one of the most underrated projects around. Love the details.
aloha
Posted by: Pius at October 31, 2005 03:16 PM
Actually, while Mac desktops are (IMHO) a total ripoff, iBooks were really cheap compared to other PC notebooks in the period of ~2001-2004. That means that lots of students at my university have iBooks (and students are not normally known to be über-rich). OTOH, I have yet to see a privately owned Mac desktop (apart from Mac Minis, which retail at 330€ at a well-known chain in Germany; but those don't really count as desktops, do they?).
Posted by: fudo at October 31, 2005 03:51 PM
The comments seem to have taken a wrong turn, but; it is possible to enjoy the Mac experience without having bundles of disposable cash. All one needs to do is to be happy with used equipment, rather than new. Me, for instance, I'm running a five-year-old G4, with a processor upgrade. It's teh cool to me, as prior to this, I was limping along running Jag on an 8500.
As for db's post, sad story. I wish I was in a position to help the NeoOffice developers, but I'm not.
Posted by: tjm at October 31, 2005 03:51 PM
This is not the 1st time things like this have happened, and is a larger symptom of Sun needing to get their act together with OO.org.
Posted by: Peter da Silva at October 31, 2005 04:02 PM
"Mac Minis, which retail at 330€ at a well-known chain in Germany; but those don't really count as desktops, do they?"
While I would like to be able to slap something better and faster than a Radeon 9200 in mine, it's kind of silly to say "it's not a desktop". It's not a game machine, but it's easily powerful enough for anything else someone using a "commodity computer" is likely to need.
It's still a bit premium priced, yeh, but it's cheap enough that if you can afford a new computer at all you can afford a Mac mini.
As for OOO, my reaction is "if office suites are so hard to write, maybe that's a sign that office suites are not the best way to go". I mean, in MS Office all I use is Word and Excel, and I only use Word because I need to be able to read and modify Word documents. Is there REALLY a huge need for programs like Powerpoint and Access for people using commodity PCs for commodity tasks?
Posted by: iznogoud at October 31, 2005 04:04 PM
As I know a bit about that cocoa port of OOo, I'd say the work is *really* hard and will need quite a lot of time to be done.
One thing that might solve... Well, help to sovle the NeoOffice/J troubles wouldn't be that the devs start investing a bit of their time to help OOo in their cocoa efforts?
They already help OOo a bit, but what if they started contributing directly for OOo.
I mean, they acquired experience from their long hard job. That might help getting quite good result, instead of hoping simply that the OOo cocoa port simply fells, before coming back and say: "hey, see, you couldn't do it!".
If they collaborated with the OOo team, I think that it would lead to a better position, and everyone could say:
"hey, see, WE did it!"
can't they forget a bit of that revulsions some have about the other team?
Posted by: Colin Barrett at October 31, 2005 04:14 PM
Yet another example of the corporate and OSS mindsets clashing. What somebody REALLY needs to do is figure out a way for OSS projects to make money, so they can pay developers to work, and do it in such a way that isn't just leaching off the corporate teat (i.e. corporate sponsorships). We have a couple of ideas rumbling around for Adium, and they are sound, but the real problem is this:
How the hell do you distribute the money?
I don't have any (good) answers.
Posted by: at October 31, 2005 04:26 PM
drunk,
I know why the luxury comment is in there, but the way its worded will piss off a lot of zealots needlessly.
Posted by: Robert Ballantyne at October 31, 2005 06:07 PM
Thanks for your detailed account. {sigh} I've been using Apple products since the days before we were amazed by Lisa.
I must have a word processor and a spreadsheet that saves in a format that my clients can use. I hate to say it, but Word has always been the superior scribe. I've paid for 2 MS Office upgrades since Word 5.1a (still the best of the lot) simply to work with Mac OS upgrades. Hear this developers:
I do not want to send any more $s to Redmond for anything. What does it take to come up with a full-feature word processor and spreadsheet program? (I was wondering why this open source project was so slow!)
