Microsoft admits being 'on the wrong side of history' with regard to open source
theregister.co.uk
theregister.co.uk
Right now, today, they're shipping literal malware on mainstream desktop computers around the world. Oh, and it's not open source either, lol!
The bloody operating system has targeted advertisements, intentionally unremovable (or at least very difficult to remove) components, telemetry, a keylogger-esque search function, the list goes on.
It pains me to see people shill Microsoft as if they've "turned a corner". All they're doing is the minimum possible in order to be taken seriously by some particularly easily swayed subset of developers whilst simultaneously ratcheting up the pain on the main product that they're known globally for.
Source? None of the things you listed below are /literal malware/.
Adware is considered a form of malware. Not the worst type of malware by any means but still “literal”.
GNU hosts a list as well; though I believe many items on it to be a stretch, I feel at least a few are justifiable. https://www.gnu.org/proprietary/malware-microsoft.html
That's a lot of space devoted to just old versions of games. I understand malware has a specific definition, but I don't know what's in these folders and we obviously didn't want them on user machines.
Deleting them required a whole dance around SYSTEM access, it was insane.
Enterprise.
Strictly not malware though.
Now with every bit of windows software hooked into the network, I see it a little differently. (same with basically all phone software)
In the past few decades, the part of the movement that has become far more popular is the very corporate-friendly, weakly-licensed side. Especially in an era of cloud computing, you can take whatever volunteers from an MIT-licensed open source product make, build whatever you want to with it, and not release the modifications or give back to the project's developers in any way. From a corporate shareholder perspective, what's not to love?
I would guess they still don't like open-source; they are just doing whatever they can to support their cloud business.
So the future is the cloud, but not Azure, that's just a hosting facility which MS can't complete concede to Google and AWS from a paranoia perspective. But Docs hits them where it hurts, and if it gets any better then outlook->gmail and office->gsuite becomes a real threat.
I have 5 Office 2019 licenses at home because my family is used to it, they're using it and happy; however I work in a big, continent-wide project and everything is done (spreadsheets to reports to slide decks to coffee) on Google Docs & Google Drive. No one is using Microsoft Office or any other desktop office software and, it works.
Lines are not that clear cut anymore.
Are you happy with the responsiveness of Google Docs? For me once the spreadsheet exceeds a certain level of complexity, it feels really sluggish, and I feel like I'm constantly wasting microseconds while in the desktop counterpart I could have it done already. Sure, the sharing function is great, but as someone accustomed to working quickly, I'm sorry to say GD still doesn't offer a smooth experience on more complex sheets.
I also use some sheets designed as multi-page input forms and it works well though, we don't collaborate on them real time.
I'm using Firefox with a 16M/1M ADSL connection BTW.
https://www.reuters.com/article/us-alphabet-gsuite-idUSKBN1F...
My comments are just for my point of view and my experience. I'm in no position to talk for the whole landscape TBH.
So what? That simply means rights are passed down. The problem with the SSPL was never that; it's that the rights that _were_ passed down are _legally impossible_ to fulfill, rendering the whole point of copyleft moot.
> The GPL was viewed much the same way the SSPL is now.
The criticisms against the GPL-ed works were about collaboration, not use. The problem with the SSPL is that I cannot legally use SSPL-ed works, in any way.
In simpler, analogical, terms: the GPL was designed to be "viral", but the SSPL was designed to be a minefield. Viruses can be harmless; minefields deny combatants, civilians, and peaceful inhabitants alike, forever.
Any GPL-ed software could be used without any restrictions on _how_ you use it. Whereas the only legally safe way to use MongoDB is to pay MongoDB Inc. to host it. So much for copyleft.
----
http://lists.opensource.org/pipermail/license-review_lists.o...
https://www.processmechanics.com/2018/10/18/the-server-side-...
https://www.processmechanics.com/2018/10/25/further-comments...
This is clearly not the case. Given a) the text of the license itself, and b) the fact that people still use MongoDB.
I think what you meant to say was “I cannot provide MongoDB as a service without open sourcing the rest of what I use directly to provide that service”. In other words, if I want to sell the usage of software which I don’t have the copyright to(remember, open source software may be free, but it’s still owned by somebody), then I have to give my source away as well. It doesn’t say “Thou shalt not compete,” the way the commons clause does. You can still provide Mongo as a service. You can still do anything you want with the software, but you might have to open source the rest of your stack.
