tecosystems - Latest Comments

Re: What Factors Justify the Use of Apache Hadoop?

Michael Dorf — Wed, 25 Jan 2012 21:15:51 -0000

I absolutely agree with you. Hadoop, buzzword notwithstanding, can be nothing but overhead in plenty of cases. I wrote a did a similar post on the same topic recently:

http://www.learncomputer.com/h...

Re: Not Dead Yet: The Rise and Fall and Rise of Java

Guest — Wed, 25 Jan 2012 16:20:13 -0000

nothing of this is news. m$ did everything to shoot java down, including slowing it down on windows and doing a copycat language.
next big language will be javascript. well, it already is. still, there are still dumb ppl who don't see it coming

Re: Not Dead Yet: The Rise and Fall and Rise of Java

Guest — Wed, 25 Jan 2012 16:18:11 -0000

c# is good for crappy stuff and bad developers. no self-respecting company will user microsoft coding "stuff"(tools/frameworks/etc)

Re: Why You Should Pay Attention to Node.Js

Benjamin Ellis — Fri, 20 Jan 2012 06:11:55 -0000

That should, of course, be 19 months!

Re: Why You Should Pay Attention to Node.Js

Benjamin Ellis — Fri, 20 Jan 2012 06:08:11 -0000

It's interesting, just 9 months or so down the line, quite how it is continuing to gather momentum. So many benefits in tooling and skills for people developing ajax applications. The number of new libraries and tools for node seems to be growing every week.

Re: How Important is Software? Generational Differences Between Software Producers

cole k — Tue, 03 Jan 2012 19:27:58 -0000

I Think software has become less relevant in recent years , with the advance of web apps.
http://worldotech.wordpress.co...

Re: Infochimps and the Inevitability of Data Marketplaces

Doug Laney — Thu, 29 Dec 2011 22:07:02 -0000

Great piece Stephen. I have been writing about the imminent emergence of data marketplaces since the late 90's. It's great to see it finally happening.

One thing tho: your piece refers to data on the balance sheet asset. This is a common misconception. GAAP does not have provisions for valuing nor reporting data assets. Even companies that are specifically in the market of buying and selling data (e.g. the credit bureaus) cannot and do not represent data assets on the balance sheet. Only in the rare circumstance of once company acquiring another and its customer list, does this get lumped into goodwill. Otherwise the growing value of data assets is in part reflective of the disparity between corporate book values (balance sheets) and market values.

Recognizing this, I have developed models and techniques for performing information valuations--quantifying its economic value--and have helped organizations build supplemental balance sheets (for various purposes including claiming M&A premiums) that do include information assets. --Doug Laney, VP Research, Gartner

Re: What’s Holding Back the Age of Data

Doug Laney — Thu, 29 Dec 2011 21:03:37 -0000

Great piece Stephen--both style and content. In the aftermath of 9/11, business victims began calling we Gartner analysts lamenting not only the horrific loss of life, but also the loss of their data. Adding insult to injury, insurance claims for the value of their data were denied with only the value of the hardware that housed it being recognized. Since that time, I have consulted to businesses to help them establish valuations for their data assets--regardless that GAAP doesn't allow them to do so. In this way at least they can determine appropriate levels of internal risk mitigation investment (e.g. security, backup/recovery). There are many other reasons to quantify the economic value of data as you have pointed out. The valuation models I've developed help organizations during M&A festivities and to market/barter with/for data sets. Happy to discuss with you or others. --Doug Laney, VP Research, Gartner.

Re: You Won’t Get Fired for Using Apache

rektide de la fey — Tue, 27 Dec 2011 18:03:56 -0000

Apache has a Maven plugin called RAT. It's one purpose is to detect files without an Apache license and to fail the build if any are detected. Apache has a review board, where patches' ownership is supposed to be handed over to Apache.

It used to be that this kind of institutional ownership was used to try to make corporations feel more comfortable about using open source software, that the Apache pennon indicated certainty & confidence and that businesses craved this.

This has changed, the need for this reassurance. The decentralized model indeed sprung up from that weakening, but what prompted that attitude change?

I'd put forth four items,

For one, open source is everywhere and we no longer regard it as alien. I'd wager 98% of people who have written code at least three times in their life have used some kind of open source code or development tool, and that 98% of software companies have at least one open source component they rely on directly or indirectly. Organizations have to opt out of open-source, and that's much harder to do than it used to be.

The FUD (fear uncertainty and doubt) has somewhat cleared. The biggest danger seems to using software that's not ready yet, and the possibility that it gets abandoned, all of which can be guarded against by downloading the source and fixing it yourself. What the past decade and a half havent shown is security risks and lawsuits for picking up OSS (unless you happen to make phones.)

Three, the programmers run the show. Not only is open source just a part of making software, but there are less business managers telling programmers how to code and what tools they need to use to do it. There are a lot more libraries than there were a decade ago, and the businesses have gotten further removed from deciding how to pick.

Four, everyone feel like IP is a primeval jungle out there, one where full and new moons have frenzied lawyers gripping knives between their teeth running through the jungle attacking at random. It's all arbitrary, so why bother fretting over the face that there is no LICENSE file for the github code you're staring at: of all the threads, joe and his weekend project are probably not gonna be the ones bringing out the elite IP attack squad to make that jungle a nightmare for you.

Decentralized seems somewhat synonymous with zero friction do-what-thou-wilt practice. It's the realization of this liberty that makes brands important, to provide guidance across an increasingly expanding and potentiated space. That's not just so you the programmer don't get fired, that's so you the programmer don't have to expend a sizable fraction of your life constantly evaluating the ever changing technical ecosystem: the brand will percolate ideas up to you.