And while you are reading this rant... isn't it time that Mac users have useful voice recognition software. If I could dictate to my computer, and pay attention to what I was saying and not to what nonsense it was writing, it would really improve my productivity. {Yeah, I know, thread drift.}
Posted by: foresmac at October 31, 2005 06:08 PM
I am nowhere near being well-off by any stretch of the imagination, yet I have a 12" PBG4. I got a small personal loan when I started back at Purdue to buy it and paid it of over time. This was partially offset by a big rebate on the iPod I also got at the time and selling my beloved iMac rev B.
While it may seem like a luxury to some, I can't imagine getting most of the work I do done on a Windows PC (I have to use these on a fairly regular basis for various jobs, so I feel qualified to make that statement). That said, I also have a sizable investment in software that I couldn't afford to replace with Windows equivalents (for those that even exist).
As for getting a DTK, why not send a well-worded plea for donations to NeoOffice/J users? if there are 200 users willing to kick in a 5-spot, That should get the ball rolling at least a little, right? Shit, a two-line post Brent Simmons made on his personal blog garnered over 1300 unique hits on a picture I took of him at . It shouldn't be that hard to get enough attention for the problem.
Posted by: foresmac at October 31, 2005 06:11 PM
Well, my bad typing skills botched that last post, but I took the picture of Brent at Adler, and I meant to link to a post I made about how his mention of the pic had gotten over 900 hits in less than a day.
Posted by: Tobias at October 31, 2005 06:38 PM
My dictionary definey luxury as "something that is an indulgence rather than a necessity". For me OS X is necessary to get work done, and I can't understand (does this mean I'm a zealot?) how the other 90% of the world do this with Windows unless they first spend luxurious amounts of time installing third party tools (read "fixes").
Posted by: AkaXakA at October 31, 2005 07:08 PM
Maybe Google can hire those NeoOffice/J guys (who, from your story, seem to rock):
Posted by: drunkenbatman at October 31, 2005 07:15 PM
My dictionary definey luxury as "something that is an indulgence rather than a necessity". For me OS X is necessary to get work done, and I can't understand (does this mean I'm a zealot?) how the other 90% of the world do this with Windows unless they first spend luxurious amounts of time installing third party tools (read "fixes").
I changed this to "premium", because this is not what should be being focused on, and if someone has a problem with "premium", I will backtrack their IP and beat them heavily. I'm not kidding. *points to bat*
Posted by: elmindreda at October 31, 2005 07:18 PM
Ah, the farce continues. I stopped caring about the latest Cocoa OOo announcement quite a while ago, but it seems things were even worse than I thought.
All the best to the NeoOffice/J guys, and thanks for the insight. I like to keep up, even though I've left that world.
Posted by: Dustin Sacks at October 31, 2005 07:50 PM
Bad Sun. Tisk tisk. Please hire these 2 people right away to build the 'official' cocoa OpenOffice.
Posted by: Robert Cameron at October 31, 2005 09:17 PM
I completely feel for the NeoOffice/J developers. I was instantly saddened to hear that their support dropped to nil with Sun's announcement.
As a Mac user, yes, I fully agree that we do pay a slightly higher premium for our hardware, and expect a bit more shine and polish from our software as well. I am appalled by the state of Office on Mac (v. X is barely usable, and I refuse to pay for 2004). I have happily spent my money on better products, and found Mellel to be a far superior word processor. I've also paid for iWork just for Keynote 2, which I enjoy more than PowerPoint. (I still have yet to find a spreadsheet that I like, but at the moment I don't have much need for one.)
However, the difficulty in porting OO.o is not that we feel there's demand for an open source presentation program. The problem is that you cannot port the word processor portion without porting almost all of presentation part, too. Much of the work of the suite is handled internally. That means that fundamental libraries that may provide the bulk of the presentation part serves only a small part of the word processor portion. But the whole library must be ported in order for the word processor to run.
Also, IIRC, NeoOffice/J is based on the 1.x code base, and OO.o is currently at 2.0, and the point increase introduced some architectural changes as well. It's really a shame that OO.o did not have more modular beginnings to facilitate this in a much easier way. But, in any case, I applaud the work that they are doing to bring an open source office suite to the masses.
