Tux + Linux Items

Help promote Linux and FOSS at the
Sample T-Shirt from the ERACC Cafe Press Store
ERACC Cafe Press Store

Open Source: OpenOffice.org We Have a Problem

I am unashamedly an advocate for Open Source software. I run a small business that offers custom built, inexpensive computers with Open Source software preinstalled. I urge my clients to move to Open Source software whenever it makes sense for them to do so. This includes full systems such as Linux distributions and cross-platform software such as OpenOffice.org Office Suite. However, recently I have been disappointed by a test install of OpenOffice.org for a charitable organization client of ours.

You see, this client is trying to find ways to reduce costs. One method of cost reduction I always recommend is to evaluate free, Open Source software as replacement for expensive, proprietary, Closed Source software. This client, like many others stuck in the proprietary software upgrade loop, is going to have to move off of Microsoft XP Professional sometime before 2014 when it will no longer be supported. Their computers will almost all have to be replaced to be able to run Windows 7 well if they decide to stick with Microsoft. Of course I am asking them to evaluate Open Source, and they are. One of the decision makers has been running a Linux distribution on his new home PC with a VirtualBox install of Microsoft Vista for about two months now. He volunteered to do this so he could be well informed for making a decision about the future direction of the charity's software needs.

During a recent reinstall of one of the charity's office PC systems following a bad Microsoft Patch Day that only affected that PC it was decided to install OpenOffice.org on the XP Professional system rather than the expensive, proprietary, Closed Source software that had been on there. I gladly did this and put on the latest release of OpenOffice.org at the time, which was version 3.1. There were some complaints from the user as it did not do what her previous office suite did. She had taught herself to use key bindings in that other software to do things faster. She did not know how to do these in OpenOffice.org and of course the similar functions are not bound to the same keystrokes. However, this is merely a retraining problem that can be solved easily with some hands-on training time.

One problem that did crop up though was one I had never seen before while using OpenOffice.org on Linux. When this user tried to save some of her Calc spreadsheet documents she was getting an error:

Error saving the document filename-goes-here:
Write Error.
The file could not be written.

This had never happened with her other office suite. I thought perhaps this was a problem writing to a network share. I suggested saving to the local drive. Nope, the error persisted. The only thing she could do was close the document without saving it. She was able to copy the spreadsheet information to the clip-board first, then a subsequent Calc document did save it. So the data was not lost, but this was a regularly occurring problem. When OpenOffice.org 3.2 came out the user downloaded and installed it on my suggestion that bugs would be fixed and new functions would be available. I was correct about new functions, but this bug was still there.

This was now a serious concern and a broadside against my recommendation of OpenOffice.org for this client. I know Microsoft has deliberately made competitor software not work well in the past. I began to suspect this was an attempt by Microsoft to kill OpenOffice.org on the Microsoft desktop because we all know Microsoft is very concerned that its cash cow will be killed by OpenOffice.org. With some research I discovered I was wrong, this problem is not Microsoft perfidy. It is unconscionable, but none the less true, that this is perhaps a long standing bug in OpenOffice.org itself:


I actually found references in the bug Issue 69993 to this exact error message going back to 2006(!). Now, I am aware that this error message may be triggered for several different conditions. One suggestion in the OpenOffice.org forums is a full temporary directory is the problem. But the temporary directory is not full in some of the cases. However, the fact remains that some people cannot on the first try save files to perfectly good file systems with a temporary directory that is NOT full. The fact also remains that these exact conditions have persisted for more than two years through several versions of OpenOffice.org. This problem does not just show up on Microsoft systems. It also rears its ugly head on Apple OS X and I saw at least one reference to this on a Linux system.

Understandably, we advocates for Open Source cannot have this type of long standing bug and expect to be taken seriously. A company that was relying on its software sales to make money and pay employees from those sales would have to fix a problem like this or go out of business. In Open Source we are not relying on sales to make money so some things that would be a software killer in expensive, proprietary, Closed Source software is allowed to continue under Open Source software. The voluntary nature of Open Source development means that in some cases unless someone cares enough to solve a bug then that bug will languish in the source code unfixed. It is not like one has to worry about being fired if one does not fix bugs in OpenOffice.org or most other Open Source software.

Yes, I know that much of Open Source development is now done by folks in paid positions in companies like IBM, Intel, Novell, Red Hat and so on. However, these paid programmers are working for the interests of their company, not the interest of free, Open Source software itself nor the Open Source community of users. While we should appreciate the efforts to develop Open Source by these companies we should also be aware they they do this work out of self-interest for their own benefit. Our benefits from their work in the Open Source community are just side benefits. It is still up to those of us that are voluntary programmers to take charge of bugs, own them and fix them as quickly as possible.

No, I am not a programmer and I do not have time to become one, so suggestions that I fix it myself are not welcome. This is true of the greater majority of software end-users.

Where am I going with this? What is my goal? Actually, I wrote this article as I thought it out. I did not outline it nor come up with an end-game goal for it. I simply wanted to air out what I see as a problem with "free" software development and perhaps get some people motivated to take action in cases where the software bugs are persistent but difficult to reproduce. In this case with OpenOffice.org, perhaps there is finally a resolution for this bug:


But we will have to wait for OpenOffice.org 3.3 to find out. It cannot get here soon enough in my opinion. If OpenOffice.org 3.3 does still have this bug then my client will likely see that as a reason to stop considering Open Source. That will be tragic because they really do need to cut costs and free, Open Source software can definitely help with that. But should the bug persist a "No" will be unavoidable in this case.

This article has this many unique views:

free hit counter
free hit counter

Notice: All comments here are approved by a moderator before they will show up. Depending on the time of day this can take several hours. Please be patient and only post comments once. Thank you.

Edit Mon May 10 18:34:12 CDT 2010: Clarify the sentence about temporary directories by splitting it into two sentences with more detail.


77 comments to Open Source: OpenOffice.org We Have a Problem

  • Captain Canuck

    Actually, it seems like 3.2.1 is the next milestone.
    3.2.1 Release Client came out the other day, check it to see if the problem still exists.

  • Captain Canuck (comment #1), thanks for the heads up! I’ll get the client to download that and check it out.

    Edit: Unfortunately the 3.2.1 “milestone” is not yet an actual production release. I can only use production releases for most of my company’s clients. Including the one mentioned in this article.

  • eug

    Gnumeric 4 Microsoft Windows

    There is a build of 1.10.2 available. Please report bugs here. Here is a list of known bugs for the Win32 build. Note: this build does not work on Win9x and since GTK+ will probably never support Win9x, that is not likely to change in the future.

  • Rob

    Your complaint can just as easily be applied to closed source software. Hard to figure out bugs don’t discriminate based on licensing.

    Perhaps you could start a donation drive for financing the fix of this particular bug?

  • eug (comment #4) thanks for reading.

    The point of using OpenOffice.org is it is an all-in-one replacement for proprietary office suites. It comes bundled in a single install package for Microsoft systems. It is also integrated somewhat. While Gnumeric is a decent spreadsheet application it is not a part of an all-in-one office suite like most of our clients want and prefer.

