“Microsoft will make available the preliminary versions of technical documentation for the protocols built into Microsoft Office 2007, SharePoint Server 2007 and Exchange Server 2007. This documentation, which defines how these high-volume Microsoft products communicate with some of its other products, is 14000 pages and is in addition to the 30000 pages posted when the software giant first introduced its new Interoperability Principles last month. They will be made available April 8.”
Will this help OOo implement more features into its code from all this documentation, or is more to show the structure of the programs with the layout and design, to show how they run and operate according to the protocols?
These are network protocols, not protocols for program operation and layout.
I think he (the person whom you replied to) is talking about, for example, exchange where by compatibility can be bought about in Thunderbird, and adding protocol support in OpenOffice.org to allow direct support for Sharepoint collaboration from within the various parts which make up the OpenOffice.org suite of programmes.
With that being said, 30,000 pages tells me a number of things; there is alot of backwards compatibility issues, there are alot of diagrams and they have to be incredibly verbose because of the complex nature of the protocol (more complex than it probably needs to be).
With this being said, I think the problem with this (the reliant on Microsoft) will cause pain at a later date because there is no assurance that they’ll continue to update the documentation when they update their protocol, change or fix a security issue within the protocol? I think the more important way to counter Microsoft is to create our own ecosystem of open standards that address the issues which Microsoft products – so that future improvement isn’t hamstrung to the whims of Microsoft.
The one advantage the community has is that Microsoft’s dependence on backwards compatibility with every previous version of Windows ever created is that protocol changes will still have to assume older versions. So even if they change it, they can’t scorch the OSS community without either scorching their own customers, or forcing said customers to apply arbitrary patches that MS knows from experience they may be hesitant to (ie. in the enterprise space).
Agree wholeheartedly, but we have our own ecosystem of open standards already, and it’s not working. It’s time for the community to do a little “embrace, extend, extinguish” of it’s own. When 98% of the computing world is using product A, it’s almost futile to try and convince everyone that switching to product B is better. Instead, product B should act like product A, but introduce improvements and benefits that product A doesn’t. So an app like korganizer or evolution can adopt compatibility with MS server protocols, yet ensure that users can migrate those services to open-standards based ones without forklifting the front end.
Or something like that. Anyways, it worked for Microsoft and made them what we are, so instead of constantly taking the high road, maybe the community needs to start using Microsoft’s strategies against them. It’s, frankly, the last thing they would ever expect.
Easily done. Create specifications where by if any part of the specification is improved on, the specification has to be revealed. Lets say Microsoft implements an openstandards protocol and then extends it, they’re then expected to reveal the specifications of that extension.
The benefit of that is that it would still allow proprietary vendors to develop support in proprietary products, whilst ensuring that if they do extend the specification they’re only obliged to disclose the nature of the extension – not the code itself. Thus, they keep their secret sauce, and compatibility is still maintained.
Correct me if I’m wrong, but don’t we have something similar (on Linux at least) in the form of D-Bus? Is this Microsoft’s version of it and can D-Bus be implemented in Windows, or is this something else entirely?
DBUS has nothing to do with this; DBUS is merely an ipc to allow communication between applications – which replaces what exists in GNOME and KDE.
There needs to be implemented protocols to allow communication between the OpenOffice.org and Microsoft products, and Microsoft only accept their protocols when it comes to communicating with Microsoft products (for obvious reasons).
So I’ll repeat, Dbus has NOTHING to do with it.
I’m not sure about MS intention, I simple don’t trust them. I can not make money with it, this is against the open source definition.
…from http://www.microsoft.com/interop/principles/osspatentpledge.mspx“&… .
“If You engage in the commercial distribution or importation of software derived from an open source project or if You make or use such software outside the scope of creating such software code, You do not benefit from this promise for such distribution or for these other activities.”
This is probably the main problem with their open specification promise… They will make everything available (sort of) and then say “if anyone implements our protocols outside any pet project, will have to pay us”.
This is monopolism, this is bad will from their part…
That seems excessively large to me. There spec is probably crap and really just released to ease fears.
I am pretty sure there are diffinitions, and examples and all sorts of stuff.
Fixed it for you…
Seriously, unless you’ve seen this stuff before, how can you be sure at all?
We^aEURTMve all seen this before, haven^aEURTMt we?
I figure this is like putting a treasure chest on top of mount everest. It is hard to imagine the amount of worker-hours it will take to put all that documentation to use…
Nothing benefits Microsoft more than open source projects writing to their protocols — especially with the patent deal that is tied into this. Because companies like Red Hat rely on the open source community, when the community builds against Microsoft protocols, Red Hat can’t redistribute the software without paying the patent price. That is why it is important to notice the patent license for commercial redistribution and patent protection for open source developers. Another attempt at taking down Red Hat from Microsoft.
That fine print hasn’t been released yet, according to the article. You should probably wait until they release even a hint of what the patent restrictions will be before criticizing them.