(I wonder how immensely long an audit of the code of the 2.0 OO.o release would take, and flag portions that could be implemented as native system calls ... and slowly crafted into a gorgeous Cocoa application ... as I feel the developers are probably slowly making progress towards ...)
Posted by: Justin Williams at October 31, 2005 09:32 PM
Two things:
1. It's a shame that the majority of the comments in here are focusing on DB's choice of using the word "luxury" to describe the Mac.
2. Politics and power plays seem to be the norm when it comes to any large open source project. Tack on the fact that a large corporation is involved, and the politics and power play level is too high for anyone to deal with.
Posted by: dhaveconfig at October 31, 2005 11:53 PM
It sounds like it would almost be easier to start a new OSS office suite for the Mac that could read/write OOo documents than it would be to continue with NeoOffice or OOo/Cocoa?
Posted by: SamR at November 1, 2005 03:12 AM
s/aestethics/aesthetics
s/work and progress/work in progress
On 'donation momentum': How does this impact the ability to have a volunteer engineer? It's a genuine question.
On the license: I'm not sure what DB's driving at with this one, but if there were only two primary contributors to NeoOffice/J, it should be relatively simple to relicense as LGPL. Unless those two contributors are more interested in license politics than OS X office suite availablility?
I understand and respect the argument that the contributors have the right to choose the license under which their work is made available; I disagree with the implication that Sun is intentionally hindering the application by choosing an incompatible license.
On use of the term 'clusterfuck': One of my favourite words.
On OOo in general - my gut feeling is that there won't ever be an 'OS X port'. I would suggest we'll probably end up with a native (hopefully Apple?) application that reads and writes the ironically nick-named 'OpenDoc' format. I realise this is a crapload of work, but suggest it may well be comparable to maintaining a continuous port of a complex unix app.
Posted by: peterk at November 1, 2005 03:59 AM
Sorry, NeoOffice doesn't cut it for me, because I need to print embedded .eps. Now it's not their fault, several other little projects have run into the problem that you just can't get a hook deep enough into Apple's Quartz/pdf API. So I've gotten used to X11 interfaces, Gimp, FontForge, etc. OOo2.0-X11 however suddenly can't see my old faithful Type1 PS fonts...
If you want an example of an OSS project with a "nice" Mac port, look at Amaya, the W3C (mostly) compliant web authoring and proofing tool. V.9.2.2 shows what can be done. And while you're shaking a stick at Sun ;-) shake one at Apple too. Seems to me that Apple's X11 shouldn't just sit there as a virtual environment, like Classic, but it should hook a lot deeper into Quartz/Aqua to make all X11 apps look and feel more Mac-like.
It can't be too hard, Amaya and NeoOffice did it.
Posted by: VIM-PoSSE at November 1, 2005 05:32 AM
On the license: I'm not sure what DB's driving at with this one, but if there were only two primary contributors to NeoOffice/J, it should be relatively simple to relicense as LGPL.
From forums, problem is the OOo group view NeoOffice to be freeloaders who should have no standing with the group because their code cannot be reused. Sun can get no value so cares little about the project. True if it would be LGPL situations could go different.
Posted by: Mike at November 1, 2005 05:46 AM
I don't agree with the comment several people have made that Pages is a replacement for an office suite such as Neo Office. An office suite encompasses much more than merely a word-processing program.
And, frankly, despite the hard-sell in the Jobs presentation, Pages leaves plenty to be desired. It's really more a low-end desktop-publishing program than a word-processor. It would be great for someone turning out regular newsletters for a church, or something like that. But, as a word-processor, it's limited. And even more to the point: why lock oneself into *yet another* proprietary format? Pages will export to .doc (though not very well). But, after all, .doc is proprietary, too.
I simply don't believe that a .pages document is a good bet for long-term storage. If word-processing is someone's primary "office" need I think he or she is better off using Abiword, which will save in .sxw, and moving to OpenDocument when that's fully supported.