    That said, I will pose it as an option for this charitable organization client. They may be willing to go for discrete applications instead of a bundle like OpenOffice.org.

  • [...] Open Source: OpenOffice.org We Have a Problem ? The ERACC Web Log blog.eracc.com/2010/05/10/open-source-openofficeorg-we-have-a-problem … http://bit.ly/at2TZ7 [...]

  • Ah, a FOSS “programmer” calling himself BrainInAJar brings out the knives on reddit in this comment:

    Here’s my take on it.

    Who gives a s??t. If you don’t pay for it, and you don’t contribute code to it, you’re a drone at best and should be ignored.

    I write open source because I feel like it (an itch to scratch) and in hopes that other people will make software that I use better. If other people want to use it then that’s cool but affects me not at all negatively or positively.

    What a narrow minded, self centered “take on it”. I contribute my “abundant” free time to help answer questions for free on Freenode IRC channels for Open Source software and Linux distributions. I started a local Linux User Group (LUG) and do the work to organize it for free. Yet, this dink says if I don’t code and I don’t pay, I’m a “drone”. This sort of crap attitude needs to stay out of the FOSS community. It serves absolutely no purpose other than to alienate all of us who are not programmers, and probably some who are.

  • Enginerd

    So you are a great advocate of open source software but not to the point of actually fixing anything? You couldn’t even hire someone to do it? I’m sure it wouldn’t be super cheap, but if you really care about FOSS it’s not a big sacrifice.

  • Mash

    One thing you should remember is that many open source programmers have no great desire to see their program take over the world.

    The motivation for you to “help answer questions for free” may be to help “spread the word”, or “save costs” or something similar. Only you can explain your own motivation. But it’s helpful to remember that the ORIGINAL purpose of open source software was freedom to tinker. Freedom to program yourself, solve your own bugs, etc.

    The motivation for someone to release something as open source can often be tied to this. I give you source to solve your problem. When you find a bug, you fix it and give the fix to me. When you improve the program, you may solve my problems too. We share our programming resources.

    So, you as an advocate, as a person who performs support, someone who “started a local Linux User Group”, what is the motivation for that programmer to help you?

    That in a nutshell seems to be what the rude “BrainInAJar” was saying. From the perspective of some programmers of open source software – you contribute nothing to THEIR goals. A “drone”.

    When I was reading those bugs, the most common request was for a “repro”. I see you saying that you experienced it as “a regularly occurring problem”. I see nothing in those bugs about you offering exact steps to reproduce this problem. Is it perhaps that you can’t find a way to make it happen either?

    Sometimes in these situations you can say “I have this happen about once every two weeks, I’m willing to run debug versions, please help me help you find the problem”. And then you’re contributing something. And then the programmer can provide you with a version of the program that explodes with extra information (instead of the unhelpful “Write error”).

    But I don’t see you contributing that either.

    So, Gene, may I ask… what HAVE you done to contribute to fixing this problem that you so loudly state “We” have.

  • Enginerd (comment #8) thanks for reading.

    What part of “… I am not a programmer …” did you find it difficult to comprehend? I cannot fix broken source code since I am not a programmer. What part of I provide “free support” on IRC and “free support” for a Linux User Group (comment #7) are you unable to understand? Here is another thing for you to understand, if you will make the effort, I cannot afford to hire a programmer to fix someone else’s bugs in FOSS. Most people are in the same position I am in financially. We do good to pay the bills and feed our families.

    Not all contributions to FOSS are done by programmers or moneyed investors that can hire people to fix someone else’s mistakes. Until some FOSS programmers get over their “it is ‘free’ so suck it up or fix it yourself or pay someone to fix it” bad attitudes I suppose we will keep having substandard software with four year old (or older) bugs.

    I sincerely appreciate those FOSS programmers that take responsibility for their code and actually work to make things good for plain old end-users. I will do all I can to help those people do what they do. I’ll even learn their software and provide free support for it when I can. Support is as important as writing code. The others with the nasty attitudes and self-centered thinking, not so much appreciated.

  • Mash (comment #9) thank you for reading.

    *Sigh* I would and do run debug versions of FOSS that I use when I can and I do provide bug reports for them if I see bugs. In this case, I have never seen the problem myself but I did see it on the charitable organization’s XP Professional PC. I have a screen-shot to prove that if you wish. I did not say I personally had it recurring for myself. I personally have never had this happen with my own files on my Linux desktop PC.

    As for testing, I do not have XP Professional on anything but a virtual machine here. The VM does not have enough space to install OO.o and I am not certain that the bug would occur in a VM. But the system in question is not running in a VM anyway. That said, the bug is well reported by others. I even give URLs for the reports on OO.o’s bug tracker. I could understand your attitude better if I were reporting a bug that no one else ever saw.

    Edit: If I am providing free support for users of your software that frees you up from having to provide that support. Support which some programmers would rather avoid I think. :)

  • Mash

    I don’t think you read my comment.

    Cause I included in that something that YOU as a NON-PROGRAMMER could do to help fix the problem. (Offer to run debug versions that would help the programmer find the problem)

    In fact I included two things. One of which you’ve already been asked for. Did you see you had been asked to do it? (I’m talking about repro steps – posted back in 2007 by the people who were trying to fix this issue)

    I understand you are unable to pay to fix the problem. That’s part of the joy of free software – you didn’t have to pay for it originally either. But the programmer has to eat too. His motivation may be related to how much you help him eat. Whether you pay, or whether you write bugfixes for his software that makes his job easier – both of those help him eat.

    Support helps the USER. You are helping many people with the work that you do and the effort you go to. But – are they the programmers of the software? Do you teach the users how to write good bug reports or do you just suggest to upgrade to the latest version? Do you help out on the bug forums, following up with the users to get better repro steps, and maybe testing it out on your computer so that the programmer spends less time there? Or do you just post a “I have this problem too” report that the programmer has to read but doesn’t help him.

    So I ask again – what do you do to help the PROGRAMMER solve this issue?

    PS. Have you ever seen the bug system of a closed source product? Do you know how old the bugs are in those systems? And do you know how many bugs never make it past the support lines of those systems? With the global visibility, open source software does much better than you seem to think.

  • Mash (comment #12) uhm, are you prescient or something?

    I was still writing my reply to you when you posted this new comment. I think you thought my reply to Enginerd was to you?

    Edit: Here is an example of my helping solve FOSS bugs for your reading pleasure: http://blog.eracc.com/2009/01/16/why-i-think-open-source-will-win-in-the-end/ While that is not the only example it is the quickest one I could point to for you. :)

  • Mash

    Yikes, I responded too fast – I thought your previous comment was a response to me.

    Yes, ok, so this is a non reproducible (by you and by the programmers) problem that has been around for many years. You’re being pretty hard on the programmers for a problem like that don’t you think?

    But my point about you testing to contribute was not just about this problem. Contributing is not just about solving YOUR problems – right? Otherwise those programmers would REALLY have no motivation for solving your problem.

