How much longer do you all think Oracle will be around as a company? They burned all bridges with engineers long ago, and they're not going to be able to keep selling their blatant and expensive lock-in strategy to management.
Maybe when they finally do go under Google will be able to buy Java and MySQL...
I don't think you realize how the enterprise ecosystem works. There are legacy Cobol systems scattered all around the place. Corporations stick with their platforms for decades. This isn't the web development front where we change platform every couple of years. In corporations change is slow because there are significant sums of money invested in people and procedures. You just don't change that for the next flashy thing. And thus the future for Oracle is much more promising than it is for Google or Apple.
> the future for Oracle is much more promising that[sic] it is for Google or Apple.
That is a bold statement. Google has ~1/3 of the world who have access to the internet as regular users. Apple is a hardware maker with a very desirable array of products. I won't say I am a fan of Oracle, and I agree with you on how the lifecycle of enterprise adoption works, but that is a very far fetched conclusion.
First, large companies need to adapt quicker; while this has always been the case, the pace is accelerating. As pointed out above, Oracle has done some shit to alienate a lot of people, and have been for a while.
New companies grow up to be "enterprise". While they ard gouging and alienating their current customers and potential employees, people are NOT building on Oracle. They aren't being locked in.
As old companies (or their management) retire Oracle will have adoption rates like cobol. I am not saying tomorrow, nor this year, but Oracle is going to make big changes or see their MRR fade out with the salesy slicks who brought them on board 2 decades ago.
The internet is pretty fickle though. I still remember when people switched from Altavista to Google and from MySpace to Facebook. It didn't take all that long and look at those old companies now.
Whilst Oracle has done absolutely nothing the last few years other than crappy, me-too products they do have a LOT of luck on their side. Why ? Because nobody is touching the EDW (enterprise data warehouse) space. There are no new startups. No disruptors. Nothing. All of the excitement and growth is happening in the analytics space where companies are just sucking all the data from the EDW into Hadoop. And then doing feature extraction, modelling, machine learning etc there rather than doing it in the database.
So nobody is really going after Oracle/Teradata/SAP etc core business any more. It's just kind of forgotten arguably like Windows or Office.
You are confusing consumers facing products( Altavista, MySpace ) with corporate facing customers. The corporations have invested 10s of million or more developing products based on Oracle products. Changing from Oracle to PostgresSQL is possibly many years of work and multiple projects for a large company.
I agree the companies are unlikely to rework the current code from Oracle to PostgreSQL (or some other database), but they'll definitely think twice when starting new stuff.
When managers start a new project, they don't want to think twice. They want to grab the ass of someone else in the company who's already done the same thing and leverage their expertise to reduce the cost of the project.
My company is still stuck using CVS because when a new project starts up, they want to use the people who are already locked into the old tech and so they end up going with what people already know. Even teams that started with git have downgraded to CVS once the managers starting looking for ways reuse resources.
It's always a cost/benefit evaluation. When the costs stay roughly constant, the outcome is usually to use the same stuff as for previous projects. But when the vendor significantly increases the costs, like Oracle does in the recent past, that may have significant impact. It may be easier/cheaper to switch to a different database, for example.
I'm not saying all companies will end up doing that, but some certainly will (and already do). They may not be vocal about it, though, because they don't want to piss Oracle.
> Changing from Oracle to PostgresSQL is possibly many years of work and multiple projects for a large company
Enterprise IT is broken enough that those multiple system replacement projects were going to happen anyway because of accumulated technical debt and broken development and maintenance processes, and replacing a backend DB as part of that replacement is a rounding error in expense. Sure, it takes a lot of time for the transition to be complete, but lots of those transitions off Oracle are already underway.
Speaking as an enterprise CRUD developer wage-slave, if you need a PhD from MIT to figure out how to migrate from Oracle to a sane database like Postgres or SQL Server, you probably should work in a different field or at least a different project.
More realistically, it's department politics that are going to prevent any sort of migration.
