LWN: Comments on "German Foreign Office drops Linux (The H)"

German Foreign Office drops Linux (The H)

gvy — Fri, 04 Mar 2011 12:47:41 +0000

> Personally, I'm deeply worried about any democracy
Well, ripping this part from context, I second that. Probably redefining "demos" as "multinational corporations" would get things closer to reality.

Microsoft Office interoperability however is an uphill battle

jjs — Tue, 01 Mar 2011 05:38:07 +0000

Ive continually seen it. One of the major frustrations is the page formatting changes when I change the printer I send it to. So your complaint on Openoffice.org has applied even within MSO versions. Not only does the layouts change, but then you have the gratuitous UI changes. The Ribbon Bar with no way to go back is probably my killer.

Openoffice.org was, in my opinion, "MS Office with all the bugs, bloat, and bad help" but has come a long way. I look forward to libreoffice, which promises even more changes to speed it up and make it more usable. I will note thtt OO.o had "print to pdf" long before MSO.

Microsoft Office interoperability however is an uphill battle

jthill — Mon, 28 Feb 2011 17:38:16 +0000

fwiw, I've seen it regularly, Word files from different computers losing pagination. The usual is people wanting a few extras of their resume, they often get an orphan.

Microsoft Office interoperability however is an uphill battle

nye — Mon, 28 Feb 2011 11:40:37 +0000

>It's equally a problem with MS -> MS - changing from MSO2003 to MSO2007 broke things, the earlier change to MSO2003 broke things, etc. Even worse, the formatting changes based on printer. Bottom line - this is a red herring, because MS doesn't maintain compatibility with itself in MSO.

People keep saying this, but I can count on the fingers of one head the number of times I've ever actually seen that happen.

On the other hand we recently tried using OpenOffice for half a dozen of our machines, and the response was that it's frustratingly slow and slightly breaks the formatting of anything which has a table or a letterhead or some layout that depends on particular margin spacing - which describes about 80-80% of our documents. Result: everyone perceives OpenOffice as a slow and crappy knockoff of an ancient version of MS Office.

German Foreign Office drops Linux (The H)

blujay — Sun, 27 Feb 2011 07:54:40 +0000

What kind of company do you work for?

People are the most important of the platform

Wol — Fri, 25 Feb 2011 19:38:21 +0000

It's not static even then :-)

Even the purest, coldest, hardest vacuum is a seething quantum soup.

Cheers,
Wol

People are the most important of the platform

cmccabe — Fri, 25 Feb 2011 04:09:33 +0000

I don't think anyone thinks that "every user of a program should be involved in its development." Only the people who enjoy being involved and have something to contribute.

If you don't enjoy something or don't have time to do it, try paying someone else to do it for you. You can do this with a recent (as in, last few thousand years) invention called money. Of course, sometimes it's easier to do business with a corporation than an individual. Guess what-- you can do that too!

Only a small percentage of the users of open source software contribute to its development. But that small percentage is enough to move things forward. In the same way, it's enough for only a few people to click on advertisements, or reply to spam emails, to make those things a viable business.

Sometimes its not individual users and developers who move things forward, of course; sometimes companies sponsor open source projects because it makes business sense to them. So you see Google sponsoring Android, Apple hiring WebKit and LLVM developers, and so forth.

So basically, I say to you: "the person who says it cannot be done should not interrupt the person doing it."

Did you ever read EULAs

anselm — Thu, 24 Feb 2011 20:45:28 +0000

So if you by MS products with no support they'll gatuitously indemnify you for all problems? I doubt it - and if you read the EULA you supposedly agree to, they disclaim ALL responsibility.

Not quite. I seem to remember that they assume liability up to the price of the CD the software comes on ($5 or so).

Did you ever read EULAs

jjs — Thu, 24 Feb 2011 16:54:45 +0000

> On the other hand, when IT chooses a soup of FOSS solutions, they become responsible for keeping it up and running - unless they signed a contract with someone like Red Hat whom they can shift responsibility to. But it's not always possible.

So if you by MS products with no support they'll gatuitously indemnify you for all problems? I doubt it - and if you read the EULA you supposedly agree to, they disclaim ALL responsibility.

At least with F/LOSS you have options - to include paying Red Hat or others for support. Oh, and they sign a support agreement for your money - sometimes (if you pay enough) with SLAs (Service Level Agreements).