Posted by: Small Paul at November 1, 2005 08:40 AM
Well, all I know is that NeoOffice/J has been a godsend to me in working with Microsoft Office documents, for free, on my Mac. I'm really sorry the guys behind it are having such a hard time.
If I find a spare $1000, it's theirs. I'll certainly make a smaller donation in the meantime.
(Oh, and on the whole luxury/premium thing: this one time I was in Caffe Nero on Regent Street, just down from the Apple Store London, and this guys comes over to talk to me about my PowerBook. This is because he was thinking of buying one, as he sometimes gets the urge to buy things to help himself through hangovers.
I probably should have killed him and stolen his wallet, but I firmly believe that, because I refrained, one day fate will make me that rich.)
Posted by: Tobias at November 1, 2005 09:12 AM
The post reads as if /J was started after the first native port had failed. Then why didn't those two guys do it within OOo and using LGPL? If they couldn't convince the rest of the project back then, maybe this is the backlash.
It sounds as if OO announcements are not necessarily thought through at Sun but made by key figures who may be less than objective.
Posted by: ha! at November 1, 2005 01:24 PM
the cocoa port will not happen in any timeframe worth mentioning. this is post-xp Microsoft saying 'we are working on longhorn'.
IF, which is a big if, this comes out, i would wager it will be for oo.o3.0 sometime in '08.
by then, neo will be so smooth no one will know why sun bothered.
Posted by: icedtrip at November 1, 2005 01:33 PM
Great write-up. although i knew some of the background on this, you have pieced together a lot of questions i have had over the years. thanks.
as for x11 in Mac OS X. there have been several times i have used x11 apps. some are not that bad, albeit a little ugly compared to what you have in OS X, but Apple's implementation of their x11 front end app using the quartz wm makes most anything usable. i have personally used gnucash, xchat, and several other x11 apps extensively.
yes, a native cocoa app would be nice, but we don't even have a lot of native os x apps written in cocoa just yet. several are still in carbon. oh well, here's to another several years into the future.
Posted by: Wes McGee at November 1, 2005 02:01 PM
I'd hate to continue to derail the thread on what may or may not be a luxury, but I live in a neighborhood where a quarter of families are using computers with speeds in the low hundreds of megahertzes running Windows 95 or 98 if lucky. These machines are usually purchased from either yard sales or from the van forever parked in front of an ethnic shopping center. These are folks who can't afford parting with $400 or $500 at any one time, and saving up for it, while possible is difficult, and credit is difficult to come by (which prevents them from going Dell as well).
Posted by: nessence at November 1, 2005 02:08 PM
Google is hiring two Mac developers. However, I'm sure google has way more projects lined up for Mac Developers than anything with OpenOffice (Google Earth is a big request for Mac). Just food for thought.
http://www.google.com/support/jobs/bin/answer.py?answer=27557&query=cocoa&topic=0&type=cocoa
http://www.google.com/support/jobs/bin/answer.py?answer=27558&query=cocoa&topic=0&type=cocoa
Let's also keep in mind pages+keynote is $79, not $399 (Office), $699 (InDesign), or $945 (Quark). Anyone with a powerbook or a mac mini can probably drop $79.
Posted by: Michael at November 1, 2005 05:09 PM
Let's also keep in mind pages+keynote is $79, not $399 (Office)
Well, so it ought to be cheaper: it does so much less.
And if price were the only consideration, Neo Office J costs $0 - unless one can afford/chooses to donate.
But for me that's not the issue. The price doesn't matter so much, and I'd have even paid Apple more. But this stinks:
http://tbray.org/ongoing/When/200x/2005/09/18/Apple-XML
I've now downloaded and played with Neo Office J, and I like it. The wp seems a little better than Abiword, too. I'd prefer it to be fully native not Java, but what I'm being offered here is open formats. That's a good thing from the POV of portability and long-term storage.