Fine print may have been the wrong terminology, but the fact of the matter still remains that some of these are patented protocols which will require patent licenses to use the protocols. Don’t forget that Microsoft is trying to court the OSS community here. It’s a move against Red Hat and IBM and others who benefit from the community work that haven’t signed patent agreements with Microsoft.
From the article:
In June, Microsoft will also publish a list of the protocols that are covered by patents, and will make available a patent map containing a list of the specific Microsoft patents and patent applications that cover each protocol, when the final version of the protocols are available, Robertson said.
The company will also release the final patent pricing and licensing terms at that time. ^aEURoeAs we work over the coming months on feedback on the protocols themselves, we are also going to be completing the patent map for each of these protocols,^aEUR Robertson said.
Once these much trumpeted multi-deca-kilo page specification dumps actually result in quality 3rd party implementations, I will take them seriously. And not a moment before.
In the mean time, I think we should take up a collection to buy the folks at Microsoft a tanker truck of Pepto-Bismol, because they seem to have developed a nasty case of PDF diarrhea. They should have known better than to try to eat the continent of Europe.
Edited 2008-04-08 21:11 UTC
So basically what you’re saying is that you’ll judge the quality of the MS spec based on the technical abilities of the developers who use them in their own products? Yeah, that’s fair.
How disingenuous of you to intentionally misreprsent what I said. History demonstrates clearly that there are plenty of developers in this world capable of taking a spec, written and presented in good faith, and creating something good and useful from it. That fact is not in question.
It is the “written and presented in good faith” part which I question.
If some of the talented developers, whom we know to exist, are able to run with this and create something of value, I will gladly give credit where credit is due. And that includes giving credit to MS if events prove their current campaign to be genuine. Until then, I am withholding judgement, but remaining skeptical.
Why do you have a problem with that?
Edited 2008-04-09 00:00 UTC
He didnt “misrepresent” what you said at all. Producing a decent spec does not guarantee someone will write an equal or better implementation. You’re basing the quality of the spec on an implementation instead of letting it stand on its own. Hell, before even examining it yourself.
I have a problem with you basically assuming that the problem automatically lies with Microsoft. I am not going to get in some sort of OSS vs Closed source pissing match, however all of the MS Office products have 20+ years of coding behind them…why would the spec be anything BUT extremely complicated and verbose? If it took MS 20 years to get these products to where they are today, it stands to reason that duplicating them would be a very difficult, expensive process.
I also hate to burst your bubble, but history actually shows that for every one good developer such as the ones you are alluding to, there are a hundred worthless ones just waiting to screw up their interpretation of a spec. I’d venture as far as to say that maybe 1% of the developers in the world actually have the technical breadth to wade through the MS specs and understand them, much less actually make a working product from them.
You cannot judge a specification by the product alegedly produced from it – thats simply ignornace.
Edited 2008-04-09 19:21 UTC
Isn’t that up to the 3rd parties?
This mean that in a near a future we will could have a better Evolution?
I’ve personally been interested in using the new SOAP interfaces in Exchange 2007 for development of a Linux based client that would suit my needs as a Linux user in an Microsoft office. I currently use Evolution, but I run into problems with it and eventually use OWA or Outlook in a VM.
The SOAP interfaces could be subject to change, but I’d imagine it would be faster to develop for than implementing the protocol again. I’d also imagine it would be a dream world compared to the current WebDAV interface.
Does this mean Mozilla devs can finally give Thunderbird access to an Exchange Server through MAPI?
Correct me if I’m wrong (and I probably am) – but isn’t MAPI just the API used to talk to MAPI-compliant mail clients like Outlook? In other words – it is used for talking to the client from a 3rd party product, not the server itself. It wouldn’t give you direct access to Exchange Server at all except via the Outlook client already installed on the machine.
BTW, MAPI is a public API already with even multiple other APIs that are implemented using it (CDO, etc.) – thus nothing prevents Thunderbird from using this other than it’s uselessness in that respect.
Edit: Unless you meant IMAP – which Thunderbird already can talk to Exchange Server through
Edit2: Upon quick lookup – you probably meant MAPI/RPC (AKA Exchange RPC) – which is a different beast.
Edited 2008-04-08 22:43 UTC
before i say anything, can someone point me at the supposed huge stack of documentation? i could not find any. if i want empty promises i will buy an amiga
It looks like if you start here:
http://www.microsoft.com/interop/
And dig for a while, you can eventually land here:
http://msdn2.microsoft.com/en-us/library/cc203350.aspx
Which appears to be the source of said documentation, although I didn’t investigate much further at that point.
…..while sitting on the can.
the whole thing is a waste of time. The deal with the EU went thru only because they got the EU to basically honor software patents just for Microsoft protocols. All the open source projects that fought hard to get the open specs got stabbed in the back at the last minute.
This is probably the same dreck the EU denied was acceptable 3 times around. The stuff then wasn’t SPECIFICATION, it was code examples, data traces, developer notes… all proprietary “IP” and not an instruction manual for how the product communicated that was asked for.