People are the most important of the platform

pboddie — Thu, 24 Feb 2011 15:37:08 +0000

*Your* position is that FOSS requires people to be programmers. I never claimed that; only that it was *one* option.
Backtrack to first post in thread and scenario, where you pick the best solution available for any given activity, and feel like you need more features in it.

I think others have already stated that you get the choice to either modify the software in-house or to pay someone else to do it. So, no: FOSS doesn't require everyone to be a programmer. That's like the ridiculous Bill Gates remark about why an open source office suite supposedly doesn't make sense, completely ignoring the ability Free Software users have to decide themselves how they get their software fixed or improved.

In response to the suggestion that people pay others to do the work, you wrote the following:

Software development does not work like that. It's not standardized that heavily, especially if talking about *free* software. Also, you are much more likely to hire a proprietary developer who will develop a feature in theirs closed-source app specifically for you, than to find a person able to do the same work in a FOSS project.

Here's the key point: with a proprietary product you can only go to the original developer and ask that they fix or improve something, and if their mindset is "selling licences to lots of people is how we make our money", then they may well tell you to buy the next version, which may not solve your problem at all.

Meanwhile, it's laughable that you can't find people to fix or improve Free Software projects. I'm sure that if you approached a number of projects with money they'd be quite happy to do work which may or may not end up in the upstream project. And with various projects, there are usually a number of contributors who comprise a fairly effective marketplace for the purchase of services.

People are the most important of the platform

Del- — Thu, 24 Feb 2011 12:27:48 +0000

You need to pay attention to what I am telling you.

>If you mean buying several versions of the product separately, I don't have any problem with that. Code has no fixed price, and by the way, is rarely sold, rather it is licensed.

That is the business model of proprietary software. For FOSS you typically pay for support and development. Exactly why do I need to tell you this? I am sharing with you real world experience. Either embrace it or come up with something way more substantial than babbling.

I can also share with you that the business model of proprietary software creates a lot of overhead expenses that FOSS projects typically avoid. Instead of talking to suits you typically talk to developers when you fund FOSS.

>Try asking KDE community to add new features to KDE 3.5 branch because you don't want to migrate to KDE 4. It's illusion that you can easily fork the code. Forked code is dead if you don't fork the project, i.e. developers, as well.

If I wanted added functionality to KDE3.5, I would contact the Trinity project, and I am pretty sure they would be interested if my suggestions were at least vaguely sensible and backed by funding. My experience with the KDE community has been related to Kolab+Kontact. They were very forthcoming, and easy to collaborate with. I would recommend them to anybody as a full replacement for outlook+exchange. As long as your suggestions are sound, you will also find funding new functionality quite smooth and very affordable there. Give it a try with Outlook+Exchange, and see what MS will charge for implementing SyncML.

>when IT chooses a soup of FOSS solutions, they become responsible for keeping it up and running - unless they signed a contract with someone like Red Hat whom they can shift responsibility to.

We have a huge contract with Red Hat ;-) Actually Red Hat supports the huge pile of software shipped with it, so it is way more of a soup on the windows clients. Professional support is available for FOSS solutions, but (again) the business model is different. I repeat, it is lack of competence on the subject that is the main obstacle today, and you are a prime example.

People are the most important of the platform

kevinm — Thu, 24 Feb 2011 04:36:15 +0000

...and of course the owner of the code has their own goals P, Q and R that likely diverge from those of their customers.

People are the most important of the platform

jjs — Thu, 24 Feb 2011 04:13:47 +0000

> Simple logic: exclusive ownership of code -> more money in a single place -> more abilities to recruit people to work towards a single goal

Simple, logical, and wrong. You're failing to understand - THERE IS NO SINGLE GOAL. My company has a set of goals. Your company has a set of goals. I want the software to do x, y, and z because it moves towards my goals. you want the software to do a, b, and c because it moves towards your goals. If you have a single company that can afford to do 3 items, which do they choose? They may choose x, a, and m (something from someone else) because it matches their goals.

However, I can freely hire someone to add x, y, and z - and then contribute upstream. You may not care. But another org out there wanted w,x, and y, so are happy to get x and y, and they pay their developer to do w.

People are the most important of the platform

jjs — Thu, 24 Feb 2011 03:49:24 +0000

> Support may be free if included in license price. Doesn't actually matter. You'll need to pay for the effort, in one way or another - and it's better to pay with money.

Same strawman used by Best Buy and others for Windows - "It's free because you already paid for it" - no it's NOT free - you paid for it.