    And there is some (selfish) motivation for you to test regularly too. If you contribute regularly by testing, you will become better known to the programmers, and when you report a problem like this, they’ll remember you’ve helped them in the past, and be more willing to spend time and effort to figure out the un-reproducible problem that you seem to be having.

    Anyway – your last comment was much more reasonable. Your original post didn’t seem to understand that the problem you’re complaining about has been around for years because it’s HARD TO FIND. That means (btw) that not many people are experiencing it.

    Open source really doesn’t have a problem because of bugs like this. It has a problem because of people who think bugs like this should receive programmer response of the “drop everything else to help me” nature. And that’s a problem of evangelism. Which is your job. And this blog isn’t helping.

    Please help the programmers Gene.

  • There’s nothing in this world that that discourages me from spending time on software than some user’s indignation that I’m not spending my free time fixing his computer. And that’s really what it boils down to.

    > No, I am not a programmer and I do not have time to become one, so
    > suggestions that I fix it myself are not welcome. This is true of the greater
    > majority of software end-users.

    “I’m not a mechanic, and I do not have time to become one, so suggestions that I fix my car myself [if I'm not willing to pay someone] are not welcome. This is true of the greater majority of car owners.”

    Do you see how ignorant that is? That’s the attitude the BrainInAJar is responding to, and I can’t say I blame him.

    This isn’t to say that it’s not tremendously useful for society to have this extra layer of customer service and technical support, making some of this work available to a non-technical public; it most definitely is. But don’t confuse yourself thinking that you’re doing the developer a favor by doing this. He’s the one doing you a favor.

  • Carey Underwood (comment #15) thanks for reading.

    Okay, you want to use car analogies. I can do that. ;)

    I have a different scenario for you. I did not build a Toyota automobile that suddenly accelerates, crashes and causes people to die. The Toyota automobile manufacturer did that. The problem was one that was hard to reproduce. Yet, Toyota the automaker (software developer? :) ) did the work to find the bug, recalled the cars and repaired them for free. No one had to pay a mechanic for that.

    See, if you write the software then you own the bugs. Even when it is “free” software. If that is an unpleasant reality to you, it is still reality. I do not want to discourage anyone from doing what he or she likes to do when it comes to writing code. However, I do think excellence should be striven for even if the code is “free”. I help find bugs in FOSS when I can and provide reproducible evidence when I can find it. I have done so in the past and will do so again.

    Someone also needs to be willing to report when things are not going well. A persistent four year old bug in a major FOSS project like OpenOffice.org is very not good. I knew I would take some heat for writing this article. I am willing to take it to advance the conversation about where FOSS is and where it should be going.

    One other thing. Not all end-users can take the time to find your bugs by running debug code. It is absolutely not reasonable to expect all of them to. Those of us who can and will do this already do. That will have to do for now. :)

  • All,

    Some people have stated in comments elsewhere (reddit) that I make money off of FOSS yet I am unwilling to support FOSS. There is no evidence for that statement and it is blatantly untrue. I make my living mainly from support of small business end-users. I will get paid for providing support whether or not the end-users have FOSS on their systems. I also make a lot less from hardware sales where I provide a service to preinstall systems for those who cannot or would rather not do this themselves.

    No, I am not making a living from “selling FOSS”. I sell services and hardware. I advocate FOSS and support FOSS, I do not sell it nor make my living from it. I actually earn more from fixing problems on Microsoft systems than I have ever made from FOSS.

  • Brian

    I don’t think the second analogy to cars is reasonable. Toyota did not give the cars away, they advertised them as safe and reliable, and OO.o can not kill you in a fiery “Can not write.” Toyota was acting both in the best interest of their brand and with a legal (and moral) imperative to prevent loss of life. It was literally a case of life and death.

    I think ‘Carry Underwood’ is right in pointing out that the programmer is doing the users a favor. While I don’t think it would be the case with much larger projects, OO.o for instance, putting this kind of negative attention to FOSS programmers could lead many to keep their code on their home machine. Certainly many small projects are useful to niches of people. Do we really want to, as a community, blame the creators of our foundation for problems with language as strong as “this is what is wrong with FOSS?” Is this what we want perspective programmers to see as what to look forward to when your project becomes popular?

    I can certainly understand your frustration at seeing a bug go unresolved, but bug triage is *not* what is wrong with FOSS.

  • Brian (comment #18) thanks for reading.

    Personally, I believe a “repair a car” analogy does not work here either. Because no one I know is giving away cars. But that is what I was given to work with, so I ran with it. My apologies that it didn’t work for you as well. ;)

    The fact is OpenOffice.org is a major project for the FOSS community. It is a “brand name” if you will. Anyone wanting to be involved with a project that important must absolutely be willing to get dirty for the end-users, not just for self-gratification. Sometimes that will also mean accepting blame for long standing bugs.

    Look, FOSS programmer people, I am not against you. I sincerely like and appreciate the labors of love you create for all of us to use. But I absolutely will point out problems as I see them. You are welcome to disagree with me but please do not place intentions on what I say that I have not explicitly stated.

    I am not making a blanket statement that all FOSS programmers “owe” the rest of us anything or are self-centered jerks that only care for themselves. What I am saying is the major projects, like OO.o, cannot have serious problems like this that last for four years. We as a community need to acknowledge this and find ways to fix it. Not behead the messengers (please, no beheading) that bring the bad news.

  • Adam

    Ha! I just ran into a VERY similar issue using windows 7 and office 2007. It refuses to allow me to save to documents that I’ve recently saved to. It affects all components of office also. Don’t even get me started with the outright misfunctioning bugs of visio 2010.

    If your using XP and office 2003, I can see the appearnce of “no bugs” but between everything I’ve experienced after switching from ubuntu back to win7… windows ain’t got nothing as far as being bug free on Ubuntu….NNOTHING

  • Gregory

    The fact that Open Office is a major brand name and headliner for FOSS is a problem, the fact that many don’t see why is also a problem.

    Why? It’s slow, crash-prone, non-native feeling GUI, mangles documents from other software, doesn’t support common key commands (yes, retraining, I know, but MS got this right 20years ago when competing with Lotus ffs), and it’s very, very, very buggy.

    That wouldn’t be a problem if it was software for developers – but it’s not pushed, mentioned or even targeted that way.

    Office Suites are for end users, they are for the people that use MS Office, and those people *might* be able to use Open Office, if all they do is really basic spreadsheet work and average Word Processing, but anything else and it becomes clear that OO.o just isn’t a quality replacement.

    I want to use it, but it is terrible. The “Fix it yourself or shut up” crowd are… well, I’ll put it bluntly, short-sighted small minded idiots.

    Not everyone has the time, the skill, or the resources to debug software.

    I might be an excellent project manager, or Java coder, but some random C++ project that is going wrong won’t get fixed by me. Why? Because my skills aren’t applicable (god knows that most OSS could use better project management though).

    If the point of FOSS is to tinker then stop evangelizing and just accept that FOSS is for toy software that is fun to play with, but isn’t professional.

    Oh it’s more than that? Well yes, obviously it is, so treat it as such.