Like others said the whole thing is poisoned by their claim of random patents over unspecified parts of this documentation… it’s all mixed together and they aren’t bothering to say what’s got patents and what’s just spec. To any open source project this is worse than useless. The ones that need these specs, like EXT2win, Abiword or Gnumeric, or Samba already have better documentation than Microsoft in most cases (because Microsoft calls the commented code “documentation” when it’s really not) What these projects need to wrap up will only be in undocumented code samples… exactly what an open source project CAN’T use.
At this point we should be working to bring OSS tools to Windows (and Mac) users. Open file systems, data formats, and programs in droves. Rather than trying to hack our way into their world. Stop trying to connect to windows stuff and force the non-free side to keep up. Put our data in ODF, OGG, and use directory services that are openly developed and work 100% with open source projects. It’s time to move on without them. I understand Windows in particular is completely locked down in some areas that make adding new types of drivers needed for file systems nearly impossible, but that’s their problem. Much like the IE7/IE8 debates it’s time to simply take the W3C and WHATAG spec and walk away… tell people to get Opera/webkit/firefox if they want to be on the net and stop living 7 years ago!
Let Windows and Mac deal with opening files and connecting to shares. I keep including Mac because Apple is very willing to take from open source, but unwilling to share the fame or spotlight by using proprietary file systems and formats all over on top of free software. (I’m sure it’s because they can pay, so shun the little guy) Apple has little ODF support and no vorbis, theora, no ext2/3 etc.
But 44000 pages is *NOT* documentation for some protocols. It is a waste of time, where you probably will spend days just trying to find a single line that explains what you want to know to get you program working.
I have had enough problems finding info in programming books I have that take up an entire shelf. Now this is an entire wall of books.
Is there a good search engine for this?
MS have released these “Office 2007” protocols.
That will get a group of OSS coders frantically coding away, using these protocols.
In 2009, MS will change the protocols.
OSS coders start coding for Office 2009.
In 2011, MS change the protocols again.
OSS coders start coding for Office 2011.
Treadmill time….
The only way to get off the treadmill is for so many businesses and individuals to use (say) OO.org or Gnumeric/Abiword that MS (and their ever-changing formats and protocols) become irrelevant.
Unfortunately, that’s unlikely to happen for a long, long time.
Edited 2008-04-09 02:30 UTC
14,000 pages??? If a picture is worth a thousand words, it seems to me that about 14 UML diagrams would suffice. Or, wait, maybe they put more than 1 word on a page… Anyway, you get the point. Verbosity at its worst.
Edited 2008-04-09 03:28 UTC
These “knowingly” patented protocols cannot be used in FOSS software and shipped by commercial vendors without royalty payments to Microsoft. It would be painting a clear picture that FOSS doesn’t mind violating patents that are clearly visible.
I think most FOSS developers realize this (since some of the developers are employees of these companies) and therefore choose not to embrace them.
Now, perhaps these specs are useful for commercial Linux and Windows developers.
In all my years as a programmer I have learned the difference between a good programmer and a bad programmer, and also the difference between good software and bad software is organization and simplicity. If it takes 14,000 pages to describe protocols, they’re software must be the crappiest, disorganized and complex software in the history of the world
I’m no MS fanboy, but I was reminded of the complexities that the MS formats/protocols have to overcome as they evolve thanks to an interesting article over at Joel on Software (http://www.joelonsoftware.com/items/2008/02/19.html).
I dare say they are big documents because they are big subjects.
got a bitch slap from European Commission it seems..
use OpenOffice/AbiWord/Koffice (or Latex/Lyx/Scribus …..)
I’ve been writing networking protocols for a while now and the one thing you can always count on is that the documentation for them is needlessly verbose and complex.
Most likely, you end up dealing with multiple layers of documentation to boot. First you have the plain text RFC which is written in such boring repetitive detail. Then most of the time, someone takes the time to take that RFC and write a word document on it with proper diagrams to clarify the RFC for human consumption.
Top it all off with examples and giant tables. From what I read this covers several computer programs, each probably having many protocols. Top that off with MS support for backwards compatibility which only increases protocol complexity…it could reach 14 000 pages.
How in hades can describing an application’s network communications ‘protocol’ require 140000 pages?!?
I mean, seriously. There are only so many things that can occur that requires network communication in Office 07.
I can see 1000 pages pretty easy, to cover several components of the office suite ( Outlook, Calendar, etc… ). Then again From what I have seen of how Microsoft creates it software, it is likely because stupid little rules to prevent interoperability are in play.
Things such as every X communication switch endian-ness, invert ASCII meanings, or perhaps they have simply decided to make the docs convoluted and Office 07 wasn’t geared at preventing interoperability at all ( Yeah, RIGHT, though possible given what has been going on, I’d like to see them do this for Office 97 ).
Oh well, I guess I’ll grab it and see what is going on in there.
–The loon
(BTW, I’d bet that the 14,000 pages is more lines of text than there are lines of code to make it all happen, making the code itself what should have really been released… unless all that networking code had been wide-spread through Office 07 or they were worried about security flaws.)
EDITED to correct saying 97 to 07.
Edited 2008-04-09 16:33 UTC