BTW - most free software developers gladly accept money. Or you can do the effort yourself, or wait. you have options, which you DO NOT HAVE with proprietary.

Another point - assume a company/organization can fix 1000 bugs at a time. Now you come in with but 1010. It's critical to you. Guess what - you have to wait (or pay extra) to get your bug fixed. If BPC (Big Proprietary Company) doesn't want to fix it because others are, to THEM, more important, you're out of luck. With Open Source, you got to another developer and pay to get the bug fixed. YOU HAVE THE OPTION.

Most ocmpanies use proprietary not because they want proprietary but because "No one ever got fired for buying [IBM, Microsoft, etc]". It's not technical, it's legal/political. As F/LOSS extends, company bosses will see the excuse "we're waiting on [insert BPC here]" is a smokescreen - the correct answer is "we've hired [insert F/LOSS company here/name of internal programer] and expect fix by xxx date".

People are the most important of the platform

jjs — Thu, 24 Feb 2011 03:34:20 +0000

> That's because of proprietary developer's expertise with their own software

Zend, Asterix. Companies that use their expertise in their OPEN SOURCE products to produce business.

Also your comment illustrates the great thing about Open Source - you can hire whoever you want to work on it because you have the source code.

Finally, most modifications needed to tailor to your business are ALREADY BEING DONE - 80% of programers DO NOT WORK FOR SOFTWARE COMPANIES. They work for everyday companies integrating apps into those companies business and writing custom code for their specific needs. With Open Source, they continue doing that, with better tools and options.

Microsoft Office interoperability however is an uphill battle

jjs — Thu, 24 Feb 2011 03:28:54 +0000

It's equally a problem with MS -> MS - changing from MSO2003 to MSO2007 broke things, the earlier change to MSO2003 broke things, etc. Even worse, the formatting changes based on printer. Bottom line - this is a red herring, because MS doesn't maintain compatibility with itself in MSO.

Which, when you think about it, given their monopoly, is a great way to drive upgrades by forcing people receiving new documents to upgrade.

People are the most important of the platform

RCL — Thu, 24 Feb 2011 01:24:07 +0000

> *Your* position is that FOSS requires people to be programmers. I never claimed that; only that it was *one* option.

Backtrack to first post in thread and scenario, where you pick the best solution available for any given activity, and feel like you need more features in it.

>> Yet this kind of power concentration is needed in order to afford major projects
> Not necessarily. See also the co-op. If enough people are interested enough to fund a proprietary software company, they're interested enough to fund a Free software company.

Simple logic: exclusive ownership of code -> more money in a single place -> more abilities to recruit people to work towards a single goal.

Keeping people organized by non-monetary means seems to be less successful (compare number of developers in similar projects, e.g. GIMP vs Photoshop).

> Certainly, proprietary software and its clear monopolistic tendencies tend to concentrate lots of wealth in the hands of one entity compared with the general populace. This sounds awfully anti-democratic.

What has democracy to do with business? People may have equal rights but their abilities aren't equal.

> You get free support? Seriously? I always have to pay.

Support may be free if included in license price. Doesn't actually matter. You'll need to pay for the effort, in one way or another - and it's better to pay with money.

>> FOSS tends to provide flaky-shakey DIY experience that doesn't look reliable to decisionmakers.
> What are you even talking about? Please cite source and logic to back up your assertion here.

The very article we are commenting on.

> Meh. Methinks I'm being trolled. *plonk*
> I think we're going to have to agree to disagree. This is better for both of us. :)

I had the same impression. If *you* could only pick the most important pieces of my posts and not try to attack every other sentence, the thread wouldn't look like conversation of two trolls.

German Foreign Office drops Linux (The H)

rqosa — Thu, 24 Feb 2011 00:44:43 +0000

> a lost cause like Ogg Vorbis

Why do you say that WebM is a "lost cause"? (Remember, Vorbis is the audio codec part of WebM.)

People are the most important of the platform

RCL — Thu, 24 Feb 2011 00:44:22 +0000

I think that the nature of free software makes it harder to profit on it. Which is not to say that there are no wealthy companies living off of FOSS. But what matters for lobbying is relative wealth.

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 23:25:11 +0000

(I did finally boot my gNewSense FSF bootable membership card. The intel wireless requires proprietary firmware, so it's not in. I'd have to switch it to atheros, iirc. The good news is that YouTube supports WebM, which is in Firefox 4, so that would handle streaming, if others used it. (Again with the external market forces).