"The internet is pretty fickle though. I still remember when people switched from Altavista to Google and from MySpace to Facebook. It didn't take all that long and look at those old companies now."
Past performance is no guarantee for the future. We zillions of PC manufacturers before we got the likes of HP and Dell, and twenty years later, we still have them. On the desktop OS/application front, the same happened with Microsoft becoming the main player.
The common theme, I think, is that things are chaotic while there is room for everyone to grow, but get more predictable once the entire pie has been divided (aside: that's an argument in favor of Uber's "land grab at any cost" strategy)
Oracle, Google, or Facebook may eventually disappear, but it won't happen at he time scale at which Altavista or MySpace did.
At a prior employer, the CIO was so pissed off about Oracle licensing that he seriously considered freezing Oracle versions (as a prelude to migration off the platform) and buying third party support from Rimini Street.
If enough businesses pursue that option, then the cash cow of annual support dries up and Oracle is in a world of hurt.
"If enough businesses pursue that option, then the cash cow of annual support dries up and Oracle is in a world of hurt."
They could just raise licensing costs. They've already won on API copyrights. Businesses would have to move everything off Oracle entirely with new API's to avoid legal risk. Such moves, esp if Oracle docs suck as much as claimed here, can cost a fortune with huge chance of breaking stuff. Oracle just has to keep cost of licensing lower than that of a move.
They also found a new way to apply that same model with 0365 and their continual effort to add new features that only cost a "few dollars more per user per month"
Squeezing locked in customers is the Enterprise Software way... all of them do it.
>Squeezing locked in customers is the Enterprise Software way
No doubt about that, but for basic infrastructure like DBMSs the customer of the future will be another tech company. Try extracting the same margins for a database system from, say, Airbnb or Uber as from a mid-sized bank founded in 1888.
I don't disagree in general, but Oracle is a tech company and technology has a particular function in our society that wasn't invented by MBAs.
The role of technology is to drive productivity growth, which is ultimately the only thing that improves living standards (or at least has the potential to do so).
A pizza sold in 1987 is just as useful as a pizza sold in 2017 measured by its nutritional value. But the same is not true of technologies from 1987.
Oracle's lockin doesn't just come from their database. There are also behemoth business/logistics/financial applications that exist only as PL/SQL packages and a bit of Java UI. Moving off Oracle Applications is a lot harder than switching databases.
Even at their most dominant, Microsoft kept adding things customers wanted. AD was a big leap forward for customers, when Microsoft could have rested on the domain model. SQL Server kept adding capabilities, Exchange, etc.
All designed around Vendor Lockin, then putting the screws to business with various and complex licensing
CAL's for example would never be possible with out Vendor Lockin, What kind of extortion scheme do you have to have to sell a server licence then charge for every thing that wants to connect to the server.....
CALs - you mean like literally every successful SaaS solution on the market today? CALs are alive and well and aren't EVER going to die. Salesforce.com? CAL. Openair? CAL. Workday? CAL.
I don't think that says a whole lot about Oracle's growth opportunities as everyone moves to the cloud.
Direct licensing costs are not all that matters. The restrictions on architectural choices and the uncertainty created by these complex licensing structures could hurt them a lot more.
You are totally correct about how change is slow in enterprise, and this favors Oracle tremendously.
But I don't think you realize the extent of Google's ubiquitousness. Innovation is a survival strategy, and between the companies you mentioned, Oracle comes last in that front.
Google's lack of a rich enterprise cultural workforce consisting of account managers paired with a culture of sales and services for profits makes a great deal of enterprises nervous since this is how they're structured for business for several decades. This is where Google's partners that are more versed in such traditional, engineer-hostile come in, but then you wind up where massive companies start to rely upon maybe 1000-person companies that will be crushed by the onslaught of horrifically mundane, non-scalable problems endemic to a customerbase consisting of non-technical people and sub-par engineering talent as standard. These companies know that they can't use these smaller companies and are not run by stupid people (usually), so they have really little choice among technology vendors that make most of the HN crowd cringe.
