Slashdot Mirror
SAS Mocked For Recommending 60% Proprietary Software, 40% Open Source (infoworld.com)
This week SAS wrote that open source technology "has its own, often unexpected costs," recommending organizations maintain a balance of 60% proprietary software to 40% open software. An anonymous reader quotes InfoWorld:
How they arrived at this bizarre conclusion is hard to fathom, except that SAS sells more than $1 billion worth of proprietary software every year and presumably would like to continue, despite a clear trend toward open-source-powered analytics... In a Burtch Works survey of over 1,100 quant pros, 61.3% prefer open source R or Python to SAS, and only 38.6% opting for SAS, with that percentage growing for open source options every year.
Worse for SAS, a variety of open source data infrastructure and analytics tools threaten to encroach on its bastions in data management, business intelligence, and analytics... Nearly all innovation in data infrastructure is happening in open source, not proprietary software. That's a tide SAS can try to fight with white papers, but it would do better to join by embracing open source in its product suite.
"In the paper, SAS correctly argues that open source versus proprietary software is not an either/or decision..." writes InfoWorld, but they note that the report also "put the percentage of open source adopters at a mere 25%, which is pathetically wrong." The article suggests a hope that the report "is the product of a rogue field marketing team, and not the company's official position." Adobe's vice president of mobile commented on Twitter, "I just wonder who in their marketing dept thought this was a good idea."
Worse for SAS, a variety of open source data infrastructure and analytics tools threaten to encroach on its bastions in data management, business intelligence, and analytics... Nearly all innovation in data infrastructure is happening in open source, not proprietary software. That's a tide SAS can try to fight with white papers, but it would do better to join by embracing open source in its product suite.
"In the paper, SAS correctly argues that open source versus proprietary software is not an either/or decision..." writes InfoWorld, but they note that the report also "put the percentage of open source adopters at a mere 25%, which is pathetically wrong." The article suggests a hope that the report "is the product of a rogue field marketing team, and not the company's official position." Adobe's vice president of mobile commented on Twitter, "I just wonder who in their marketing dept thought this was a good idea."
What do the other elite forces think - what do the seals use ?
Nullius in verba
Hell yeah! It knows you are reading it!
You might get tickets to a playoff game, courtesy of the vendor's sales guys.
Hmm now that I think back on other projects in the enterprise that I have worked on that is not a bad estimate.
Even back in the 90s most of the time we would just buy controls are parts of the system we did not have expertise in. Fast forward to now we are usually stuck with some closed source thing and glue it together with opensource.
If you can do 'greenfield' you can usually get much higher. But if you end up in something like an 'oracle shop' you are going to be using closed source if you like it or not...
...they're quite right. Open Source is not magic pixie dust. As long as software is made with the same broken techniques, the same broken tools, by the same broken people, it will continue to be just as broken as proprietary software. I think after a decade and a half of pro-FOSS FUD it's finally gotten to the point where people are ready to admit that the promise of FOSS has fallen well short of the mark due primarily to a lack of market incentives to ensure software is produced using best current engineering practices.
Consequently, whatever your particular need, you may find that a FOSS application fits the bill where a proprietary one wouldn't, or vice versa. It just depends on exactly what functionality you want, and there's no hard and fast rule to guide you. You literally are forced to try different packages, see which ones are buggy, and then pick the one that's right for you.
This is, indeed, a bizarre conclusion. There is no correct percentage for open source software and proprietary software. Instead, use the best tool for the job. If that tool is open source, that's great. But if the best tool for the job is proprietary and closed source, I'm still going to use it. The costs both in terms of licensing and setting up the software have to be considerations in deciding what's the best tool for the job. Most of the work I do runs on Linux (open source) and uses open source software (NetCDF, Octave, NCL, etc..), but sometimes I need to use proprietary software (ArcGIS -- which I run on Windows, WDSS-II, etc...) to get the job done. I never concern myself with how much software I use is open source and how much is proprietary. I care that the work gets done, the software is reliable, it's efficient, I can afford it, and that the results are of good quality. Mostly I care about access to the source code if I have a need to modify the source code. Sometimes I need to extend or modify software that someone else has created to complete a task, and that's when it matters whether I have access to the source code. But in that case, it's because open source is inherently the best tool for the job. However, most of the time, it really isn't a big factor whether the software is open source or proprietary. I'm pragmatic about licensing, and just want something that's going to do a good job while being easy to set up and maintain. IMO, that's how it should be.