At least this conversation got me to finally boot it up. It's kinda neat to have a membership card that actually *does* something! :)

Can't wait to rescue someone's PC with it. heh.

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 23:17:40 +0000

Meh. Methinks I'm being trolled. *plonk*

I think we're going to have to agree to disagree. This is better for both of us. :)

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 23:16:38 +0000

> 1) You basically propose that every user of the program should be involved in its development. This does not scale (violates basic principle of human specialization).

Ah, I see. You're attacking a straw man, I'm afraid. I've tried to make it quite clear that not all users need be developers, just like not all washer or fridge users are repairpersons.

Rather, we pay others to do this for us because it's not something we've chosen to specialize in.

The problem is that this option does not currently exist due to Free and Open Source Software having insufficient marketshare to make software-repair a viable model.

Of course, like with appliance repair and upgrades, if you have the knowledge, you can do it yourself, but you don't *have* to.

> Most users cannot be developers, yet they expect the program to fulfill their needs, and complain when it doesn't.

Complaining is all they *can* do if you use proprietary software. If you use FOSS, you have other options, as discussed above.

> There are dishonest people who will violate FOSS licenses.

Yes, of course. There are people who violate proprietary licenses too (glaring and the not-so-glaring, like benchmarking early .Net implementations). Heck, I'd wager you can find a number of very cheap copies of Windows 7 on the streets of China.

The purpose of copyright is to shut this down (until the idea is old enough that it's no longer worth it to prevent others from copying the
author's ideas.)

>>> much better than artificial FOSS-only environment,
>> Again lacking the logical argument behind your assertion.

> Try using gNewSense.

What of it?

> There was old saying about USSR, the country I was born in: socialism successfully resolves problems that aren't known anywhere else. That perfectly describes experience of gNewSense.

> There are a lot of applications that companies use in their day-to-day life, which aren't business-critical yet still valuable. Being proprietary, they only require a license and occasional call to support if things go bad (or migration, if developer goes out of business, but that's rare).

You get free support? Seriously? I always have to pay. Or hope they deign to respond. Rather the same as Free Software.

"Only require a license"? Seriously? Try Matlab Student edition (not just a license; you have to have the documentation CD mounted in the drive. Or Maple: you have to have eth0 and it has to have the right MAC address.

Proprietary software developers going out of business or being bought out and terminated or "assimilated" into the parent company's portfolio is rare? Really? In what world?

> Being FOSS, according to you, they would require employment of contributors or at least programmers - even if your main business isn't software development.

No, that's a gross misstatement of my position.

*Your* position is that FOSS requires people to be programmers. I never claimed that; only that it was *one* option.

> For the same reason why we cannot have "anarchy" as a government.

First, FOSS is compared to communism (complete dominance by a government) and now anarchy (complete absence of a government). Which is it?

> FOSS, by definition, does not make it easy to accumulate power around a single entity.

You mean, it tends to limit monopolies? Yeah. I'd call that a feature, not a bug, and it's *very* far away from proving that FOSS is anarchy. Rather, it proves that FOSS is anti-monopoly, which I'd agree with. And perhaps even anti-oligopoly.

Contrast this with proprietary software, which is by its very definition a monopoly. Which is the free market here?

> Yet this kind of power concentration is needed in order to afford major projects

Not necessarily. See also the co-op. If enough people are interested enough to fund a proprietary software company, they're interested enough to fund a Free software company.

> or influence politicians.

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 22:53:59 +0000

> And in the latter case you'll have to manage the project yourself.

So that would be, "Yes, you have that option, but it's hard to do by yourself." Which is true. But it happens if there's enough support for it. Take for example GCC/egcs, XFree86/Xorg, and OpenOffice/LibreOffice. None of these would have been possible with proprietary software.

And this was accomplished *despite* not having anywhere *near* a majority presence.

> Responsibility becomes manageable if every piece of software is supported independently (by its provider).

You get support for your proprietary software for free? Where?

Everything I've seen is either:
1) Pay for support
2) You may get a little support for minor things if you're sufficiently nice and it's sufficiently small
3) Your problems may get fixed in the next release. Or they may not.

Which is what you get in FOSS too, except, if you can convince enough people to join you, or you have sufficient resources to tackle the problem at hand, you can get it done without the vendor's direct help (and you can perhaps get it sent upstream; many times the hackers would love to do something but don't have the resources to do everything that would be nice to do; your oppositional proposition is only one, perhaps deliberately negative scenario.)