Honestly, it’s not like Mongo didn’t have a lawyer look at the license (ie write the thing). If it was as horrrible and perilous as you say, they couldn’t use it because all the lawyers of all the other companies who use Mongo would get up in arms.
The only “safety” you lose is that of owning the copyright to closed source software, which if you really care about that, then why do you care how “closed source” Mongo chose to be?
> b) the fact that people still use MongoDB.
Just because people are using something doesn’t mean they’re doing so legally.
I'd like to refer you again to the three links I cited in my parent post.
> ... the fact that people still use MongoDB
That is no basis for legality. If we're to use popularity as a measure, I'd refer to my classic argument about the popularity of competing religions that each denounce the other.
> I think what you meant to say was ...
No. As I've been saying all along in this thread, the license is legally-unenforceable. But you seem to be insistent on this other point, so I'll take the bait, for now, and hopefully I'll do a good enough job of illustrating my point this time.
> "I cannot provide MongoDB as a service without open sourcing the rest of what I use directly to provide that service"
1. "open sourcing" != releasing under the SSPL, which is what's demanded by the SSPL. The SSPL is not an open source license.
2. "... use directly to provide that service": this falls under contract law, not copyright law. Licenses fall under copyright law, not contract law. Licenses dictating terms beyond the scope of the copyrighted work, which is what the SSPL does, is copyright abuse and legally invalid. Copyleft licenses, including the GPL family, work strictly within copyright law; the terms they dictate apply only to the copyrighted work. Yes, derivatives of copyrighted work fall under the same copyright. No, "... use directly to provide that service" does not qualify as derivative work and thus does not fall under the copyright of the original work.
> ... didn’t have a lawyer look at the license (ie write the thing).
Lawyers across the world write legal briefs, notices, claims, affidavits, contracts etc. and lose all the time. Simply being written by a lawyer does not make something valid or correct.
> If it was as horrrible and perilous as you say, they couldn’t use it because all the lawyers of all the other companies who use Mongo would get up in arms.
Most users of MongoDB do not do carry out due diligence before using it anyway, or, historically, MongoDB wouldn't have had as many users in the first place. For a neutral example: companies and institutions across the world use pirated works all the time, and not because they actively want to tempt legal trouble.
> The only “safety” you lose is that of owning the copyright to closed source software ...
This is not how open source licenses work — open source licenses do not confer ownership of copyright by mere usage — so I don't know what your point is. Perhaps you were confused by how MongoDB has in the past demanded (and probably still does) copyright transfer before accepting contributions from outside, but, I assure you, that has nothing to do with open source licensing.
I think that there's some room here.
The links that you gave had two examples, the Lasercomb case and the Sandisk case. The SSPL isn't like either of those.
From your first link: > For example, in Lasercomb, the license agreement required the licensee to > agree not to develop a competitive computer-aided design program for 99 > years. Lasercomb America v. Reynolds. 911 F.2d 970, 15 USPQ2d 1846 (4th > Cir. 1990). In Practice Management Information Corp. v. American Medical > Ass'n 121 F.3d 516 (9th Cir. 1997) the AMA licensed “CPT” health care > codes (in which the AMA claimed a copyright) on a condition that the > licensee not use any other such competing codes. In Assessment > Technologies v. WIREdata. 350 F.3d 640 (7th Cir. 2003), the copyright > holder tried to limit the licensee’s access to the licensee’s own data > stored using the software. This defense is almost always applied to > prevent a copyright holder from leveraging its copyright to impose > anti-competitive practices.
In other words, what we're looking at here is, "Now that you have accepted our license terms, you can't build a competing software, with no out" What the SSPL says (paraphrasing) is "If you can't offer our product as SaSS on an SSPL stack, then you can't offer it (see 12. No Surrender of Others Freedom)." It doesn't say, "by using our software, you are not allowed to build a competing product."
Let's look at SanDisk:
>SanDisk had a patent licensing program that required any licensee to provide a grant-back license to any subsequently-developed patents in the same field of use. Two courts examined SanDisk's program under both the antitrust and patent misuse angles.
In other words, SanDisk said, "if you use our license, we can use anything else in this space that you license." MongoDB is not saying, "now that you have licensed Mongo, you MUST SSPL regardless of whether you are currently using Mongo." Again, point 12 comes into play here. You are allowed to simply /not/ use Mongo. You have an out. The requirement of having a stack under SSPL or compatible license is a precondition for being allowed to provide Mongo as a service, not a limitation of what you can do after the fact. In other words, if the company sued over SSPL can't or won't release their stack, then they just can't provide Mongo as a service. That is, I think an enforceable outcome within the bounds of copyright: "if you do X, you can't use our copyrighted material." Thus I think your point of the SSPL being unenforceable is incorrect. Of course, this is interpretation, IANAL, and we won't truly know the result of this until it hits a court.