This seems about right. Once you've introduced proprietary software into the mix, a huge amount of your time is going to be spent fighting with the software vendor, waiting for updates from the software vendor, working around the idiocy of the software vendor, etc. So, even though 90% of the company runs on open source software, you still need 60% of the workforce to deal with the proprietary software.
The percentage here isn't the story, the story is that they are recommending open source.
Fifteen years ago, that wouldn't have happened: open source was a communist virus.
"First they came for the slanderers and i said nothing."
I came across a box of personal papers from the late 1990's that had print out of license keys for dozens of programs that I bought back then. Many of those programs have open source counterparts. Except for some specialized software, I generally don't buy software anymore.
When Adobe writes "I just wonder who in their marketing dept thought this was a good idea." let's be clear about this—Adobe's main source of revenue is user-subjugating software (proprietary software) just like SAS. So Adobe isn't arguing that a user ought to prefer FLOSS, even reject proprietary software. Adobe's objection comes down to either quibbling over percentage points in SAS' recommendation or rejecting the recommendation altogether on the basis that any discussion of this is likely to bring to mind the very thing proprietors don't sell users and don't want users thinking about—software freedom.
Proprietors rely on FLOSS so they can't complain too much about it. Adobe's RAW camera software, for example, depends on dcraw, a FLOSS program which, as its developers put it, "has made it far easier for developers to support a wide range of digital cameras in their applications. They can call dcraw from a graphical interface, paste pieces of dcraw.c into their code, or just use dcraw.c as the documentation that camera makers refuse to provide".
Digital Citizen
Ha ha, you get a piece of code that you cannot see inside, and kid yourself its 'best engineering practice'??? No, it's code from version 1 decades ago, re-hacked a gazzillion times as marketing demands changes to market regardless of customer need.
And "Current"? As if 'best practices' is a fad that changes over time! Do you think code should be re-written whenever the latest development fad comes out? Nah.
The basic premise is correct, but it has nothing to do with engineering practices. It's the simple a) "do I need a feature that is only available in proprietary software" b) if I build my product around it, what happens if that company goes out of business/changes direction/screws me around, or that developer stops maintaining/changes direction/screws around with it. Sometimes proprietary is the way to go, sometimes FOSS is better.
I've heard before about the "hidden costs" of open source software. What utter crock. Closed source software has:
1) The same or worse hidden costs:
Their support largely consists of other users in support forums, with the majority of the cost absorbed by the client organisation.
Licence management costs are compared to zero as a baseline, and litigation for accidental breaches of licence is a real and catastrophically expensive danger for closed source only.
In terms of the effectiveness of the software, commercial software is largely chosen by those ill equipt to make the choice, based on marketing rather than any sensible criteria, so it LESS likely to be effective (and no, your favourite example of photoshop being nicer than GIMP or whatever doesn't change this general point, because that is consumer software in a completely different domain).
Lock in! Your bosses subscribe to the sunk cost fallacy. If you work out that it is worse than open source alternatives, you're still stuck with it because "we bought it so you better use it!". Then when it's time for contract renewal "we don't have time to swap" so you have to renew. Bullshit.
2) More up front cost:
Again, open source sets the standard at $0, and to take my most hated example of business software that is shitter than numerous open source alternatives (ClearCase), you can start the bargaining at about, what was it? $4k per head? They don't make it easy to find the cost but I think that was it. And if you are one of those people going "oh I don't understand why all of my co-workers hate clearcase because I have no trouble getting it to work and it has this one feature that is really nice in a particular use case, so..." do you actually imagine that to be worth the cost?