  • phil

    Volunteer work is always like herding cats. If my cat doesn’t *WANT* to do something, I usually can’t make him do it. Programmers are like that, too. Heck, I can’t even make myself work on my own projects as hard as I would like to. I’d have to be *REALLY* motivated to dive into a non-reproducible bug hunt if the bug wasn’t bothering me personally.

  • Alphaville

    Giving this a tiny bit of thought it seems kind of obvious to me that your client should evaluate OpenOffice.org like it would any other competitor to Microsoft Office. What I mean by that is, STOP suggesting that your client (or anyone) evaluate all open source software based on 1 open source program.

    Yes, I understand you were pointing out a general problem with FOSS, which I can generally agree with. But, if your client uses the OO.org bug as a reason to stop using all Open Source they’ve missed the point entirely. Evaluate each program on its own merits for its own business purposes and see what works best.

  • The trick with car analogies is to make sure they’re exact :p

    My boss has every right to tell me to fix a particular bug. He’s paying me after all. You’re not, and so you don’t.

  • Peter

    \A company that was relying on its software sales to make money and pay employees from those sales would have to fix a problem like this or go out of business.\

    Ha. I laugh at this; Word for Mac has had a bug for 12 years (!!!) that causes saves to fail if you do too many in a single session. I have lost hours to this bug, and when I found out its history, I was understandably pissed.


  • angelo

    I’m sorry to read about your problem, I am a programmer, I do this for a living. I can understand your problem very well. I have been a Linux user for a very long time. The difference between closed source and open source is pretty simple, people who write closed source software have to fix the bugs or they lose money. people who write open source just move on, fixing bugs isn’t fun at all. I wish you luck in getting this fixed, but honestly I have little expectations it ever will.

  • Matt

    XP will not just stop working because it is no longer being supported but “a retraining problem that can be solved easily with some hands-on training time” will cost money.

  • dhany

    “What I am saying is the major projects, like OO.o, cannot have serious problems like this that last for four years”

    Yes they can. Stop being delusional.

  • Mike Harvey

    Go find your OpenOffice exe file. then right click on it and open properties. Poke around the menus, look for a compatibility mode selection and select an XP version.
    Attempt to recreate the problem. I had a similar problem with another app where this solved the problem.

  • I think programmers and non-programmers are somewhat talking past eachother here.

    Gene, you’re not a programmer. The reason why programmers ask for “steps to reproduce” is not because they want proof that the bug actually exists or not. It’s not that they won’t believe you.

    They need it in order to be able to diagnose the problem and fix it. If a programmer cannot reproduce the problem, but only have a screenshot and a description of the Write Error, they can only make rough guesses at what causes the problem and because OO.org is such a big project, that makes it even harder.

    Think of it like this, I had some strange intermittent problems with my TV cable and phone connection. I’d call up the support line, they sent someone over. Of course the problem never occurred while someone was actually there. So all they could do was do a global check of things that might be wrong, signal strength, noise levels etc. They didn’t find anything, but did tune a few bits in the hope it would fix it. It didn’t. They needed to come over five times over the course of several months before the connection happened to fail while a mechanic was present, before they could fix it [and I'm still not 100% sure it's fixed, but it's a lot better now].

    So if you want to help the project, put in some effort on finding a way to consistently reproduce the error. That’s difficult, but it’s not programming work, think of it more like detective work. You can do this even without programming skills. If you write a bug report that describes the steps that someone (anyone) should take in order to consistently reproduce that bug on their own machine–be it in a VM or perhaps on a particular installation of XP–as long as a programmer can use it to recreate the bug at home.

    Only when they can do that, they can reliably pinpoint the cause of the problem and fix it.

    Without being able to reproduce the bug, the programmers are pretty much shooting in the dark, even if the bug has been reported multiple times already.

    And the good thing is, even without programming skills, you *can* contribute by doing some detective work, comparing configurations on clients where it happens and does not, up until you figure out how to reproduce that bug. Cause basically if you don’t do that, it means that one of the programmers that will eventually fix that bug will have to do it.

  • J. Young

    Just for a bit of constructive input: how could this bug be solved?

    I too find the “fix this for me so I don’t lose money” attitude quite troubling.

    When I make money off using opensource apps I always contribute some of my income.

  • eug

    Free Microsoft Office Alternatives or Replacements


    Edit: I fixed that URL for you, eug. I have no problem with URLs of that nature. Gene

  • Mark

    What I am saying is the major projects, like OO.o, cannot have serious problems like this that last for four years.

    Except that they do, and as a 30-year veteran of the software world, I can tell you that those kinds of bugs are a fact of life in every program out there. It doesn’t matter whether it’s open source, proprietary, large, small, insignificant, important or anything else.

    Seriousness is a matter of perspective. You’re defining it by how the problem affects you and your clients, and obviously for you it’s a show-stopper. But if you step back and look at it from the perspective of the entire population of OO.o users, it’s minor. There may be other problems that are affecting more people, and those will be the ones that get the attention. Software projects, be they open or proprietary, aren’t blessed with infinite resources, and those that are available have get deployed in ways that do the most good. The fact that your clients haven’t gone out and bought the latest versions of Windows and Office with shiny new computers to run them says they’re in the same finite-resources boat.

    With open source, you get what you paid for, sometimes more. The value in it is that you have the option of fixing bugs yourself or, if you’re not qualified to do it, paying someone to do it for you. Whether or not those costs are worth it depends on your situation. There’s no magic bug-fix fairy that solves these problems; someone with the skills to do it has to be paid or donate his time to the project. For your bug, neither has happened yet. This is not a uniquely open-source problem, either. I’ve done several projects where I’ve had to negotiate and purchase source licenses from vendors to fix problems they were unwilling to touch, often at a cost of tens or hundreds of thousands of dollars. My clients or employers were willing to do this because they understood the value in it: what they got in business was a good return on their investment.

    If OO.o doesn’t work for your clients in their particular situation, they need to weigh their options, which are to live with the problem, fix it or find another solution. All of those options have direct or indirect costs associated with them; you and your clients have to figure out which one causes the least pain.

    Open source software is a gift from all of the people who worked or paid to develop it. And as my mom always said, you can’t look a gift horse in the mouth.

  • As a professional software developer, reading through that bug report looks like they are making excuses. Of course it is not reproducible. Obviously the problem relates to the environment on your PC, which the developer does not have.

    I deal with bug reports all the time and if I made an excuse that they are not reproducible, I wouldn’t be working very long. The next step is to look at the source code for all possible reasons why a write error would occur, and recover each and every one of them.

    I’m not going to do this because I’m not interested in using OpenOffice, and its source code is abtracted to absurdity. (There’s probably a method like Uno::AbtractIo::Saver::NetIo::Hal::WriteSys::FileWriter::BufferedFileWriter::Save() that nobody wants to look at.)

  • Bj?rn Cintra

    Realizing I might be unpopular with the open source advocates, you could give IBM Lotus Symphony a test ( http://symphony.lotus.com/software/lotus/symphony/home.nsf/home ). I know the current 1.3 version is based on an old OpenOffice codebase, but the version currently in beta is fairly recent, and IBM has done a few things to make it “business ready”.

