Slashdot Mirror
Lack of Innovation in IT Holding Companies Back?
bednarz writes to mention that Google's Dave Girouard, manager of enterprise business, is blaming a "crisis" in IT and the "insane complexity" of technology, among other things, for the lack of innovation that could allow businesses to grow. "A lot of things that people think of as core IT functions need to disappear into the ether so that the IT organization can properly focus on the value-added [activities]," he said. "Information security, as critical as it is, needs to be taken care of by organizations who live and die by it, who invest the money, time, resources and staff. Why should every company in the world have to build up their own expertise and have to maintain servers and provide security?"
Google. What a mystique! They can 'innovate' new forms of -
2 .html
r essLan=1230&lan=3
e s/000663.htmlr es/google_email_troubles_continue.html
Cross-site scripting exploits:
http://blog.outer-court.com/archive/2007-01-01-n1
http://blogs.zdnet.com/Google/?p=338
Exposure of personal and sensitive data:
http://www.finjan.com/Pressrelease.aspx?id=1261&P
Data loss:
http://dream.sims.berkeley.edu/MT/vanhouse/archiv
http://googlewatch.eweek.com/content/google_featu
Site failure:
http://status.blogger.com/
Privacy violation:
http://www.google-watch.org/bigbro.html
http://www.google-watch.org/krane.html
"Flyin' in just a sweet place,
Never been known to fail..."
Just look at J2EE, EJB in particular. For many clients, it'd be cheaper to just write your own custom remote objects system using RMI. Cheaper and a lot easier too. "Enterprise" anything is typically very complicated and poorly documented. It's sad to see how bureaucratic it's become, but a lot of these things are, I think, complicated just to make work for people like consultants. I look at half the stuff that I have to work with, and it's far more complicated to get these huge, unwieldy apps to work together than to write most of the code.
Software is written badly. Seriously. Yes, it's complicated, but by now, home computing (and related devices) should be just that, devices.
I'm an admin.
The server room work-around: multiple servers, physical and virtual, for redundancy. More for software problems, than hardware. It's embarrassing.
And I've programmed. For every decent programmer on a big project, there's 10 lazy ones. Not lazy as in not doing their job, lazy as in take a shortcut here, use an API call that you know leaks memory because it's one line versus twenty. Really just acting human on a boring job. Anything for speedy output and a quick finish - and management encourages this. Often even tells them to do it.
Small crap adds up quickly. API over API. I believe this is why the best projects have a small group of main workers. LLO - limitted lazy overhead. With a bigger group, there has to be an accountability (think buildings/structures/planes) that just isn't happening yet.
Longterm uptime: hardly ever the dev's problem. Testing is boring. New features (whether they're needed or not) are fun.
Everyone knows its a problem, and slowly, I hope it will (continue) to change.
Right now, there's a move to making servers smaller/clustered/redundant. This is good, but it's still avoiding the main issue.
It is smart, very smart, if...
You get the drudge work standard maintenance & operations done by a trusted qualified vendor.
Medical Device firms have been specializing for the last 15-20 years and their is now an ISO standard for Strategic Partner relationships.
One company does manufacturing only for a range of "Medical Device Companies" but Strategic Partners often do all the manufacturing.
Only the trade secret, bleeding edge, developmental and core competancy items are retained in house.
Well, let's follow this guy's logic.
Why should every company have an HR department? Why should every company have a payroll department? Why should every company have to maintain an accounting department? Why should every company have to maintain upper management? Why should every company need people to work in them, or janitorial staff, or mail rooms, or anything?
Let's just go with the NeoFascist extreme, where there is one company that does accounting, one company that does HR, one company does data processing, etc, and every "company" is just outsources each and every kind of work needed to be done to these companies... so executives no longer have to be tied to the office: they can just make three decisions per day, then head off to play golf.
Damn, it's hard werk being a Republican. Cut all taxes! Deficits don't matter!
The way I see it is that you have trade offs. You can make things "simple" like Microsoft does by having software that makes assumptions about what you want to do. If you have minimal and non-specific needs, then you won't have much of an issue. You'll also be able to get away with less skilled workers to accomplish your basic goals. But if you want to do something very specialized and the assumptions don't support it, then that approach will get in your way. And your less skilled workers will become a detriment if you need to move to something a bit more complex.
On the other side of the coin you have other OSes (*nix OpenVMS and others) that have a good deal more flexibility in terms of allowing you to do virtually anything. But this will require more skill in your workforce and natually more complexity. There is not currently a way to have less complexity and a high degree of flexibility and power. There just isn't. Even Microsoft is learning that lesson as they add PowerShell (previously Monad) to Exchange 2007. They've finally seen the light that what you really want is a set of powerful and small tools that do one job well (the CLI) and then you layer your "ease of use" on top of that. So I expect that future MS products will probably earn the gripe of being "complicated" by less skilled people who entered the IT industry in the 90s as paper MCSEs.
There's no way around it. Computers ARE complex machines and they become even more complex when you want to do something really unique and innovative. This is why there is no equivalent of Exchange on the Mac OS Server platform. Zimbra is about the closest thing and it's not exactly friendly. But if you're a real IT guy, that's not a problem.
-"...bad old ideas look confusingly fresh when they are packaged as technology" - Jaron Lanier (Digital Maoism on Edge.o
He's right about the problems.
It's not clear Google is the answer.