If Apple add OpenDocument support to Pages, I'll even cough up again. Not that I haven't lined Steve Jobs's pockets enough already ... but - what the heck? - Apple products are "luxury" products I'm told. But the way things are looking right now, that's about as likely as Apple's supporting ogg vorbis for the iPod.
Posted by: Dave at November 1, 2005 10:06 PM
I've never used Open Office (can't face that download :-)) but the fact that even with Sun pushing it it hasn't dented MS Office's share makes me think the thing must be a bit of dog.
Posted by: drunkenbatman at November 1, 2005 11:23 PM
Let's also keep in mind pages+keynote is $79, not $399 (Office), $699 (InDesign), or $945 (Quark). Anyone with a powerbook or a mac mini can probably drop $79.
To be even fairer, Pages is a horrible word processor, and it was never intended to be its primary role back in the NeXT days -- it's a page layout app that just happens to have a few editing capabilities. When I looked at it, it couldn't even do footnotes.
Posted by: dhaveconfig at November 2, 2005 01:56 AM
huh?
Pages certainly does footnotes...
I've actually switched to it for all my academic writing. I don't mind living without EndNote, and I've always found Mac Word since 5.1 frustrating (although the latest incarnation is quite decent).
I see Pages as a word processor with basic page layout capabilities, not the other way round...
Posted by: drunkenbatman at November 2, 2005 02:20 AM
Pages certainly does footnotes...
Yeah, you're right, I spaced and mixed up my wording -- I meant endnotes. It was incredibly slow, and Word isn't exactly a screamer in OS X to start. I needed endnotes, and that was a no-go. And then I needed paragraph autonumbering, and line spacing... Oh. My. God. no fixed leading. Line spacing is kind of important, and while this won't show up a a detriment with a lot of "normal" text it is a deal-killer when it comes to needing some characters larger than others without blowing out spacing.
I see Pages as a word processor with basic page layout capabilities, not the other way round...
We'll have to agree to disagree. :) For something like a newsletter or every day stuff, I think it rules. If you're trying to throw together a book or a large paper, no way.
Posted by: at November 2, 2005 03:01 AM
Wow, 10 pages of text for for an application I've never heard of and will never use until it is out of beta?
Posted by: remco at November 2, 2005 05:41 AM
when talking about open-source/cocoa word-processors there is one that nobody seems to mention here abiword
http://www.abisource.com/
i do not do a lot of word-processing but it seems nice to me.
About spreadsheets i use them at work but i often wonder if they are really that good of an idea . maybe a different way of doing the table/calculation thing will be nicer. I think having different apps for different things is better and more suitable for open-source
Posted by: Peter da Silva at November 2, 2005 02:08 PM
An office suite encompasses much more than merely a word-processing program.
Yeh, but for 90% of the users, that's all they actually want or need, and toss in the spreadsheet and you've got 99% covered. Hell, some of the applications in Microsoft Office have negative value. Powerpoint, for example. Oh god, I dearly wish people were still forced to mark up slides by hand...
About spreadsheets i use them at work but i often wonder if they are really that good of an idea . maybe a different way of doing the table/calculation thing will be nicer.
Back in the '70s I nearly invented the spreadsheet, but I got sidetracked by this idea that what you really wanted was freeform text with embedded live equations (what can I say, I was an engineering major), and by the time I figured out how to do that (on an Apple II, using Basic and later Forth, and no idea how to really lay out text) not only was VisiCalc on a roll, but Tk!Solver had demonstrated how horrible an idea that was.
So...
You're probably right, there has to be something better than spreadsheets... but if there was something obviously better than spreadsheets someone like me (but savvy, instead of geeky and naive, and probably a lot smarter) would have come up with it in the 25+ years since...