  • Phillip

    “Where am I going with this? What is my goal? Actually, I wrote this article as I thought it out. I did not outline it nor come up with an end-game goal for it. I simply wanted to air out what I see as a problem with ?free? software development”

    I wish that paragraph was put at the top so I would not have wasted 5 minutes of my day.

  • Bj?rn Cintra (comment #35) thanks for the suggestion.

    I am already evaluating IBM Lotus Symphony as I downloaded it earlier today. The first thing I tried to do was open an existing document on our FreeBSD NFS share. Symphony appears hung hard as I type this. Not looking good so far. However, it may have a problem that is also in OO.o with file locking on NFS shares. Now to find a fix or workaround. :)

  • Sean

    Good article and illustrative on a number of points. You are showing why any technical support person gets upset at the end-user when the problem is hard to reproduce. From a support person perspective, we can try to communicate exactly what caused the issue (You are running in a VM, how many times was the document saved ? What was the name of the document? etc.). Having had to submit bug reports myself and work off of them, I find that a high percent of bug reports are useless because they do contain any information beyond the perfunctory “It Broke, fix it.”

    As a thought, are you married to Open Office ? Will KOffice meet your needs? KUbuntu installs and runs nicely in a vm. I also saw that someone suggested a part of the Gnome suite (AbiWord/Numeric). You may want to evaluate the workflow of this charity to see if the individual applications still meet all of the needs of your client.

  • Fred

    This could be the cause. When using WinXP the user has only user rights and want to save in a location without permission. The solution, save in a location the user has access rights or make that location accessible for the user.


  • anon

    Status: VERIFIED
    Resolution: FIXED

    get the commitdiff/apply patch/build/done

  • Robert

    If it wasn’t for the interoperability of OO.o with the cashcow product from microsoft it would have died pretty soon after being opened imo.

    I look forward to OpenOffice 4.x pushing for a modular build, it will be painful but I think OO.o needs to drop some of the painful to maintain codebase.

    (I have submitted and resubmitted a bug which took which was never truly fixed, an option was just added to disable the feature which was to blame for the suite hanging, which was better than nothing)

    I’m quite glad Nokia is pushing time/money into KOffice, it is better than OO.o on underpowered devices. Would just love to see OO.o become more modular or someone to rip out the import functions from the suite and build a separate program around it so I can get rid of the thing.

    Anyway, that is just my opinion…

  • Fred (comment #39) thanks, but no cigar for you.

    The user has write permission on her own “My Documents” directory and to the network share where she can save documents. She saves files often with no problem using all the applications she uses. Other times, only in OO.o, does this one problem appear. It is not a permissions problem.

  • anon (comment #40) okay.

    As an IT consultant you go convince a small, charity office to make a build environment to build its own software from source code. Or do this with an average small business. Go ahead, try it. Then come give us some verifiable evidence that you were able to convince them this is a good idea. Or are you going to build that for them and then you own all the problems when they crop up? Sure you will.

    While I am willing and able to build from source for myself I do not recommend it for typical desktop production environments.

  • The problem here is not that nobody is willing to fix it. The problem is also not that there are no resources (money, people) to do it, so throwing money to the problem is not going to help you getting it fixed.

    The issue can not be reproduced by someone that has the skills to fix it. If you want this issue to be solved, you have to understand what the problem is, and nobody to this point knows what causes the problem. There are some theories and possibilities, but nobody sat down when it happened and eg. straced the process to see what exactly is causing the error.

    If it is a permission or filesystem full condition, then an strace will clearly indicate what the problem is. Also if the problem is caused by stale lockfiles, or missing documents, an strace might help you with that as well. However if the problem is caused by a subtle memory corruption bug, it may be much harder to find clues if you are not an OpenOffice developer. In that case the only way to help solving the problem is by trying to reproduce it on other systems.

    At least if you can produce the error, or have the error on a system, you may want to contact a developer or offer someone access to the system so he can test the issue while it happens and extract the necessary information when it happens. Don’t expect anyone to do that work for you, they cannot reproduce it…

    Salvation is within reach, but it needs you to take ownership ;-)

  • J S

    Any software will have problems. There are so many lines of code and ‘those dang computer things do just what you tell them to do, over and over again!’

    Buggy features were what made me leave proprietary software for Linux and FOSS.

    I could never leave a Windows computer running for more than a day or two without rebooting or face risks of BSODs. Meanwhile my Linux installations run for weeks and months. I just checked my FreeNAS file storage machine and it has uptime of 58 days. That was when I finished installing and configuring it.

    I’ve run my business on Linux and Open Office for five years now – engineering & technical documents, spreadsheets, and presentations with/for/to Global 50 corporations. It always amazes me how much anger, fear, and FUD there is with making a change to FOSS. Get over it and save money!

    Remember: that money can go to hiring more great people to increase sales and expand your business and profits. Create a virtuous cycle.

  • The sad fact is that there isn’t a single office suite out there that can even remotely compete with MS Office. OO.o is, in truth, one of the worst. I use Koffice myself, but MS Office is the one thing i really miss from my Microsoft days.

  • As a fellow advocate, here’s how I would approach this: I’d ask the user to drop copies of every file she attempts to save with OO.o in one of two folders – a “saved” and a “didn’t save” folder – for one day. Once she’s accumulated a dozen-or-so files in each, go through them and try to find some commonality to the “didn’t save” files that isn’t present in any of the “did save” ones. Maybe they contain certain special characters. Maybe they use an obscure excel function or macro. Maybe they all contain a certain formatting artifact.

    This would be the first step to establishing a repro procedure.

    Relatedly: I honestly don’t believe that bug trackers exist only for the convenience of programmers – so those who say “fix it yourself and leave me alone” don’t seem to see that the very purpose of bug tracking systems is to allow non-programmer users access to programmers. I’d argue that OO.o and most other FOSS titles would be far worse-off without this form of input from end-users and VARs like yourself.

  • Alex Chejlyk

    My wife ran into this problem last night on her Windows 7 computer running Office 2007. She ended up using her Mac to open and edit the document. I run Linux on all my machines and have not encountered the problem with Openoffice.org.

    I wonder if the problem may be related to the virtual machine. As far as a developer is concerned, running software in a virtual machine may not be well tested or supported.

    I have successfully switched companies from MS Office to SoftMaker Office (not free, but less expensive).

  • km4hr

    Software, like everything else, is worth what you pay for it. OO is no exception. I’ve been using FOSS software for at least 10 years because I’m cheap and it does what I want it to. But I haven’t seen it make any progress in the business world for at least the past 5 years. And with all the new versions of Linux coming out every month I don’t expect that to change. One thing about MS software is that it is consistent everywhere. It might be consistently bad but the most important point is that it is consistent, and that’s critically important. This is an advantage that FOSS will never be able to achieve. As the old saying goes, “when everyone is in charge, no one is in charge”. For this reason alone FOSS is destined to remain the toy of software geeks

  • tekjunky


    “But I haven?t seen it make any progress in the business world for at least the past 5 years”

    No? Have you been looking at Redhat’s financials?


    and I quote here:

    2010 2009 2008 2007
    Revenue 748.24M 652.57M 523.02M 400.62M

    Their revenue has almost doubled since 2006 (4 years not 5 so less time than you mention) to the tune “$748.24 Million Dollars”!