So Oracle's future is based on those large enterprises, while the more nimble, those that will be large enterprises down the road, steer away from Oracle? Maybe Oracle realised that they won't be winning the smaller one's over and now decided to squeeze the larger ones even more.
Nope. Once you're big enough to start worrying about public company financial processes, your new CFO is going to buy Oracle Financials.
This particular price change is all about customers moving away from old RISC systems and IBM and towards x86. Amazon dramatically overprices Intel cores, so Oracle wants to capture that value. Cheap licensing for x86 is subsidy for AWS and Azure. They are spiking the pricing to push you into Oracle/Sun hardware or the Oracle cloud offerings. "Hardware and Software Engineered together" isn't limited to technology engineering!
Or they realize that your small nimble companies will "grow up" and hire experienced managers who will put in place rules that require mature vendors and Oracle gets another generation of company. Happens a lot, sadly. Idealism gets beaten and the people who don't like Oracle jump ship to new and exciting nimble companies.
DB2 is a bit of a different kettle of fish for a few reasons:
* DB2 rolls new features in at their top pricing tier, then trickles them down. So things you could only get in the top tier in 9.5 (for example) are now available in Workgroup edition in 11. Oracle keeps very old features (e.g. partitioning and encryption) as expensive add-ons.
* DB2 is still pushing out new features. PureScale has (for example) moved from parity with RAC (clustering in one datacentre) to well beyond it (metro clusting across a region).
* DB2 respects pretty much any VM technology for sub-capacity licensing. Oracle pretty much require you buy OVM.
Top-tier DB2 and a fully loaded Oracle DB are pretty similar pricing (depending on the specific deals you can obtain from your account manager), but DB2 keeps moving, and many of the features you might care about can be had in much cheaper editions.
The problems DB2 face are mindshare (easier to find DBAs for Oracle, many people still think of DB2 as a "mainframe thing"), and there's still quite a few "enterprise" apps that only support Oracle DBs.
Don't forget the risk of an expensive screw-up when switching to a new system. Unless there are clear business advantages, switching to a new system from an old, archaic but battle-tested system doesn't look enticing for most companies.
> And thus the future for Oracle is much more promising than it is for Google or Apple.
I'm not sure that's true; it's may be true that Oracle has more Enterprise penetration now, but enterprises with significant Oracle dependencies I've encountered are mostly migrating out (largely to Microsoft) -- including enterprises that also have substantial decades-old COBOL mainframe legacy that they aren't as actively working to retire.
Oracle will be around for many, many decades to come. Especially since in the enterprise world nobody could care less what engineers thought. Managers determine budgets and capabilities. Architects determine what to purchase and how it fits together. Engineers implement. Pecking order that's been around as long as Oracle has.
The biggest obstacle to Oracle's future success isn't price or other databases. It's the analytics space i.e. Hadoop/Spark which is relegating big iron SQL databases to being dumb CRUD boxes. It's why both Oracle and Teradata moved into the space but have struggled to get any traction against Hortonworks or Cloudera.
Oracle actually has a product called Big Data Appliance (and has been for a couple of years IIRC), that offers Cloudera-based clusters. They also offer that as a cloud service so you don't have to actually own the hardware.
Not sure how much traction is having vs Horton or Cloudera, but I'm sure a lot of companies that already buy Oracle would look into it if needed.
But those engineers retire, new ones without memory of Oracle's past take their place, and all will be forgotten.
Remember how Microsoft was the company many engineers hated when the Halloween Documents [1] came out? Yet, they have managed to patch up that relationship since 1998.
Yeah, my new boss laughed when we complained about Oracle. He adamantly insists it was the best database in the world. He also knows only one database.
It is a fantastic database. Best is a subjective term but for a MVCC translational workload it's amazing. Note that I'm talking specifically about the database itself, not the cluster fuck of end user software built atop it (ex: Oracle Financials).
The phrase "tune it via their CBO" doesn't make sense.