>This is not how open source licenses work — open source licenses do not confer ownership of copyright by mere usage — so I don't know what your point is.
You've misunderstood me. I mean that providers lose the ability to use closed source code in the providing of the service. The statement you were quoting was said with the presumption that "you" (in the general sense) were the provider of the service, who also owned the copyright to said closed source code used to provide said service, and that the thing lost was the ability to keep that software closed-source.
>SSPL is not an open-source license
What, because some committee says it isn't? Very well, open-source in the colloquial sense, as in "I can read the source code, modify it, and run it if I wish."
> Now that you have accepted our license terms ...
> "now that you have licensed ...
That is not how licenses work. Licenses cannot be forced on either party, nor can they force terms after being granted; that'd be extortion. Neither Lasercomb nor SanDisk forced their licenses or subsequent terms on their licensees, nor am I alleging (as you seem to paraphrase) that Mongo does.
The misunderstanding you seem to have is that a license term such as "If you agree to this license, you MUST ..." can be interpreted as "Now that you have accepted this license, you MUST ...". It can't. All three of Lasercomb, SanDisk, and MongoDB's SSPL use the former and not the latter.
> You are allowed to simply /not/ use Mongo. You have an out.
As did the licensees of Lasercomb and SanDisk to simply not use the respective licensed works. That did not stop those two licensors from being in legal jeopardy, and from their licenses being invalid. And when a license is invalid, _any and all_ use of the copyrighted work under that license becomes copyright infringement.
> then they just can't provide Mongo as a service.
And here you introduce another kernel of the problem with the SSPL. It does not define what a "Service" is, leaving it vague and so open to interpretation, that MongoDB Inc. can change their mind later and sue even users not providing MongoDB-as-a-service. This inability (or omission) of the license to answer the question "What is a service" leaves me as a potential user in a very gray area. This has all been discussed many times over [0].
The only safe play is to not use the copyrighted work at all. Which leaves, as I say above, "the only legally safe way to use MongoDB is to pay MongoDB Inc. to host it."
> You've misunderstood me. I mean that providers lose the ability to use closed source code in the providing of the service. The statement you were quoting was said with the presumption that "you" (in the general sense) were the provider of the service, who also owned the copyright to said closed source code used to provide said service, and that the thing lost was the ability to keep that software closed-source.
You said, "The only “safety” you lose is that of owning the copyright to closed source software". One (owning the copyright) has nothing to do with the other (keeping the software closed-source), so I don't know why you would include the former if all you wanted to say was the latter.
> What, because some committee says it isn't?
No, because it does not meet the industry-accepted definition.
> open-source in the colloquial sense, as in "I can read the source code, modify it, and run it if I wish."
1. You can't just "run it if I wish" as you please; and
2. That is NOT the definition of open-source; that is the definition of shared-source, or source-available. If MongoDB wanted to make their source available under those terms, they should have, instead of trying to abuse the term 'open source'. If you want to use a piece of software under shared-source terms, feel free, but please don't try to jam it to others as if that's open source, because such twisting of terms harms the industry.
----
>If conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot use, propagate or convey a covered work so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not use, propagate or convey it at all. For example, if you agree to terms that obligate you to collect a royalty for further conveying from those to whom you convey the Program, the only way you could satisfy both those terms and this License would be to refrain entirely from conveying the Program.
The important bit here is:
>If you cannot use, propagate or convey a covered work so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not use, propagate or convey it at all.
That is what I mean to say when I say you have an out. As in, rather than SSPL'ing your whole stack, you can simply STOP using MongoDB, or any other SSPL'd software. Once you stop using MongoDB, then you are free not to SSPL the rest of your stack under clause 12. Under the SanDisk and Lasercomb cases, you couldn't simply un-agree to the license, because it was a software purchase, or a patent license respectively, but the SSPL is a license that is validated and invalidated and renewed automatically (see clause 8: Termination), and for which the acceptable, enforceable outcome of any litigation is always "stop providing the copyrighted work as a service".
>You can't just "run it if I wish" as you please
Name one thing I can't do with it. No, providing it as a service does not count, as I can do that, provided I manage the service with other SSPL'd code.