    That is a whole lot of progress in my book.

  • ian

    You’re kidding.
    I’ve had colleagues of the not-very-IT-savvy sort bring in MS works doc files that Word refuses to open; and it wasn’t as easy as I would have liked to find a solution. Or try opening a Word 2007 doc in Word 2003. With the ribbon, while it came in about when I stopped using Office, the one consistent thing about it was that the task button I was looking for is invariably hidden.
    And, just backing up to your first statement, Linux didn’t cost me a lot to download; in terms of neck pain prevention alone it’s worth 10x what I paid for Windows licences. You get what you pay for is a uniquely American approach to life; like all rules, it has limitations.

  • imacdvguy

    Some of the discussion here is almost bizarre.

    People want to strap up a “community”, but then artificially enthrone “programmer” above all other community stations. If we choose to do this, then we should just acknowledge an hierarchy with “user” in the place of serf, and “programmer” in the place of Lord. This would, with instant force, destroy any delusions we may have in the F/OSS community about our valuing of community. That would be a shame.

    Much of the talk on this page lacks some sanity. A user does not have to edit or modify code in order to be a valued member of the F/OSS community. The GNOME Human Interface Guidelines are designed for mere mortals because that is the target of the GNOME developers. Understanding that the GNOME developers are not the sole bearers of this normal user burden we may know that the normal computer user is, in honest fact, a target. As a target, they are a desired part of the GNU/Linux ecosystem. As a desired part of the GNU/Linux ecosystem, they are necessarily desirable. Hence their status as valued members of a community that can never be an “all boy’s club” while being useful.

    As a consequence of this desire and specific target, those of us in the developer camp must understand that many users will be less than perfectly useful in the diagnosis of software problems. It is an inevitability. We have to work around it rather than complaining. We can’t ask them, “What are your contributions?” with any ounce of conviction. It is spiteful. It also forgets a vital tenet: If you view users properly, all users are contributors. Your software is only useful if it is used. If you desire no users aside from the strictly educational, say so. But I very much suspect that many developers do not wish to have their software roundly ignored except by the special interest group.

    The motivation for all of us to help one another in the F/OSS community is exactly that, a proper sense of “community”. “Community”, properly observed, is all the incentive you need. It is what prompts many of us to write code, create documentation, spend valuable time reproducing bugs in software that the authors thought worked just fine, or just simply using and embracing F/OSS software. Which, by the way, is already a favor to any programmer who wishes to have their software used.

    Every time one of my fellow students used an application of mine during university I felt elated. Exposure through ordinary users is a positive attribute of any project.

    I’m sure I have more to say, but I’m already wearing a hole in my soap box.

  • mario

    Well, as a non-programmer you are not able to fix the bug yourself. But at the very least you could have investigated it. I mean it’s two years, this was important to you.
    Even if it looks like garbage, users can still download the source code (or browse it online – SVN), and search for the specific error message.

    See, you have dedicated a lenghty article about the whats and abouts of the problem; but the ongoing discussion here is fully abstract. This will not solve your problem at hand. Posting or linking the error message and the surrounding source code might however have fixed the problem by now. If you are too lazy to be specifc, it seems not important enough for anyone else to go to great lengths here, and lazyness prevails with nobody checking out the code.

    Have a couple of open source projects myself. And non-programming users who find problems usually DO say about where they suspect the error to lie.

  • james

    I had the same exact problem on a Xubuntu 8.10 machine. fixed by reinstalling the OS

  • Is it just possible that getting open source software to run well on open source operating systems is more of a priority, than getting it to run well on a (now orphaned), closed source operating system built by a notoriously open-source hostile corporate entity?

    You are unnecessarily adding several layers of complexity to the task, and doing your client and the open source community both a grave disservice.

    Tell the client to run Open Office directly on a well-supported Linux distribution, not on XP running in a virtual machine. You/they may still encounter bugs, but at least there’s a snowball’s chance in hell of getting someone to take you seriously.

  • Macfield


    This article is serious but in some ways very funny. And here is why:
    I work for a large telecommunication company. This company makes heavy use of Excel 2003 for much of its day to day work.

    There is a long standing fault in Excel 2003 ( and probably earlier versions) whereby Excel will NOT and CANNOT reopen a file that it has
    modified and saved. The error report given by Excel is always something to the effect of “To many borders exceeded “, meaning that the user has made to many grid drawings using Borders and Boxes in the spread sheet, and as such has busted the upper limits.

    Why does Excel not stop you from saving the changes in the first place is beyond my understanding.

    Anyhow, here is the funny bit. We have to use OpenOffice Calc to recover (rescue) the broken Excel spreadsheet, by opening the affected Excel file using Calc, and then re-save the file in XLS format. Once done, then Excel 2003 can read the file again.

    Now don,t ask me what the heck is going on under the hood, but for sure this is a long standing ( several years) joke in our business, that OpenOffice has to come to the rescue of Ms Excel.

    And by the way this problem is well known and understood at Microsoft, but it was not in their interest to fix it in the 2003 version.

    In their defence, the problem does not exhibit itself in the Excel 2007 version.

    Though not a reason for in-action, but this type of issue cuts both ways.

  • dan dunn

    MS Outlook has a bug in all flavors. Once the main file size reaches 1.8+gb the program tanks and usually keeps sending an email to the last person you sent an email to, and keeps doing it every ten minutes. I would consider this to be the same type of bug.

  • The words \I don’t have time to become a programmer\ kind of stung.. but you did find time to write this post and to respond to all the comments in such painstaking detail?

    No offence, thanks

  • Have you considered KOffice? It runs on MSWin and GNU/Linux both, and OSX as well.

    Not the same as OOo, and therefor not trying to be a clone of MSOffice, but it is good. Sometimes, being different means the transition is easier because its not expected to be identical.


  • mario (comment #53) thanks for reading.

    I have been aware of this bug for much less time than it has been around. I think the first time I understood what the end-user was seeing was sometime last month. That was the first time she left a document up that was showing the problem and I went on-site to see it. That was when I was able to copy the data from a Calc sheet that would not save and paste it into another after closing Calc (with no saving) and restarting it. Before this, she did not know to do that.

    My research into the problem following this on-site visit revealed it had been an intermittent problem for different people for at least four years. Had I known about it sooner, before someone maybe had fixed it (we won’t know until the fix is in a production release everyone will use) I would have tried to help solve it by offering to run debug code. I rely on OO.o myself to create documents and I have never experienced this bug myself. I like OO.o so I recommend it to people. A four year old bug like this makes recommending it difficult.

  • All,

    I could not care less if closed source software has similar bugs or worse bugs. I could not care less what problems Microsoft Office has in any version. The fact is I rarely use closed source software at this point. In my experience over the past several years I have seen that open source can be better than closed source. That is why I highly recommend FOSS to people that look to me for advice on software for their businesses.