That you can get support for software that the company themselves didn't write (e.g. things included in the distro) is enormously different from the hodgepodge of providers you have in Windows. This is due to FOSS licensing.

> unless they signed a contract with someone like Red Hat whom they can shift responsibility to.

Yes, this is the same as with proprietary software. If you want support, you have to pay for it. If you want things fixed, you have to pay someone for it or hope that it gets fixed when you pony up for the next release.

> But it's not always possible.

It's not always possible for proprietary programs either. Whom do I pay for support for the myriad of small, proprietary programs?

People are the most important of the platform

RCL — Wed, 23 Feb 2011 22:43:19 +0000

> Either way you end up paying for the same code several times

If you mean buying several versions of the product separately, I don't have any problem with that. Code has no fixed price, and by the way, is rarely sold, rather it is licensed.

> eventually adding new features become [...] very expensive (try asking what Adobe will charge you for bringing your feature request to Photoshop).

Try asking KDE community to add new features to KDE 3.5 branch because you don't want to migrate to KDE 4. It's illusion that you can easily fork the code. Forked code is dead if you don't fork the project, i.e. developers, as well.

Neither Photoshop nor KDE cannot be influenced that easily, in both cases you need big money: either to sign an exclusive contract with Adobe or to pay a horde of developers.

And in the latter case you'll have to manage the project yourself.

> More importantly we can continue accessing and influencing the project long term. Implementation of the software in your business is also far easier avoiding licensing issues.

Again, only license costs are really avoided. Influencing the project which decided to move in an unwanted direction is costly and forking it shifts all the responsibility to you.

> Going with the big proprietary offerings is conceived as the safe route, even to the point of total stupidity. It-departments are actually the worst in this respect. It seems like people in IT-departments fear that they might loose their job if the company went over to free software.

I think it can be explained easier. Responsibility becomes manageable if every piece of software is supported independently (by its provider). On the other hand, when IT chooses a soup of FOSS solutions, they become responsible for keeping it up and running - unless they signed a contract with someone like Red Hat whom they can shift responsibility to. But it's not always possible.

People are the most important of the platform

RCL — Wed, 23 Feb 2011 22:24:21 +0000

Again, a summary answer:

>> Some of them don't scale well or don't take into account selfishness of humans, or both.
> How so? Please state examples.

1) You basically propose that every user of the program should be involved in its development. This does not scale (violates basic principle of human specialization). Most users cannot be developers, yet they expect the program to fulfill their needs, and complain when it doesn't.

2) There are dishonest people who will violate FOSS licenses. Learn about this case: http://bit.ly/gl4kaW

>> much better than artificial FOSS-only environment,
> Again lacking the logical argument behind your assertion.

Try using gNewSense. There was old saying about USSR, the country I was born in: socialism successfully resolves problems that aren't known anywhere else. That perfectly describes experience of gNewSense.

>> Your own programmers, who are presumably not that knowledgeable with FOSS piece you want them to improve
> If it's business-critical, why aren't they already small contributors to the project already? (Also explain why, once they implement the feature, they will necessarily immediately forget everything they've learned and start over from scratch the next time you need something.)

There are a lot of applications that companies use in their day-to-day life, which aren't business-critical yet still valuable. Being proprietary, they only require a license and occasional call to support if things go bad (or migration, if developer goes out of business, but that's rare). Being FOSS, according to you, they would require employment of contributors or at least programmers - even if your main business isn't software development.

> But as you pointed out, the world is hardly "FOSS-only;" rather, there's a highly entrenched monopoly squelching competition. Therefore, "overall usage statistics" (i.e. that Linux usage is in the minority), being an indicator of *current* market conditions, has nothing to say about my "artificial FOSS-only environment."
> For completeness, I'm saying that once some threshold population of FOSS users is reached, it becomes self-sustainable, with end-user/customer-friendly features (the hire-a-developer model) that can't be found in the predominantly proprietary world we currently inhabit. Perhaps we can and perhaps we can't reach past threshold. That's a separate argument.

We can't. For the same reason why we cannot have "anarchy" as a government. FOSS, by definition, does not make it easy to accumulate power around a single entity. Yet this kind of power concentration is needed in order to afford major projects, or influence politicians. FOSS tends to provide flaky-shakey DIY experience that doesn't look reliable to decisionmakers.

People are the most important of the platform

nix — Wed, 23 Feb 2011 21:59:07 +0000

Why do you assume that companies backing free software are necessarily dirt poor? IBM doesn't strike me as not able to accumulate a significant amount of money...