So what? The automatic termination of a license doesn't make it valid, nor does it stop it from being invalid. Even in the presence of the termination clause, the SSPL asserts rights over works it does not cover. _That_ is copyright abuse, exactly like in the Lasercomb and SanDisk cases. The fact that those licenses differed in other areas does not affect this fact.
But yeah, thank you for bringing up the termination clause, as it further strengthens my point that using an SSPL-licensed work is like entering a legal minefield. As I say in my parent comment, if tomorrow MongoDB Inc. decides to interpret "Service" differently than today, invoking the termination clause, suddenly I would be forced to stop using the software. This mere possibility makes it unsafe for me to use the software now.
>> 1. You can't just "run it if I wish" as you please; and
>> 2. That is NOT the definition of open-source; that is the definition of shared-source, or source-available.
> Name one thing I can't do with it. No, providing it as a service does not count, as I can do that, provided I manage the service with other SSPL'd code.
At this point, I'm beginning to suspect you're being pedantic and willfully ignoring practical realities. You're also being selective and ignoring the rest of my argument: merely being able to run your modifications under certain conditions does NOT make something 'Open Source'. And, please, read the rest of my argument, which I reproduce here for your convenience: "If MongoDB wanted to make their source available under those terms, they should have, instead of trying to abuse the term 'open source'. If you want to use a piece of software under shared-source terms, feel free, but please don't try to jam it to others as if that's open source, because such twisting of terms harms the industry"
It's not like they've open sourced their core business, they're open sourcing things to make Windows more attractive, so that people won't leave for other environments. So this is really pandering.
And Microsoft shouldn't HAVE to open source their software, though it's good that they're open sourcing some of it, and I certainly appreciate it. Why shouldn't they have to? Because software costs money to develop. And until we have some sort of legal construction or open source business model that obviates the need for closed source (at least in the limited sense of source available, rather than commonplace definition), then Microsoft is not wrong for keeping their source closed, and I'm not sure that they believe differently.
Part of copyright is the right NOT to free software that you've written, just as it's my right not to run it. I'll appreciate their open source offerings, but I don't believe that they believe much of anything about "the right side of history".
I spent decades trying to find way to avoid buying windows licenses.
And now... they're just a harmless, if slightly stodgy, tech giant hawking their OS and office suite, throwing a few good tools over the fence, and generally being pleasant to work with. Post-pandemic, in fact, I've been working full time on a desktop that's running windows (with a bunch of X11 terminals from the linux development box, of course). And not only do I not hate it, I don't even feel shame.
- Trying to Jerry-rig ACPI standard so nothing other than Windows can run on PC platform (Bill Gates memo).
- Trying to Embrace-Enhance-Extinguish Java with their fork.
- Using obscure, undocumented extensions on everything (i.e. Frontpage was using IE specific extensions to make HTML files render nicer on IE but not on Netscape).
I think that their current plan is a long/slow game to death-hug the Linux ecosystem if they can however, Google, IBM and other players probably won't allow them to do that.
P.S.: Oh and of course, SecureBoot.
It’s my understanding that the reason for their “proprietary” format was because, at the time, there wasn’t really a format that could contain everything Word could do. Also, their proprietary format wasn’t very proprietary; it was reverse engineered, and they did nothing to stop that.
In response to the OpenDocument Format, which was released in 2005.
> If they wanted to keep it proprietary, they wouldn’t’ve created Office Open XML ...
OOXML, when first released, was still pretty proprietary: it was exceedingly complex, far more than the already-existing ODF, at points even requiring reverse engineering of binary blobs and other, not-covered-under-OOXML, Microsoft-proprietary libraries.
For example: "OpenXML on the other hand, is a high-level specification which describes the high level envelopes used to embed binary objects which are included in the content. The content itself contains the binary code which can call any function in any Microsoft library and has all permissions of the person opening the document." [0]
It felt a lot like a hurried piece of work rushed through under short notice and short schedules. In fact, Microsoft even tried to rush through an ISO standardisation for OOXML, a process during which it did not hesitate from maligning and harassing national bodies [1] and academics [2] who reviewed OOXML poorly against the ODF. The very link you cited covers a few of the issues with the standardisation process: [3].
> Also, their proprietary format wasn’t very proprietary; it was reverse engineered, and they did nothing to stop that.
That is NOT the definition of 'proprietary'. Reverse engineering is legally protected irrespective of copyright and licensing.