    So, when I discover that a hard to find bug that keeps people from saving a document has hung around in open source OpenOffice.org for four years it is a very unpleasant surprise. If “more eyes on the code” really means anything then it should also mean faster bug fixing than closed source could ever hope for. In many cases, it does mean that. In this case, apparently not.

  • Joel (comment #58) thanks for reading.

    Why would my lack of coding skill and lack of time to become skilled sting you? I am good at what I do as a general troubleshooter and consultant for SMBs. I did enough programming early in my career to know I do not have the patience nor desire for it. I know becoming a good programmer takes at least desire, skill and time. None of which I have at this point. I leave the coding to those who do have desire, skill and time. I write in my spare time because I do have a desire to write some, but not as a career. Obviously if I had more spare time there would be more than one to four articles a month here. ;)

    Another thing, too many are assuming I never do anything to help solve bugs. That is just not true. Please see my reply in comment #13.

    No offense taken. :)

  • lefty.crupps (comment #59) thanks.

    Actually, no. I had not considered KOffice. I do not use KOffice myself so it is not the first thing I think of when I recommend FOSS. Thank you for the suggestion.

  • lpbbear

    From your description, and correct me if I am wrong, the environment is a WindowsXP Professional system with a user running OpenOffice.

    You did not indicate whether the user only has “user” permissions or is a member of “administrators”. I am going to assume they are logging in as a user with their own username and password.

    What I think you should first establish is whether the issue is user centric or system wide. To test this create a test user with a new username and password and see if you can reproduce the issue.

    If the issue continues with the test user try logging in as administrator on the system and see if the issue continues.

    If the test user works properly the issue is related to some problem in the normal users profile. Perhaps migrate them to a new username and profile and remove the old one to see if the issue resolves.

    If the test user has the same issue but administrator does not then begin looking at permissions in the OpenOffice program directory. I use the Linux version of OpenOffice but I suppose OpenOffice dumps all of its main files somewhere in C:\Program Files like most Windows applications do.

    I find similar issues in Windows with many applications that have installed with incorrect permissions. Sadly too many Windows based programmers tend to be lazy and assume everyone using their software is going to have total administrative rights which leads to issues with “user” level users. Many times I have had to go set permissions wide open on specific folders for applications that cost the users an arm and a leg because some programmer somewhere was clueless about designing with security and permission levels in mind.

    Anyway, maybe those won’t help at all but give them a try.

  • lpbbear (comment #64) thanks for your comment.

    As with almost all small business offices still running XP Pro, this charity’s users are all set up as “administrator” on their own Microsoft PC systems. Unfortunately, this has to be done so their management software, designed for similar charitable organizations, will run correctly. This piece of software will also keep them having to use a Microsoft OS, at least in a VM if not as the primary OS, on most of the desktop PCs in the future. The current release of it fails to even load in WINE. The software is specific to this type of organization and it does their payroll. There is no FOSS alternative for it. I know because I’ve been looking for one for over two years now.

    Edit: I should add that there is a project that offers an alternative for this type of charity. But it does not have payroll, which is a requirement in this case. So, in reality there is no alternative for this specific closed source application.

  • lpbbear

    You should still give those suggested tests a try. It could still turn out to be a problem in the users profile. It may be a waste of your time but dot the I’s and cross the T’s ;)

    I also work in a similar fashion as you. I think of myself as filling the role of an “Implementer” in the FOSS food chain. When talk of Open Source and Linux comes up people tend to always focus on the Coders, Distributions, and End Users but the “Implementer” is almost never mentioned. For Linux and Open Source to succeed the needs of the Implementer have to be addressed. Implementers should not have to “fix it ourselves” and we are generally not coders who could do that nor do we have the luxury of time to do that even if we were coders. When we “implement” a solution for a customer it has to WORK right the first time and EVERY time and in the most efficient time possible.

    Last week I spent approx. 10 hours finding a Linux distribution that worked for a customers system that I could only realistically bill 2 hours for. Because I am a devoted Linux Open Source believer I kept at it until I solved it. Not many in my role would. They would simply grab the Windows disk and be done with it.

    THIS is the issue facing Linux and Open Source distributions now. We the “Implementers” are the ones on the front lines face to face with those potential converts. Its US they are talking to and its US they will blame for these kinds of failures, not the Coders or the Distributions. From reading your post I think this is the crux of the issue you were trying to explain. Its US the “Implementers” that take the heat for the mistakes on up the food chain. WE take the blame and its OUR necks on the line in a real financial way when trying to implement Open Source and Linux solutions. We have enough on our plate just trying to battle against the current in convincing customers to switch away from Windows. Its not our job to also “fix” the software nor should the fact that it may be “Open Source” be used as a cop out for others sloppy coding and or poorly tested distributions that fail to recognize plain jane hardware as was the situation last week for me. Those that would claim you should “fix it” yourself should be walking their talk because if they believe that you should do that, then they also should be doing that since they now are as aware of the problem as you.

    Anyway good luck with the OpenOffice issue. Hope you solve it in less than 10 hours. ;)

  • Joel

    I am just wondering where you are installing it. From Windows XP SP3 onward, it appears some applications installed inside the “Program Files” directory can have this problem if it is trying to write temporary files into that structure. The security has been “enhanced” on that directory and subdirectories. Have you tried installing it ins its own folder OUTSIDE of the “Program Files” structure? Just a guess, but maybe…..

  • For the what have you contributed crowd, and the learn to fix it yourself crowd, I have this comment…

    I’ve learned to program in various languages from Pascal to COBOL to Visual Basic to Java. I have not (and probably would not normally take the time to) learned how to program in C or C++. I’ve tried it, but the syntax and structure make no sense to me. Yes, I realize that Java uses a lot of the same syntax and structure, but it’s still different enough to make it easier to learn Java than C.

    So, the idea that the author should learn to code, just so they can fix this problem or so they can contribute something meaningful (so you’ll look at their problems) is ludicrous. Imagine this. If they learn to code, maybe they’ll just say screw your program and remake it. Then people will use their program instead of yours.

    It boils down to two things. 1) If you’re going to make a program that you’ll only work with other coders on, don’t expect your program to ever get above a niche audience. 2) If you’re working on a program with a wide audience (like O.O.o and Linux), then you need to drop the “If you don’t code, you aren’t worth a ***.” attitude. Otherwise, your project will rapidly become a niche program for a niche audience.

    For the author, a few questions come to mind.
    Does this happen if she opens one file and works on it in Calc, or is it after she’s worked on a few different files?

    Does it matter what order she does the files in? (In other words, if it happens on the fifth file that she works on, does it matter what that file is–or just that it’s the fifth file she’s worked on).

    Is it always the same file?

    When she gets the error, does she immediately close Calc, or does she try to save it to another location?

    Does it only happen in Calc, or does this affect Writer or other applications?

    Filing a bug report with those answers may help to narrow down the bug. Because it’s the little things like “It always happens on the tenth file.” or “It always happens on the file dogfood.xls” or even that “it only happens in Calc, but it doesn’t happen in other O.O.o applications ever.” will narrow the problem down. If the applications are using a common save routine or “class” for the coders out there, then it could be something in how Calc calls that routine.