The sad thing about all this is: I'm not an open source / free software zealot. I don't have a problem with the idea of paying a fair amount for something that is good value for money. My problem is that IT IS NOT THE CASE, in general, for closed source software from large vendors, and SAP, in particular, is shithouse in most cases that I have seen.
I read the summary, then decided I wanted some popcorn on hand before I started reading the comments.
Carry on.
#DeleteChrome
SAS software's primary focus is on getting maximum value from analytics. A reliable, open analytics platform underpins that focus. Combining the power of SAS with open source technologies enables you to unify disparate toolsets, eliminate silos, increase productivity, foster collaboration and facilitate business agility.
Ah, a buzzword generator. Are these people relevant to policy wankers?
“He’s not deformed, he’s just drunk!”
Say I have $1000 to spend, big guys can use K or M.
I spend $400 acquiring open source software and $600 on proprietary software. Then I am set to go.
A former client of mine was paying SAS $10,000/month to host a shitty dashboard that was updated once per quarter. It didn't even come with a vanity URL. That's the typical SAS market: gold-plated clients with unlimited budgets and almost no actual needs.
We spent an afternoon rewriting this piece of shit as a HTML dump from matlab and "deployed" it on the corporate intranet.
When you don't provide added value, you quickly become obsolete.
Farewell, SAS.
lucm, indeed.
Except for Linux man pages, the state of open source software documentation is laughable. That leaves only forums for help, but those are useless and just full of rude people.
Except for select projects/products open source is just a hobby most of the time. Any day the developer just moves on with his or her life and you're left with less than nothing.
Open source software is good only if your time is worth nothing (ex. students).
^ for those who didn't want to read that wall of text, the guy basically says "it depends"
lucm, indeed.
Nearly every SAS customer I have heard of has either sued them or engaged in some serious name calling. How on earth does SAS not only continue to stay in business, but in many cases SAS will screw up royally, engage in a public fight with a company, only to have that company expand their SAS deployment.
When I see a company deploying SAS, I usually am seeing a company that has recently been taken over by MBAs. Maybe a big family company that is moving on to the third generation. Maybe a company where the founding engineers have retired. But it take a seriously shitty bunch of management to choose SAS. The sort of management that would believe some bullshit about this 60/40 thing without a few googles of how shitty SAS is.
Assume, for the moment, that this twaddle is correct. How would you even measure what percentage of your software is Open Source vs. Proprietary? Not by budget. By packages? Does a Linux distro count as "one", then, or one per installed package? By lines of code? Good luck getting that from vendor.
SAS always thinks that they're worth the world - so SAS 60%, everything else can be open source, is what they mean. They don't make a browser, email client, or Office-type software. So let open source fill in around the edges so SAS users can complete their jobs. Perfectly understandable.
I use Windows. There, I said it. I could use Linux, but the convenience of being able to interact with others (most of whom use Windows) without really thinking about it is worth something. And one key piece of proprietary software (don't use it much, but when I need it I need it) won't run in Linux, with or without Wine, and has no complete equivalent in the open source world. HOWEVER, the other software I use under Windows is mostly (by number of applications and by time they're in use) open source. My commercial/proprietary software focuses on a couple of specialties where it's very useful, providing well-designed and maintained tools at a reasonable price that works - as proprietary software should. Everything else is open source, or at least freeware, or is very very old. So I probably have SAS' percentages reversed - I'm more like 60% open source, maybe 75% if you ignore the Windows.
I have experimented with many mixtures of proprietary and open source software and discovered the ideal ratio when creating a document is: six pages in MS Word, four pages in Libre Office. Harmony and balance. However, it does slow down our team workflow. And making every document exactly ten pages doesn't speed things up either.