Posted by: Nat at November 2, 2005 06:27 PM
I own a small business -- when we started 2.5 years ago, we couldn't afford Office -- & the Quattro & WordPerfect stuff we got bundled with the Dells was terrible. So we use OO.o on XP Pro. (Just upgraded to OOo2.) At home, my wife & I have Macs -- mine is a B&W G3 with an upgraded G3 CPU, hers is a dual G4 (which she needed for video work). Apple software was generally useless. I've had a million headaches over the years trying to use Office for layout, and am fundamentally opposed to Microsoft anyway. My wife didn't feel like we could afford Office. So I installed OOo on both our machines. I found it too slow to use on my Mac -- the X11 translation was just too clunky. She found constant printing headaches (& wasn't geeky enough for the X11 to be comfortable to use). So we both use NeoOffice at home and love it. Nearly every day I make a case for OSS--and get people to--or start--with OOo or NeoO.
Of course, it would be great to have an OO.o-like thing run natively on Macs using Quartz. If Sun does produce something (using, say, Google money -- to hire the NeoOffice boys, e.g.) it'll be terrific. If it doesn't come out of this smelling like roses, it will do irreparable harm to the OSS movement. And given db's history--one of the best summaries I've read (thanks db!), it doesn't seem likely. Sad.
Posted by: Uli Kusterer at November 3, 2005 05:27 AM
>> To be even fairer, Pages is a horrible word processor, and it was never intended to be its primary role back in the NeXT days
Your mixing up two apps there. When Pages came out there was lots of talk whether it was the old NeXT app, but it isn't. Apple just snatched themselves that tasty name.
Posted by: Long-suffering Sun Observer at November 3, 2005 04:13 PM
Sun has a long history of screwing up practically every software product/project it's ever touched (with the possible exception of Solaris). They have a habit of buying out small independant companies with great products in a huge fanfare only to leave them rotting on the shelf. Sun's software strategy has all the consistency of direction of a weathervane. I would be very surprised if OOo in Cocoa ever sees the light of day in a usable state.
All Sun will succeed in doing is undermining the good work that Ed and Pat have done.
Google should do us all a favour and buy up Sun then asset strip them.
Personally I would much rather see a written-from-scratch, native Cocoa Office Suite rather than some cruft-laden port from yesteryear.
Posted by: Jon H at November 6, 2005 02:16 AM
1. I think they're jumping the gun on getting a DTK. Is it the best use of their resources to *rent* a machine? Now? Is it that important that they have something to ship when Apple ships?
No, it really isn't.
Have they run out of other work to do on it? No. I suspect they have far better things they could do to improve the software in the meantime. If they get a DTK donated, great. Otherwise, they should chill, improve the software, make sure it at least compiles for Intel, and wait for production machines to ship.
Perhaps next June they'll be able to order an x86 Mac Mini for $450, or an iBook for $1000. Which they could *keep*.
Really, I don't see the urgency of getting a DTK. It's not like PowerPC Macs are going to stop running next June, and Apple is going to continue selling them.
They should just make their suite kicks ass on PPC, so that x86 Mac buyers will be impatiently demanding their own version.
2. Regarding the luxury status of Macs: Yeah, I suppose. But there are Windows PCs that are at least as expensive as Macs, and sometimes more so. Those are *really* luxury purchases - the buyers could get equivalent functionality from far cheaper machines. At least Mac users have that going for them.
3. Sun's Cocoa port: Its best chance of success would probably be for Sun to dust of their old Lighthouse Design NeXTSTEP applications, update them to Cocoa, change the names, and use that code as the basis for any suite components not represented among the Lighthouse apps, which would be built new.
This would be especially effective if any of the old Lighthouse developers are still at Sun. (The only Lighthouse person I know is still at Sun is Jonathan Schwartz.)
Posted by: Jon H at November 6, 2005 02:25 AM
Michael writes: "But for me that's not the issue. The price doesn't matter so much, and I'd have even paid Apple more. But this stinks:
http://tbray.org/ongoing/When/200x/2005/09/18/Apple-XML"
Bray's point appears to be that, if you aren't going to hire a bunch of high-priced XML consultants, you should use XML, and if you just need a file format, you're a crook bastard.
Well. I don't recall seeing *that* as the rationale for XML's original development.