    Have a great day:)

  • Spenny

    End users are told by the open source advocates that they are fools if they use proprietary software, and Open Source is the way to go. It is hardly surprising that there is such a divergence from expectation to reality.

    What we have here is a typical failure of the software geeks (and I speak as a C++ programmer of a couple of decades who only wilts slightly at STL template code) to understand that real human beings should not be expected to have software skills, including diagnostic skills, much as we don’t require programmers to have a sense of hygiene or interpersonal skills :)

    We don’t expect car drivers to be able to fix ECUs, and we don’t even expect them to diagnose the problem – the car manufacturers build the sensors in to work it out for themselves. A car doesn’t work – you pay ?60 a year to someone to take it away an take you home (and that ?60 service is often not called upon). Cars are complex bits of software, I’ve only got about 5 switches to control the lot, and I am oblivious to the 20 or 30 CPUs that lurk under the bonnet of the modern car.

    Computers are supposed to be a godsend – things where you require no brane to use them, point, click, job done. How come that typing a letter can be such an adventure compared with driving a car?

    The reality is, commercial or otherwise, that the current generation of PC software stinks – that is whether it is Windows or it is Linux or Mac or even iPhone. Users are expected to tolerate iffy interfaces, unreliable software and what is worse, when it doesn’t work, they get the blame for not being capable or helpful.

    If open source is a just hobby for you, well great – but don’t blame the users for failing to cope with crappy code and not having the skills to assist fixing it. What is so unreasonable about being able to type a letter and save it without needing to spend hours of time wondering why it might not work? Once you write code designed for end users rather than yourselves, then, whatever it says on the tin, you have created an obligation to provide something that works without the user jumping through hoops, otherwise it is not office software; you have not produced code that is fit for purpose.

    If you are not prepared to accept that Free means Free to end users, then don’t “sell” it as free – and if you think that the average office worker under pressure to do too much in too little time has more obligation to the programmer than their bosses, then you really are living in a different world.

  • Shane Kerns

    Hey Gene,
    I wouldn’t go tracking down every comment on this beautifully written blog and tell them you are not a programmer. The open source community needs non-programmers just as much as programmer. I am a programmer myself and judging by some of the atrocious comments written by some of the so called FOSS developers I can surely say that they are dumb asses whether or not they contribute. You are right in stating that such kinds of idiots are the ones that taint the FOSS community. Basically what they are saying is that if you are not a programmer then you cannot and should not use software at all, be it open source or not. These so call FOSS developers are idiots with a capital “I” which dates back to their over bloated egos.
    You provide a good service to FOSS, keep at it and good luck.

  • yonnie

    This whole topic sure brings up some raw emotions. It also brings up an old issue – paid programmers.

    Why not start a “hot” issue squad of programmers who can be paid by donations to resolve problem issues? This probably wouldn’t work too well for OO.O because it’s so complex, but it could go a long way toward addressing long-standing bugs and the fact that many volunteer programmers cannot volunteer all of their time without starving to death.

  • Andrey

    Open Office works for me, never had any problems with it. However, it is Java based and Java has a long list of known glitches, my favorite being the inability to run “java -version” in a folder that does not have write permission on Windows.

    Now, how can you expect, more so, pass this to clients, that there is or can ever be some free software for them? Free software emerges as a cooperative effort. So, if there is a hard to reproduce bug and you cannot make a debug build and find out what does not work using a debugger, just to show that to developers, you are out of luck.

    Your best hope is to get something free with strings attached. I guess you must inform your users that there are free alternatives to proprietary software but the debugger is part of the learning curve.

    As I see it, there can be no peaceful coexistence. The “users” either change or we ALL pay monopoly tax for EVERYTHING. Right now, while you are whining, Microsoft is SUCCESSFULLY introducing the Android tax.

  • Lazo

    Why dont u try the novell version of openoffice.. http://go-oo.org/

  • Linuxiac38

    I very much appreciate all the work done by all of you: Users who support Linux and FOSS; IBM (invests $1 Billion per year!); other corporations who invest; the many who offer it all up for FREE on their expensive server space on the WWW;
    all the fine folks on all the help forums; and the millions of folks who talk and promote the spread of FOSS, GNU/Linux, and the BSDs!

    Special thanks go out to Microsoft for several fantastic contributions:
    1. Microsoft runs 15,000 leased Akamai Linux Servers;
    2. Microsoft pushes customers to FOSS, GNU/Linux, and the *BSDs, through lock-in, update failures, the Million Microsoft Virus programs, and, by being the ONLY OS on Planet Earth that is NOT a parallel processing, concurrent multi-processing, multi-user, stable, fast Operating System!
    3. Continuing to run Hotmail/MSN.com on 13,500 GNU/Linux and BSD servers.
    4. Running all the corporate internal networks on 400 Linux routers.
    5. Continuing efforts (in advertising, in media placement stories, at court trials) by
    Microsoft Corporation to show by comparison that Microsoft’s corporate psyche is
    inferior to the passion of 500 million Linux, BSD, and FOSS, programmers,
    and users, in 184 nations!

    Your discourse also did remind me, and enlighten me, about my duty and responsibility to continue to work to maintain all that FREE stuff, that came with no guarantee, no warranty, yet has performed so flawlessly for me since 1996!

    In 1996, tired of the many disgusting issues and traits of a proprietary world of locked-in corporate greed, that caused me so much labor to maintain the computers of many hundreds of friends, family, schools and businesses, we ALL switched to Linux or BSD and FOSS! I convert more than 400 folks each year, to FOSS, in schools, businesses, and homes.

    I couldn’t do it without all of YOU! We all thank you profusely, for cutting off our shackles, and freeing us from drudgery and extended hours of slavery to the MS monster!

    BTW, yes, had a file in OO on Linux Mint 8, that would NOT save, though others since that one, on that system, save just fine!

  • George

    If you are looking for something free and just as good as OpenOffice, you should rather have a look at SSuite Office for a free office suite.

    Their software also doesn’t need to run on Java or .NET, like MS Office and so many open source office suites, so it makes their software very small, efficient, and very easy to use. :)


  • George (comment #76) thanks for the suggestion.

    That software is for Microsoft Windows only. The requirement is that the software must be cross-platform capable. The client in this case is considering a switch to Linux on their desktops for all office tasks. The exception will be their existing charity management software with payroll which is only available for Microsoft. That will be run in a virtual machine until the time it either works on Linux or can be replaced with FOSS/CSS software that works on Linux.

  • All,

    The management at this client instructed us to put Microsoft Office back on the user’s PC. Unfortunately they do not have the financial resources to pay us to troubleshoot bugs in OpenOffice.org software. Neither do we have the ability to troubleshoot this for free since client support and troubleshooting is how our business makes its income. This is a tragic end to what I had hoped would be a positive experiment for this client.

    The good news is they are not going to just give up on FOSS based on this. But this is delaying their decision and putting off any other experiments until sometime in the nebulous future..

Leave a Reply




You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Follow the directions below to post a comment if you are human. After 3 failed tries reload the page to start with new images.