Maybe they just mixed up the numbers from the Burtch Works survey.
by Cyphase ( 907627 )
Yeah, it's been many years since I worked in IT. I can see it now that it took me several minutes to decipher "SAS". In academia we use and love open source, and abbreviations such as "SAS" mean little to nothing. At first I thought "the British SAS? Or is it the Scandinavian Airlines (that would be more plausible)?" I guess it's not only me having left the world of IT industry but also the arbitrariness of the statement of 60% proprietary + 40% OS. I haven't had to deal with such BS in over a decade.
"The agriculture ministry is not in charge of Gundam" - Japanese ministry official.
If you update packages and deploy in production without testing your product, I wouldn't blame the uni freshmen but your poor practice..
Every time there's a story about OSS software being less than perfect, someone always trots this tired crap out. "Oh if it isn't want you want you can just fix it!" That is complete bullshit and you should know it. If you don't, you are hopelessly naive.
First off, most people are not programmers and many do not even have the request problem solving, analytical, and mathematical skills to become one. If you aren't a programmer, you can't just go and fix software. Becoming a programmer isn't magic either, you don't go and read a book and then you are good. It takes years of experience to get proficient, and decades to really master and is something you need to spend a lot of time on. If you think you are some hot-shit programmer and you "picked it up just by reading" and "just do it in your spare time" then guess what? You aren't near as good as you think you are.
Second, even if someone is a programmer they may not have the requisite skills or knowledge to deal with a piece of software. Not all software is created equal, not all problems are the same to solve. Someone might be a programmer who's actually pretty good, but knows about making database code because that's what they do. However if they are trying to implement an algorithm for processing audio they might be lost because they don't understand how that works, it is another set of knowledge.
Finally, even if someone does have the skills, knowledge and experience to do it, maybe they just don't want to spend the time. We all have only so much time to spend in a day, maybe they are not interested in dropping a bunch of time to fix something that is to them just a tool. They'd rather pay to have one that works and spend their time on other shit.
So knock it off with the "oh it is open just do it yourself" crap. That is extremely silly, and you know it.
"First they ignore you, then they laugh at you, then they fight you, then you win."
What is SAS by the way? I know what R and Python are though...
An even simpler rule: Only use propriety software for which you have a reasonable and tested open source alternative, in case you really need this piece of software and the copyright protection decides you can not use the software this week, after you bought the software and are paying maintenance fees for it.
I usually send an e-mail that they have 24h to jump on a plane (Autodesk is on the other side of the world) and fix it, after that I will use my rights according to EU law to reverse-engineer the copy protections (collaborating with other over the internet, so effectively publishing the "solutions" as well). If they indicate they need more time, we have to discuss damages, which is more costly than flying a team over here every day.
This has happened many times, so now we are moving to Rhino 3D for 2D drafting work, just to get more sensible licensing, free conversion to networked licences and a much more productive interface as a bonus.
There is also GNU PSPP as an open source SPSS alternative and Octave as a Matlab alternative. For more stuff: https://theouterlinux.com/rese.... I tried to pick as many cross-platform open source software as I could. Any other suggestions would be awesome.
Everyone knows by now open source is a cheap way into market with a product. I know its not popular here to denounce open source, but its clear some people find support worse and lacks function over a proprietary software title. People use Libre Office because they are cheap, not because its better than Microsoft Office. Obviously open source give you a chance to find that more perfect application. But how well its supported or how long it is. Well that's something else.
...to commercial solutions in this context is that there are many more experienced users around. Until recently I was working in science and was trained in data analysis and statistics with open source C/C++, Python and R frameworks. Now I am looking for a job in industry and encounter two types of job announcements: those who want experience in data science with Python and/or R, and those who want experience in data science with SAS, SAP or other commercial frameworks, which I can immediately strike from my list.
Due to their focus on corporate environments and lack of free amateur licences, an average mortal being cannot afford the high costs of such systems just to learn how they work. So the only way to get any experience with them is to work in a company which uses them. In contrast, I know teenagers that are already apt in Python programming.