Oracle DBA here who hates their licensing, practices, etc. as much as the next guy, but I have to hand it to their query optimization. I've seen their optimizer get better with every major release, and yes, regression testing will always turn up a few problems but Oracle gives you the best tools to analyze and override the decisions that the optimizer is making when necessary (and I'm not just talking about basic hints here)
If you changed the case of a table name, this means you had to change whatever was generating the SQL since mixed or lower case requires table or column names to be surrounded by double quotes.
So, you changed code and didn't test it. Yup, definitely Oracle's fault.
Except that's not what I'm referring to. As with any cost based optimiser, it's only as good as the statistics it has to work with, and if you've finally carefully tuned your database with an appropriate histogram on a critical table, and found that after an upgrade they "tweaked" the CBO to fix a statistics calculation then you are right back to looking at explain plans to fix your performance issue.
Sometimes you just aren't going to pick up this sort of thing in testing as you'll only know about it under full production load. Their CBO is pretty powerful, but by and large it's often difficult to know why it makes certain decisions. And that's even when you use their full suite of profiling tools!
In all fairness, I can't really blame Oracle on this one. That's a fairly well documented issue, because Oracle places the hash value of each query into the shared SQL pool, and if CURSOR_SHARING is set to exact and you've stopped the query from aging out then Oracle will see it as a different query.
That's not even really a CBO issue. But I hear your pain :-)
> Oracle sells to managers and executives and are very, very good at it
I agree, but the point is that they are getting worse. Their cloud shift is not panning out as well as they wanted, I think, and they are actually opening opportunities for competitors: if you have to make significant efforts to migrate your existing solutions, you might as well move somewhere nicer. And by basically threatening their own partner ecosystems (which won't be needed in the new oracle-cloud era), they're losing steam in sales and presales.
(Of course, it doesn't help that Safe Harbor first and Privacy Shield now have been shot down, so European businesses are not keen on moving to US-based clouds. The sort of businesses Oracle pines for, does care about these things.)
Isn't gross income a bit misleading if they need to buy out larger and larger companies and harvest their products/good-will with an increasing portion of that gross income?
I've had two jobs in the past that have been in Oracle shops and from my experience those companies were so fully invested in a lot of the Oracle suite (think WebLogic, ESB, the database etc) it would be like turning a huge oil tanker around to pivot away from it.
And don't forget the reams of PL/SQL that often sits at the core of a lot of business logic
Eh, that's true just about everywhere... Code = Money.
You spend some money, get new and hopefully better code, and with some luck you will get a return on the investment, ie more money than you spent on doing it.
Unless you are doing it for fun. Or don't know what to do with all the money you have.
I laugh at how the only thing discussed here is open source software and the cloud. People forget that there's a whole entire world of enterprise software and systems.
As someone else mentioned earlier, mid to large size companies don't replace systems like people do here with Javascript frameworks. It is really expensive and costly to replace systems that have been around for decades. The business logic tied up in some of these systems is so vast and complex that it is cheaper to keep them around than to replace them with something new.
I would restate it that there is deep business logic encoded solely in a system that grew organically over decades. It's not always that it would be hard to rewrite it all... It's that no one knows or has documented anywhere except this organic code base all of the requirements. And specifically no one has encoded WHY.
So rewriting and keeping the same logic is very difficult and the business risk is very high due to all the unknowns: features, time, compliance, correctness, performance.
I agree with the sentiment, but they do make fantastic acquisitions though. of course Sun with Java, MySQL, and VirtualBox, also Dyn, Ravello, Addthis, Crosswire just to name a few
Since reputation of Google here is it shuts down projects/products. Do you mean Google buy Java/MySQL and shut them down? I do not see any other reason to buy either of those considering they can and are using Java and MySQL as and when they need.
Oracle has a huge client base in the gov't and healthcare sectors (at least in the US). Gov't and healthcare also lean heavily on solutions vendors who tend to use Oracle. This is a sector that is easily 10 or 20 years behind the curve in most places, so Oracle offering their me-too product line-up is safe and reassuring.
Maybe when they finally do go under Google will be able to buy Java and MySQL...