Reply to post: Re: “Office applications have not been specifically designed for unattended usage at scale”

Re: “Office applications have not been specifically designed for unattended usage at scale”

Multi-user “shared” (concurrent/simultaneous) editing is only offered by Microsoft “as part of the product” via Microsoft Teams architecture.

Simple sharing of documents via OneDrive or SharePoint was never designed and offered by Microsoft as a *product* but rather as a limited feature with many asterisks (particularly related to the simultaneous experience). SharePoint does not imply simultaneous editing of the same objects.Many users can get a copy of the same object and edit it independently, but then they would need to integrate their edits back together, and they better have established beforehand who would be editing what (it’s called data governance rules, in this case they can’t be enforced, and have to be validated when the different, updated copies of the document are being merged/integrated back into one, which is not managed by SharePoint.

Nor shared (concurrent/simultaneous multi-user) editing of documents truly supported via OneOrive sharing architecture. Multiuser editing is possible, but not truly simultaneous. Even Google Docs ultimately can not provide such experience, since multiple people physically cannot be updating the same letter in the same document at the same time and observe how they cannot do it in a true real time for everybody. There are actually very few same things that people can do at the same time in an uncoordinated fashion, and in such cases usually no more than two actors can participate without “stepping on each other toes” — so, accordingly to that expression, one of such activities would be pair dancing, another sex, and as the number of participants grows, if the same specific goal to be maintained, the degree of planning, coordination, or real-time communication necessary grows too, be it a verbal communication, as in a group debate, or non-verbal or combination in that other case. And still, there are always some ground rules. Another example is playing music or performing as group — it’s highly orchestrated, choreographed and rehearsed, if the common goal outcome with contributions from many to be expected.

There is one special case of collaborative activity which is not pre-planned, communicated, or orchestrated in the real time, yet it extremely harmoniously happens in the absolute real time with a common goal output. The common goal is maintained, but the exact collaborative output is never predictable, and there are quite a few underlying rules, or even a framework to which all participants conform to produce the output albeit exactly unpredictable yet highly coherent to the common goal. It’s Jazz, and it’s obviously art, although it based on a use of lot of common technology, i.e., rhythm, musical harmony, other musical disciplines, and acute listening of everyone and everybody as a whole by every given participant, which could certainly could be considered as form of very deliberate communication; in many respects it’s not much different from a productive group debate, although it’s hard to imagine verbal human debate produce the same high-value output as Jazz. Not goal oriented and deliberate group human communication could be very harmonious and pleasant, but it cannot be as productive.

So these are the ultimate natural limits of what’s possible in a collaborative processes of different kinds. Basically, naturally it nearly never happens before more than two people, and it happens there, because nature programmed it.

You can extrapolate from these examples. Computer technology can help a lot with coordination and communication, especially substituting interpersonal communication, particularly the information exchange and speed of its visual representation, but the problem stepping of each other toes without coordination (rules) and synchronization/serialization (sequencing), and orchestration (all together-governance) remains the same. Below this topic is discussed specifically in application for collaborative electronic content creation using computer systems.

Specifically for Office document shared collaboration, the best architecture that Microsoft came up with so far is Microsoft Teams framework. Try simultaneous remote multi-user collaboration on the same documents In the Channels via Microsoft Teams (see details here, and pay attention to the difference between In-channel document shared addicting and just cloud document sharing — it’s vastly different modes: https://is.gd/JfgpY0).

But if you need a true, robust, multi-user document editing capability, there is a number of industrial publishing packages which offer this capability, it’s this functionality has been around, even on PCs and other computers before Windows arrived (Xerox was the first to offer this, and it was the first to develop graphical user interface, GUI, which Apple and Microsoft, in Windows, famously coopted), for more than three decades.

Although this software is usually focused on visual editing rather than semantics (text), text is obviously part of it. Still, in such software the “edit mode” used to be granted to a specific user on a per-page basis; for the cross-page formatting, one would need special permissions, including allowing content to spill over to next pages and push their content down. Or the edit permissions could be given within an editorial piece spanning several pages, but the individual editors, who only could edit the pice one at a time, do not control the layout of the piece across pages (this a normal process in magazine editing.) For such an industrial multi-user collaborative process to work, there must be a robust system of governance in place.

If you know anything about enterprise database management systems work, you would understand, that only one person (or process) can edit/update a record at one time. Any number of users can simultaneously edit their own copies of the same record, but under normal circumstances, whoever saves their version last, wins, due to something called “serialization” (so, obviously, saving has to be coordinated: if the record has been changed by someone after you made your copy for editing, you cannot overwrite a newly saved version before merging the record content into your copy and making sure that your edits are done upon the latest the version of the record. Regardless of the fact that no data is overwritten these days, just new copies of the changed data are saved, still it’s necessary to know which saved copy is the “latest” or “current” version of truth. So all sorts of data governance schemes defined how the editing of the “same” piece of data is done.

Word do documents are XML documents, which are basically a nested hierarchy of objects, as far as their content, as as far as their layout, that is dynamically defined on the screen, this type of document editing/production is called WYSIWYG, or “what you see is what you get”. It’s dynamic, but not predictable (you can’t refer in the beginning of the text to a specific page and guarantee that what you are referring to will be in that page, as the document continues to progress. At his is opposed to a publishing (a paper of magazine layout, where the article breaks on one page, where it states that it is continued on specific other page, and that is guaranteed. In Word do document for online publication, one can use dynamic pointers to the later placed in the document, regardless where they would end up in the document. The footnotes are also managed dynamically.

The point is though, that in the WYSIWYG editing environment, what one person sees of one document may not necessarily be the same what another person editing the same document would see, and if anybody were to print the document at any point in time, or save a personal copy, what they would get is also unpredictable, if a bunch of people are editing the same document, especially if they are editing potentially the very same text as another person or people happen to be editing. (The effect would be something like when you give control of your screen, keyboard, and mouse to another person, and then try to do something yourself — two or more people cannot edit the same piece of data, be it a word, a sentence, a paragraph — whatever the atomic unit of data is chosen by a given multiuser software, and the updates that everybody does cannot be replicated and seen on everyone-else’s screen in an absolutely real time, especially if one is changing what another one has just changed. There has to be some coordination. There are no miracles, even in quantum physics, one thing cannot be *and* be observed at two different places at the same time (if this works for an analogy).