----
0: https://slated.org/ooxml_dissecting_the_binary_blob_problem
1: https://www.infoworld.com/article/2654142/iso-publishes-offi...
2: http://deepakphatak.blogspot.com/2008/05/this-is.html
3: https://en.wikipedia.org/wiki/Office_Open_XML#Standardizatio...
s/Microsoft/IBM/g
s/Facebook and Uber/Microsoft/g
The lesson to me has not been "Microsoft's now the good guys", but "infotech builds monsters". In a commercial, property-based world, the leading firms will be monopolies behaving atrociously, for extremely entrenched systemic reasons.
As for tech's leaders: the industry not only attracts monsters, but selects monsters for success, and turns those who have (or appear to) good intentions into monsters.
The good it spits out, destroys, or drives to madness.
You can see largely useless features going in just because some corporation wants them.
I wonder what will happen if the destruction is complete. Will things revert to closed source entirely or are the corporations happy with being the puppet masters of OSS?
Not saying that it can't exist, but it doesn't exist, and that is why.
^^ with competing interests
Microsoft's interest in all things linux is a tad worrisome IMHO.
How 'bout instead of the "keep your enemies closer" strategy, MS start releasing their products as open source and leave Linux be.
Why does an extension need to be open source for a program to be called open source?
If someone writes a closed source plugin for LibreOffice or GIMP, does that mean they aren't open source?
However they are not one of the big 4 as you have listed.
I have heard MS employees talk about open sourcing some of their bigger products and the conversation always comes down to the issue of the third party code (libraries, etc) they use that cannot be open sourced along with the main product.
I'm not defending Microsoft in any way just giving an explanation I have heard over the last few years. I believe Mark Russinovich has talked about open sourcing things a number of times if you want to look into it further.
Microsoft tries to aggressively buy its way into existing mature projects.
They do have gitlab ce that has just a subset of the functionalities from Gitlab Enterprise that is open sourced.
the first line even say: in doubt use gitlab ee
the point here is that core has two distinct meaning in the op question in gitlab pr..
Gitlab ee is their core product, core here meaning their main product..
the fact that the core of that product, here meaning the base the product is built upon, was released as open source does not make the whole product open source..
So, gitlab core\main product is not open source.. But the core\base of that product is..
Frankly: yes. MS has changed it's colors but the nature of the beast is still exactly the same, don't forget that the current CEO was right there when all that stuff was going down. Browser wars, SCO trial, MS tax and on and on.
And your evidence is?
> don't forget that the current CEO was right there when all that stuff was going down.
And? Just because he was there doesn’t mean he agreed with it all.
Microsoft has provided much evidence that they actually support open source, and you all are sticking your heads in the ground and ignoring it all, while also providing no evidence to show the contrary.
Microsoft is, quite literally, famous for their EEE strategy. And, as a major corporation, their decisions are based on forecasted returns to shareholder value. Those two facts support the idea that Microsoft purchased GitHub as part of an EEE strategy on git. What we should be arguing about is how strongly they indicate that is what Microsoft is playing at; history has already shown that it is a definite possibility.
Microsoft still wants their phone and a phone needs apps. I bet somewhere inside MS a team is working hard on a version of Electron that comes with the OS and that does not have all these memory issues that the current version have and that behaves like a native app (multiple windows etc.). When a large % of apps are using that MS going to release a new phone.
EEE -> Embrace Extend Exfiltrate
or maybe:
III -> Install Identify, ...Income!
(google makes billions just by identifying individuals)
Microsoft is doing a lot of positive contributions, but I don't think they will ever become an open-source company. And we shouldn't expect every company to become open-source either. Any company that benefits from open-source should support, invest, and contribute to open-source.
[1] https://en.wikipedia.org/wiki/Bundling_of_Microsoft_Windows
Microsoft was on the "wrong side of the history" wrt. open source only because others figured out how to take advantage of the shared labor programmers put towards open source and work around the markets where Microsoft had near absolute dominance - so Microsoft was forced to play ball. In markets where Microsoft still dominates (e.g. desktop) their open source efforts are of the "let's opensource the Windows Calculator, we love Open Source, teehee" variety.
Basically, Windows wasn’t written to be open source, so the effort involved to do so is insurmountable.
(my point is that "can't", "too much", and "insurmountable" are rarely words that make sense in the realm of software, especially given sufficient time -- and any missing interfaces can be removed and re-implemented; in fact it's likely that existing open source alternatives already exist for most of them)