German Foreign Office drops Linux (The H)

cantsin — Wed, 23 Feb 2011 21:52:33 +0000

Free software does succeed on the desktop when it only needs to comply to open standards (such as Firefox and Chrome/Chromium) or when the FLOSS model actually provides interoperability benefits (VLC playing almost any media file as opposed to proprietary media players).

Microsoft Office interoperability however is an uphill battle. The people discussing here likely do not work in environments where you have to share complex Microsoft documents using Open/LibreOffice. We're not only talking about daily formatting screw-ups - which often enough are so bad that in most cases, you cannot edit a co-workers document without destroying letterheads, tables, form fields etc. -, but also the non-existent support for Microsoft macros in the free alternatives. These things cannot be fully compatible in free software since Microsoft's implementation is non-standardized and over-complex.

ODF once brought the promise of ending this mess, but it has not gained corporate momentum, and likely is a lost cause like Ogg Vorbis. Even standardizing on it inside an organization doesn' t help since using it across different (Open Source) applications such as OpenOffice, Abiword and Kword results in the same incompatibilities and formatting screwup as with Microsoft documents.

It's realistic to see Linux on the corporate desktop as a lost cause except for a few niche scenarios.

People are the most important of the platform

Del- — Wed, 23 Feb 2011 20:54:40 +0000

>Software development does not work like that. It's not standardized that heavily, especially if talking about *free* software.

True, you may not be able to get stupid suggestions or poorly implemented patches accepted in an open project. This is typically a good thing though.

>Also, you are much more likely to hire a proprietary developer who will develop a feature in theirs closed-source app specifically for you, than to find a person able to do the same work in a FOSS project.

I have extensive experience being responsible for funding multiple proprietary and FOSS development projects. What you are saying simply does not match reality. For the proprietary projects they either die or get bought by large players. Either way you end up paying for the same code several times and eventually adding new features become impossible (if they die) or very expensive (try asking what Adobe will charge you for bringing your feature request to Photoshop). As long as they are runners up you can get fair agility though, but even then you cannot make several vendors cooperate easily. For open projects it is relatively easy to unite several communities (as long as it makes sense), and my experience is that we typically get much more value from the investment. More importantly we can continue accessing and influencing the project long term. Implementation of the software in your business is also far easier avoiding licensing issues.

The reason not more companies take advantage of the open model is mostly lack of competence on the subject. Going with the big proprietary offerings is conceived as the safe route, even to the point of total stupidity. It-departments are actually the worst in this respect. It seems like people in IT-departments fear that they might loose their job if the company went over to free software. This may be caused by the perception that Foss saves money, an argument I hate. It is *not* about saving money, it is about using the money for better services, increasing value. Actually, it is very important to continue spending with FOSS alternatives, simply because the same money brings more value. It makes sense.

People are the most important of the platform

HenrikH — Wed, 23 Feb 2011 20:03:02 +0000

> Small businesses can't realistically make use of the "Free to modify" Freedom of FOSS, but large businesses and government units most certainly can

We are a 5 people business in the finance industry and we make fine use of the "Free to modify" Freedom of FOSS, we even hack the kernel on occasion. Also giving your clients the benefit of helping/modifying our software (GPLed) is awesome, due to this very fact we have now almost completely removed a former monopoly competitor with over 100 people due to our higher pace of development since the customers are involved.

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 19:14:05 +0000

>> you are deluded if you think that world is in some kind of temporary, imperfect state, and will magically "correct" itself once certain hindrances are removed.

> The world is *always* in a temporary, imperfect state. Ignoring the once-again only-inflammatory "magically" and scare quotations around "correct", I don't think that's what I'm claiming at all.

For completeness, I'm saying that once some threshold population of FOSS users is reached, it becomes self-sustainable, with end-user/customer-friendly features (the hire-a-developer model) that can't be found in the predominantly proprietary world we currently inhabit.

Perhaps we can and perhaps we can't reach past threshold. That's a separate argument.

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 19:06:33 +0000

> The world is *always* in a temporary, imperfect state.

Argument:
The world is not perfect. I think we can agree on this. I would also argue that it will never be perfect, as there's always some possible way for it to be imperfect, such that the final, perfect state can never truly be achieved. By some metric. (It can, however, be improved, impaired, or kept the same. Perhaps view it as a real number; you can get as close as you like to the number, but never get *to* the final, perfect state, by fractional improvements)
I suppose there's no reason you can't be in the exact perfect state, but I don't find it likely, as there are too many ways to screw it up, and being an epsilon from perfect won't cut it.