Posted by: JKT at November 7, 2005 12:37 PM
Hi, I'm involved in the user support at trinity for NeoOffice (i.e. I'm no developer, just an end-user who helps out) and I want to clarify a few things that many people have posted:
1. Patrick (and, IIRC Ed) did work for Sun in the past (on the StarOffice for Mac project, too), so when saying "these guys should just work for Sun/Sun should just hire these two guys..." be aware that Patrick and Ed already know the score with respect to Sun and just may not be interested in working for the company anymore and vice versa.
2. Patrick and Ed are not going to touch the Cocoa "port" with a barge pole because a) Ed has already been down that road - he was one of the two developers working on the original attempt for christ's sake - and he knows exactly what is involved and why it will (probably) go absolutely *nowhere* and b) it will be re-inventing the wheel as it will have to duplicate completely all the work that Patrick has done to get NeoOffice/J where it is today and why should they bother with that when NOJ already does it?
3. If Sun really wants a native Mac port, then they would be better off supporting the NeoOffice project instead of trying do repeat everything NeoOffice/J has already achieved again. It should also be stated VERY, VERY CLEARLY, that SUN HAVE NOTHING TO DO WITH THIS PROPOSED COCOA PORT - it is a VOLUNTEER EFFORT AT OOo, NOT AN OFFICIAL SUN PORT!!
4. Wrt printing .eps - I believe the switch to Java 1.4.2 has allowed support for this in the next release of NeoOffice/J.
Posted by: Peter da Silva at November 9, 2005 09:33 AM
XML
AUGH
Let's take SGML, make some syntax changes that make it a lot harder for humans and slightly easier for computers, and then make the content subject to so many complicated rules that even major software companies have to hire contractors or use commercial libraries to successfully generate and parse valid XML.
Blow that for a joke, you might as well use ASN.1 or RPG layouts on 80-column card images.
Posted by: heayvboots at November 18, 2005 07:18 PM
Wow, I was just looking up the neooffice website and stumbled across this. That is amazingly depressing because Neo is really, really great software, imho. There are some rough edges in configuration, but I'm actually getting some users interested in it here because Word and Excel tend to blow up frequently & randomly on incoming files from clients (the major thing we use it for is just viewing crud other Office-ites send us). It would be a real shame if Neo disappeared because, as mentioned, it actually works!
BTW, one easy way to "Macify" it would be to "un-grey" as much of the GUI as possible. For example, I immediately change the app background to a white-blue and people immediately feel more at ease with it.
I'm definitely praying for Neo to get enough traction to continue!
Posted by: Late to the conversation at December 15, 2005 03:09 AM
I agree with an earlier poster that I am utterly uninterested in the broader range of the office suite functions built into the software. I am interested in a program that is OSS, and is the same price as NEO, and just a solid word processor.
I've used Neo to some degree, and find it to be very slow, but, I use it because I agree with OSS and don't care to support MS. I will never make a presentation with it, a spreadsheet, or a database. Just papers and letters.
I generally understand that a Mac may cost more than a PC on the initial purchase price. But for that price increase comes a lot in stability and in the end, usability and simplicity in the user experience. As a convert of almost 5 years who has had to use MS OS throughout that time, I rejoice at the lack of viruses and the like, and the fact that when I use my computer, it doesn't freeze or crash on me. From my initial purchase price of just over $1100 (yes it was not the top of the line) I wouldn't trade my experience with my friends who suffer through their $800 Dell or whatever for that price difference.
Posted by: depuis 8 mois at June 18, 2006 03:00 AM
It's interesting to look back on this 8 months later and see what has come to pass: the new lead of the OOo Mac port unilaterally revoked Ed Peterlin's OOo cvs access and unilaterally moved the NeoOffice listing to a hidden page on the OpenOffice.org wiki, and he has bludgeoned the entire OOo community into walking around with their fingers in their ears, mumbling "We are the only one" and pretending that NeoOffice does not exist.
On the other hand, NeoOffice 2.0 Alpha is the only office suite available that runs natively on Intel Macs ;) Soon with even more Aqua widgets, too!
I find it difficult to believe Sun would FUD an offshoot unless they were serious.