Re: [Fsfe-uk] File format import/export

[ian]

On Mon, 2003-08-04 at 15:25, Simon Waters wrote:

> Specifically the question of when will MSO get an Import or Export for
> OO format documents.

Why would MS produce import and export filters for OO.o? Answer if it is
in their interests. When MS Word was a minority product and Wordperfect
ruled the roost, it was important to have WP import and this remains as
far as I can see in Office2000. Apart from various text formats when I
go to save as there is still only a WP filter for saving files - oh and
Works. MS might want to import OO.o docs but I'm not sure they want
people exporting docs in that format. Besides, the OO.o import and
export filters for Word are very good and improving so is their really a
need from the MS perspective? If they do it, it just provides free
publicity for OO.o. If governments and large corporates adopt OO.org
file format as an international standard and insist that all
applications use it, then MS would have to or lose sales but of course
there is a chicken and egg issue there.

> Ultimately more utility and less interoperability issues occur if OO is
> installed at the other end. I believe this is the route we should press
> for, as it will give people a better experience of free software than
> they have had of proprietary incompatibility*.
> 
> Technical barriers to this (bandwidth/disk space) are I think a red
> herring. 

Agreed. OO.o works reasonably on 64 meg and will run on Windows 95.
Office 2003 will not run on any of the MS Operating systems except
Windows 2000 and XP. That is going to be a big crunch. Will all those
Windows 95/98/ME users out their upgrade their hardware and buy a new
operating system and then fork out for Office 2003 when they can do
everything they need on the current machine and can have OO.o free
knowing it will work on all versions of Windows and Mac, Solaris and
Linux too? Let's hope even a significant minority go the OO.o route
because then even if they do upgrade their machine they are likely to
stick with it. Also there is some incentive for the OO.o developers to
make the code more efficient, there is really no incentive for MS to do
this because the more resource hungry the more people have to buy new
PCs with new copies of Windows on them.

> I may be stuck with ISDN, and old PCs, but for general purpose
> Office computers sold in the last 2 years OO fits fine, as it will for
> newer home PCs. Actually it runs kind of okay here on P200's with 64MB.
> Broadband is now widespread. OO beta version is on the cover disk of
> Linux User and Developer this month BTW, I assume just the Linux
> version, haven't looked at the CD yet. One off 40MB downloads are not so
> painful.

In any case you can buy it on disc for a fiver and then burn as many
copies as you like and pass it round all your friends. I just had an
E-mail from someone who is clearly not that technical but he waited 5
hours to download it after reading an article I wrote in Living with
Technology. So I don't think its as doom and gloom as all that.

> Microsoft typically provide import formats for competing applications,
> as this is a requirement to migrate to your product, so expect an Import
> function for OO from MS when they decide the potential market gains
> exceeds the increased ease of life for OO users of such a function.

Double edged sword though because it also makes it easier to lose as
well as gain people. See above.

> Whilst I agree that Office format documents are not the technically best
> way of doing consultations or surveys, or getting multiple document
> modifications to review and merge, it does utilise tools readily
> available to the civil servants in question, which may make it the most
> practical solution. So if they are going to do it this way, I'm with
> Ian, it ought to be free (gratis) software they use to allow others to
> consult without paying an IT tax to Microsoft, and I suspect it would
> alos have to be FREE software to support a sufficiently broad range of
> platforms.

You could say that the Government have an obligation to do this. After
all they use pdf because you can download a free reader. Ok it takes
longer to download OO.o but the principle is the same. I think in
Switzerland the Tax collectors are sending OO.o to people to fill in
their tax forms on the grounds that its cheap enough for the revenue to
send out a CD with the tax forms. 

> Not to say a good general purpose survey/form design/consultation
> website couldn't find a niche in eGovernment. But it has to be easier
> and cheaper, or better than just wacking out a word processed document,
> to both author and responder - remember the government doesn't
> experience the responders pain when replying to government consultations
> - - if anything more pain means less responses, and thus less work
> collating those responses, even if it results in worse government.
> 
> The openness of the document format may be important to software
> authors, from a practical perpective if I want to do some "red-lining"
> style co-operation with Ian, I probably want him running exactly the
> same version (or as close as possible) of whatever software is in use.
> 
> - From a practical perspective free (gratis) software has a big edge here,
> as it avoids the difficulties of everyone having to get budget to buy
> the same software. If FREE software can't win in this area it isn't
> going to win anywhere.

Yes and upgrades are an issue too. MS Office 97 has been file compatible
with 2000 and XP to this point. MSO 2003 changes that. What happens when
people with MSO 2003 start sending attachments to people running office
97 (Still millions of them out there) MS hope it will force them all to
run off and buy a new computer with a new OS and then upgrade to Office
2003. I'm not so sure. It could be a big opportunity to get OO.o a much
bigger user base and establish it as a very serious rival to MSO -
sufficiently so for Govs to adopt it as the natural choice. The MSO is
doomed in the longer term. 

> We need to distinguish document format descriptions carefully....
> 
> It may be "open" as in fully documented and people are not prevented
> from writing interoperable implementations , it was my understanding
> Postscript and PDF (at least old versions of PDF met this).
> 
> It may still be proprietary in control of the standards, although I
> think this is less of a problem as free implementations are not obliged
> to go to newer standards. However it may be that proprietary features
> are added (such as patent encumbered, or encryption!) which standards
> controlled by more broad committees may not tolerate.
> 
> However a lot of this comes down to the trust you can place in the body
> controlling the standard, and even that may change with time, like the
> proliferation of companies on committees defining Internet related
> standards.
> 
> Quite how much faith to put in companies like Adobe is a difficult one,
> although I think perhaps they now begin to recognise that "goodwill" in
> the company as a manager of the standards is as, or more important, long
> term than their short term business interests. However I think Java
> story is suggesting that the market may find that proprietary control
> unacceptable. Microsoft may have shot themselves in the foot by making
> such an issue over the management of the Java standards by SUN.
> 
> Am I alone in thinking that a free software implementation, may be a
> better "standard" in some areas than a written standard, I certainly
> think this was the basis of SMTP's success, because as a mail protocol I
> doubt it would ever have got through a telecomms industry protocol
> committee, but sendmail as a free interoperable implementation made SMTP.
> 
> Similarly Winboard protocol became a major player in the Chess world by
> happening to be how GNU Chess 4 and Xboard/Winboard talked, others could
> refer to the programs as a reference implementation. That is not to say
> I don't think makin the protocol or file format "explicit" is a bad
> idea, many of the flaws of these protocols comes from making them
> "compatible" with the implementations that existed when they were
> standardised.
> 
>  Simon, who was pondering this as a result of thinking that KDE 3.2 is
> beginning to really show the snowball effect of free software, with a
> fully integrated Desktop and related web/mail/PIMs features that begin
> to make proprietary solutions look rather poor in comparison. Although
> they could do with more big business users to get the hardware vendors
> behind products like Linux PDA syncing software.
> 
> * And ultimately it is about better software for doing a job, although
> better is a broad term.

Its partly about long term vision too because some software that migh
not be as good at doing a job to-day is worth backing if it is likely to
do the job much better and less expensively in the future.

-- 
ian <address@hidden>