He doesn't explain how THEY deal with complexity except by throwing money and servers and data centers at it - which is pretty much how everyone else deals with it (and which is self-defeating eventually).
Self-defeating - almost a definition of the human condition.
Take IT in small business - or don't. Don't even get me started. I'm constantly getting clients who complain to me how long it takes for me to solve their problems. While I don't say it directly, the real issue is that EVERY small business using computers has done it wrong from day one. If they hadn't screwed up from the beginning, they wouldn't need to call me to straighten out their mess.
I doubt there is ONE small or medium business that hasn't screwed up. They bought the wrong machines, they bought the wrong OS, they bought the wrong application software, they bought the wrong networking hardware, they set up everything wrong, they didn't plan anything, they didn't ask anyone how to do it, they don't train, they don't back up, they don't maintain anything, etc., etc., ad nauseum.
Then they wonder why it takes a tech two days to do something as apparently simple as rebuild a PC.
It's because they are SO fucked up NOBODY could have done it faster.
NO human can take responsibility for their actions.
Richard Steven Hack - This sig is TOO GODDAMN SHORT TO DO ANYTHING USEFUL WITH! MORONS!
Complex systems require a certain amount of vertical application expertise by definition, at least if an effective level of support is the desired end result.
Often this expertise takes several *years* to develop effectively.
In many industries, that expertise also tends to be company-specific in nature (and not just industry-specific) because the application software is very tightly coupled to company-specific business rules and procedures. Other companies in the same industry may do similar things using a very different set of procedures, requiring different software (and different expertise).
The time required to grow the required company=specific expertise for effective support often makes the outsourcing of support for those applications impractical unless you spin off your in-house support staff as a separate support company. That sometimes works well.
Mainframe/UNIX Bit Twiddler and long time Windows/Linux Hobbyist.
The Theorem Theorem: If If, Then Then.
- Servers, e-mail, security, and whatever else "Google Apps" offers now or in the forseeable future is a small portion of any company that has any kind of serious IT going on. Most companies have a lot of custom apps, or at least highly-customized "off-the-shelf" solutions. I don't see Google offering a solution to reduce costs in that area.
- Any cost savings from outsourcing will not translate into increased corporate R&D. The PHBs and bean-counters will just give themselves raises.
- The PHB raises are not just because of greed, but because "forward-looking" on today's Wall Street means the next quarter -- thus no R&D.
- The biggest barrier to innovation is corporate culture. The Google exec even touts that as a benefit within Google, but fails to mention it as a barrier in companies other than Google. I still remember the first time many years ago that I made what I thought was an innovative suggestion within a Fortune 500 company, and then hearing the response, "that's nice to have," and feeling flattered. Only a few minutes later I came to realize that in corporate-speak "nice to have" really means "not nice to have" because "nice to have" is meant to mean "it's nice to have but not needed to have, so we won't be doing that".
- Sarbanes Oxley. Nothing much else happens in the IT departments of publicly traded companies these days except for SOX compliance.
- Innovation is easier when everything's in house. Having servers, e-mail, security, etc. outsourced creates barriers to deep integration and high degrees of customization.
Despite the wrongness of the Google exec's assertion, it will be well received by PHBs because they'll get to cut costs while simultaneously pretending to be in favor of innovation.Dead on. But it gets worse, the remaining R&D is focused on short term Iraq related technology particularly IED countermeasures.
Would you rather own your home, or rent it? Would you rather rent a car or own it?
Those are bad analogies because that is apples and oranges comparisons.
By default if given the choice you should own a home (even though the bank still owns it for the next 30 years) because house prices increase whereas you car price decreases.
This is what makes leasing attractive to some, but that is almost the same as owning.
By what you really mean would be would you rather take public transportation to work or own your own car to work.
The traveling to work is not the end result of your actually work and not even an actual process of it other than getting your feet to it. Given the choice of maintainece you most likely can't fix your own car anyways so if it breaks you have to hand it off to a 3rd party.
If public transportation breaks, you'll see a delay but you won't actually have to pay for it directly and chances are they will route a new bus to your pickup.
But you can't add your own custom products to the public system and you can't make it go to exact places at any given time like your normal car. So both have their benefits and detractors.
"I am the king of the Romans, and am superior to rules of grammar!"
-Sigismund, Holy Roman Emperor (1368-1437)
How is a company supposed to innovate if they outsource their staff? Solutions that are honed to very specific company needs can only realistically be achieved by staff that is there who has the business's interest in mind. Outsourcing is a step backward for innovation, not a step forward. Taking security as an example, since it's in the headline of the article....my team has saved the company quite a bit of money, time, and effort ... and is able to put management tasks down to less-skilled operations staff rather than have us involved by coming up with a custom firewall/vpn solution built around iptables and openswan. It works flawlessly at numerous branch office sites, and we even use it internally to segregate networks. Any outsourcer would not only provide less effective service in general, but would have used expensive to license and support hardware and software that likely is a real pain to manage.
Many companies farm out sales. I can think of a couple easy examples.
First is manufacturers. Many use indepenent sales reps (manufacturer's reps) rather than an in house sales staff. They also use distributors who handle product sales to retailers.
Next is travel related businesses. They use all sorts of commission based plans to farm out their sales.
There are generic sales firms that will sell *anything* for a commission. You give them the leads and they will hit the streets. I've dealt with those guys. They are mercenary.