The world is in a temporary state. The world in the next time epsilon is different from the world as it is now. I suppose the heat-death of the universe will eventually get close to static, (but of course, it's still only static in the infinite limit, i.e. never)

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 18:54:20 +0000

> Some of them don't scale well or don't take into account selfishness of humans, or both.

How so? Please state examples.

> you are deluded if you think that world is in some kind of temporary, imperfect state, and will magically "correct" itself once certain hindrances are removed.

The world is *always* in a temporary, imperfect state. Ignoring the once-again only-inflammatory "magically" and scare quotations around "correct", I don't think that's what I'm claiming at all.

Perhaps the differences in perspective is one of you're mired in the now and I'm interested in what might be if.

> natural forces (monopoly being one of them) resulted in current state of software industry

Yes. I believe I've claimed this from the beginning.

> it is functioning well

Not really, I'd argue. Particularly at how nobody but Microsoft can win in the desktop space, which is rather the point the article makes.

Is a market where only Microsoft can win because everyone knows them and they have deep pockets to buy politicians best for the average user than a real, competitive marketplace unencumbered by proprietary lockin and where the user's only recourse for software defects is the vendor? I'd say that it's not.

> much better than artificial FOSS-only environment,

Again lacking the logical argument behind your assertion.

> evidenced in this article

The article rather clearly states that the transition back to Windows is not technical but political.

> evidenced in [...] overall usage statistics.

But as you pointed out, the world is hardly "FOSS-only;" rather, there's a highly entrenched monopoly squelching competition. Therefore, "overall usage statistics" (i.e. that Linux usage is in the minority), being an indicator of *current* market conditions, has nothing to say about my "artificial FOSS-only environment."

The competitiveness is also not purely based on technical merits, as you agree. Rather, entrenched monopolies are stomping on smaller competitors due to vendor lockin both mental and technical and due to their deep pockets.

People are the most important of the platform

RCL — Wed, 23 Feb 2011 18:31:31 +0000

I have no time to address all of your questions, so here's my summary:

You don't look far enough into your own predictions. Some of them don't scale well or don't take into account selfishness of humans, or both.

Also, you are deluded if you think that world is in some kind of temporary, imperfect state, and will magically "correct" itself once certain hindrances are removed.

In fact, natural forces (monopoly being one of them) resulted in current state of software industry - and it is functioning well, much better than artificial FOSS-only environment, as evidenced in this article and overall usage statistics.

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 17:19:01 +0000

> Note that I'm not talking about technical superiority. I'm talking precisely about "market" superiority, and FOSS is unlikely to obtain it, because it tends to be dispersed and is rarely able to accumulate significant amount of money for a single developer

This is quite possibly true, and tragically sad. I weep for the future that could be. Perhaps I should write a short science fiction story about it. :)

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 17:17:18 +0000

> You can have developer of some unimportant software implement and maintain a feature for you for a few thousand euros.

Aside from flavoring words, e.g. "some unimportant" which are misleading and solely useful for attempting to malign Free and Open Source Software, where do you get the price? Are you seriously suggesting that adding a checkbox in a dialog is the same as implementing and using a new networking protocol?

> Your own programmers, who are presumably not that knowledgeable with FOSS piece you want them to improve

If it's business-critical, why aren't they already small contributors to the project already? (Also explain why, once they implement the feature, they will necessarily immediately forget everything they've learned and start over from scratch the next time you need something.)

> Don't people see advantage of centralization of app expertise in a single place, that is, the developer?

Ah, I see. You're still entirely wed to the notion that software has to be proprietary, inside of one organization.

In FOSS, the underlying premise of your statement is that the application is developed by an organization, consisting of developers who solely work on that project to produce new versions of it. This is not (necessarily; like git there are a number of models you can use) the case in Free Software.

> Are you serioulsy suggesting ministry to maintain a broad range of free software

A small ministry? No, they'd hire someone to fix their bug or band together with other ministries in a co-op fashion to hire someone to fix the bug or implement the feature. Or, like in proprietary software, ask the upstream developers to do it or just stay silent and wait for it to hopefully happen.

> starting with obscure drivers for the most cost-effective Chinese hardware

In a Free Software-prevalent world, drivers for the hardware would be produced by the vendor and kept upstream in the kernel (and documented, in code or in spec). Just like it is with Windows now, or do you really expect a piece of hardware to be released without Windows or Mac support? except that when the vendor goes broke, the driver still gets updates for as long as a sufficient base are using the hardware (or as long as you can continue to pay for it being kept around).