In the current situation, where there is a lack of data scientists, companies that use open source software have a clear advantage over those who use proprietary software.
As a community we have had the wrong focus with the fight over open versus closed software. It doesn't matter. While I prefer my software open source there are times when it just won't be made unless a commercial entity sells it under a closed source model.
However the important fight that we have long neglected, and continue to do so for the most part, is for the open access to our data. I don't care how I created my data. It is mine and I should never be held hostage to access it. When an application is released it should include a document that explains how the data is stored so that other applications my access it. That way if work stops on the application for any reason (the person on an open source project stops working on it, a company goes out of business, ...) then it may be possible that other applications exist that support the format. By having the file format available it makes it easier for other developers to write import/export routines.
There should be laws that protect people from determining the format of data files in order to provide support in their applications. In addition if they publish the details of the file format they will be protected from civil lawsuits.
Too many companies use the format of their data files as a way to lock in customers and with every new release they change the document format. The data is mine, not the company's. I should be free to use the program of my choice to work on my data. And if someone sends me a file I shouldn't have to be required to own a copy of the application just to view the file.
Make a better application and I will buy it but only if I can move my data over to it.
A few reasons that commercial software is bad:
1. Better support. Commercial vendors like IBM, Microsoft, and Oracle have nothing but contempt for you when reporting bugs. At least with open source, you have a good chance of getting the developers to fit it, or failing that you can fix yourself. With commercial offerings, you are stuck.
2. Generally better quality. Even though OSS stacks are newer than their commercial predecessors, most metrics now seem to suggest that OS's quality has overtaken commercial quality, and continues to pull even further ahead.
3. Vendor lockin - The Most Expensive Blunder for IT Incompetents, simply isn't an issue.
4. No involvement with companies that fund patent aggression, and attack your business, like Microsoft.
5. No license agreements that allow commercial vendors to audit you license compliance.
6. Cheaper to deploy and maintain - at least running Linux desktops organisation wide in a large deployment, instead of Windows, can save hundreds of thousands in license fees, and massively reduce IT resource demands to support. Same goes for databases.
7. Weeds out incompetents. Low skilled IT people will struggle with non-windows tech.
Look, they are selling closed software. Frankly it's shocking they recommend any open source. The numbers seem random to me, but then I did not read their so called report.
excitingthingstodo.blogspot.com
You're a stupid motherfucker, whining about the length of a post that was pretty short. You also seem to lack basic comprehension. You don't contribute anything useful to this site or to society in general. Fuck off.
WHAT? A commercial company releasing a pseudo-intellectual white paper to support a position that benefits them? I'm shocked! Besides, what is this, 1998? FUD? Really? You know the company that started the whole FUD thing just added support for Linux to their OS?
I had a problem, so I hired SAS. Now I have 2,174 problems and no money to fix them.
SAS' motto should be, "Our Cure Is Worse Than The Ailment"
Just cruising through this digital world at 33 1/3 rpm...
My company HEAVILY uses SAS. They use it for the predictive model. They just recent upgraded to SAS Visual Analytics. That requires Python and runs (no shit) 39 web applications on one Tomcat instance. It literally takes the server 45 minutes to boot up! It's ridiculous. It's purely a Java system that uses Flash for gods sake! It's cost is UNGODLY! for two servers (test and production) each with 24 cores was $100K+++++.
The Truth is a Virus!!!
The vast majority of line of business applications are proprietary. Just as an example show me all of the quality open source job cost accounting packages out there that a medium to large sized contractor might look at.
sometimes software has unexpected costs. Never heard of that before.
Imagine if I wrote a high school essay claiming Hester Prynn in Hawthorne's Scarlet Letter was an aloof buffoon because I heard some kids telling jokes about her before class started.
If you want to make a graph to support your insults, maybe you should make sure the graph itself is not stupid. Being a self-righteous cunt is not enough, you need to dot your i's and cross your t's. I'm sure you'll do better next time now that you are aware of it.
lucm, indeed.