> not ending with the latest "law library" or accounting helper applications.

Why not? It's the same as any other piece of code. You find someone who you can trust and who specializes in that piece of code (or your hired person does).

Of course, this only works if there's sufficient customer base. Which we don't have right now, which is why nobody thinks of it. Users having power to modify code or hire a local or trusted partner to do it is not well-known amongst the general populace.

People are the most important of the platform

RCL — Wed, 23 Feb 2011 17:14:02 +0000

>> If free software were easier to use or more profitable to develop, it would be used instead - and there are some areas where it's the case.

> Right, and if some technology were superior, we'd no longer be using coal-fired plants. This is false; markets are extremely complicated and interwoven, even if you ignore out-of-bend (e.g. governmental, both to and from) influences.

Note that I'm not talking about technical superiority. I'm talking precisely about "market" superiority, and FOSS is unlikely to obtain it, because it tends to be dispersed and is rarely able to accumulate significant amount of money for a single developer (which could then use it to lobby for it).

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 17:04:09 +0000

> It's not standardized that heavily, especially if talking about *free* software.

What's standard? There's a huge amount of vendor-specific bits... just like in software (where you use libraries, perhaps analogous to standard parts)

> Also, you are much more likely to hire a proprietary developer who will develop a feature in theirs closed-source app specifically for you, than to find a person able to do the same work in a FOSS project.

Are you talking about now, where the vast majority only use proprietary software? Yes, of course that's the case. I discussed this. It's due to a lack of sufficient customer base (and a culture of not having this option), not due to some innate lacking in FOSS.

> That's because of proprietary developer's expertise with their own software.

No, that's because of how the current culture and market are. There's no reason a software repair/enhancement shop shouldn't also be experts in their chosen software.

> Users use proprietary software because it's easier.

I would claim it's because it's all they know (culture) and because of current market realities. There's absolutely zero reason proprietary software *must* be easier.

> If free software were easier to use or more profitable to develop, it would be used instead - and there are some areas where it's the case.

Right, and if some technology were superior, we'd no longer be using coal-fired plants. This is false; markets are extremely complicated and interwoven, even if you ignore out-of-bend (e.g. governmental, both to and from) influences.

People are the most important of the platform

RCL — Wed, 23 Feb 2011 17:02:38 +0000

> With FOSS, you don't have to plead for special attention from the program's vendor to add features, your own programmers can do it.

You can have developer of some unimportant software implement and maintain a feature for you for a few thousand euros. Your own programmers, who are presumably not that knowledgeable with FOSS piece you want them to improve, will take more time and money, especially if employed on permanent basis.

Don't people see advantage of centralization of app expertise in a single place, that is, the developer? Are you serioulsy suggesting ministry to maintain a broad range of free software, starting with obscure drivers for the most cost-effective Chinese hardware and not ending with the latest "law library" or accounting helper applications.

People are the most important of the platform

RCL — Wed, 23 Feb 2011 16:52:33 +0000

>> Small businesses can't realistically make use of the "Free to modify" Freedom of FOSS, but large businesses and government units most certainly can.

> Why not? You hire people to do things for you all the time, such as dishwasher or car repair and enhancement.

Software development does not work like that. It's not standardized that heavily, especially if talking about *free* software.

Also, you are much more likely to hire a proprietary developer who will develop a feature in theirs closed-source app specifically for you, than to find a person able to do the same work in a FOSS project.

That's because of proprietary developer's expertise with their own software.

> The main problem is that the vast majority of users don't have access to this option because they use proprietary software.

Users use proprietary software because it's easier. Masses follow the path of least resistance, believe me. If free software were easier to use or more profitable to develop, it would be used instead - and there are some areas where it's the case.

People are the most important of the platform

Trelane — Wed, 23 Feb 2011 16:30:29 +0000

(or fix bugs, in the dryer repair model. There's of course no reason one can't soup up a dryer, but there's no culture for it, unlike cars. Heck, my kenmore came with a parts and electronics diagrams, which I've used quite a bit to fix things that were broken in it. Because it was a Kenmore, the repair shops wouldn't touch it (supposedly, Kenmores can only be repaired by licensed repairpeople or something))

In this model, like other repair places, they'd have specialists with different pieces of software. They'd receive service bulletins (security or bug fix updates) and send patches upstream.