Little last word on this, if Apache does want to maintain a presence as a reliable base for developers, quite a few Apache project wikis need a lot more authoritative and prescriptive content on building and running Apache powered systems. Apache has more commercial contributions than it used to and the danger is Apache allowing commercial entities to sell Apache-ified code while keeping the project structure and support out of Apache. Apache the brand needs a lot more than just code to serve needy app developers.

Re: Bottom Up Adoption: The End of Procurement as We’ve Known It

Guy W Hancock — Tue, 20 Dec 2011 06:44:04 -0000

When I referred to "fit for purpose" what I was aiming at was the "my iPad does this great thing and so I want to use it" (which is user/individual-centric) as opposed to the "I need a solution that I can deploy across hundreds of users and afford to support and maintain it whilst ensuring it does what they ALL need" (group or company or organization focused).
In a lovely world we would be able to see the early adopters, then try their solution and if/when it is right or suitable, deploy it to a wider audience.
The problem with lots of little, individual solutions is that it creates lots of overhead for management and maintence. For example, if you use your iPad at work and have a problem ... but the helpdesk doesn't support it what do you do? Likewise with applications, where is the data stored? Is it secure? Is it in a suitable location? And what happens if you leave and I need to know where all your work is? Organizations have wider responsibilities for data/information that they have to manage.

I'm only trying to expose the wider issues that must be considered from an organization standpoint, though not usually from an individual's.

Re: Bottom Up Adoption: The End of Procurement as We’ve Known It

steve o'grady — Mon, 19 Dec 2011 09:59:11 -0000

Respectfully disagree. Most of the above technologies are designed to appeal on an individual basis, whether to developers (e.g. open source) or users (BYOD). When this happens, questions of "fit for purpose" - if I understand your definition - become secondary as adoption has already occurred. MySQL or EC2 or the iPad may or may not be fit, but what they are is heavily utilized.

Re: Bottom Up Adoption: The End of Procurement as We’ve Known It

steve o'grady — Mon, 19 Dec 2011 09:55:54 -0000

My focus was on the mechanisms of the trends as opposed to their actual content, and was thus including Yammer, Chatter, etc under categories such as Software-as-a-Service. But you could certainly make the case that they deserve their own group because of the unique nature of their appeal.

Re: Bottom Up Adoption: The End of Procurement as We’ve Known It

Guy W Hancock — Mon, 19 Dec 2011 09:05:17 -0000

Nice piece except that at no point does it address the question of "fit for purpose" which, for an enterprise is far more than one user saying "this does what I need".

Re: Bottom Up Adoption: The End of Procurement as We’ve Known It

Krish — Mon, 19 Dec 2011 02:41:36 -0000

Another category where Bottoms-Up adoption forced the CIO hand is social services like Yammer, Chatter, etc..

Re: You Won’t Get Fired for Using Apache

Sam Kleinman — Thu, 01 Dec 2011 10:45:53 -0000

It's a truism to describe git as decentralized, but I think that's too simple. I suspect that most organizations use git in a way that is pretty centralized, and it seems that github tends to facilitate this usage pattern, even if the github service is technically agnostic on this question...

As an aside the git project itself is a member of the Software Freedom Conservancy, which provides a function much like the Apache Foundation to a collection of projects.

Re: You Won’t Get Fired for Using Apache

Stephen Walli — Tue, 29 Nov 2011 08:53:41 -0000

Okay. I said a whole lot more about this in a discussion on the foundations email list, and then turned it into a blog post: http://www.networkworld.com/co...

Re: You Won’t Get Fired for Using Apache

Stephen Walli — Tue, 29 Nov 2011 07:54:55 -0000

@sogrady, I think it's the perception of IP cleanliness, not quality that's important. While many individual developers are comfortable easily sharing software using FOSS licenses as social contracts, most corporate organizations live in a [rightly] more conservative legal climate and need the assurances of a well documented and understood IP process before they will assume the risk of use and participation.

Re: You Won’t Get Fired for Using Apache

Chas Emerick — Mon, 28 Nov 2011 10:34:15 -0000

More and more, the "Apache brand" is one that serves as a warning label for me, not a safe harbor. Quality runs the gamut across Apache-managed projects, from the sublime to the shoddy. Insofar as Apache seems to be the dumping ground for all sorts of projects (how many web frameworks need to be under its umbrella? Which bits has Adobe sloughed off to Apache this week?), any connotations of quality that may have applied in the past is fundamentally behind us IMO.

The old saying about CA, "where software goes to die", could be applied to Apache in many instances…"where open source software goes to die".

Re: Deconstructing Red Hat’s OpenShift: The Q&A

Stodge — Mon, 21 Nov 2011 18:30:10 -0000

Can I install openshift on my own machines and my own cloud?

Re: Why You Should Pay Attention to Node.Js

alejandro — Mon, 14 Nov 2011 13:05:07 -0000

I'm planning to move to an EC2 to start deploying Node.js , it is possible to run PHP ,ruby and NodeJS ?

Re: Why You Should Pay Attention to Node.Js

Vincent RABAH — Mon, 07 Nov 2011 05:15:59 -0000

Hi Stephen,

It's really a good introduction to node.js. I use it daily and it's a awesome framework to use on server side. What I really love in node.js : it's Javascript ! One unique language from server to client, allow efficiency, fast dev, money :)

I'm a French node.js addict, you could read my blog post about Node.js here : http://www.it-wars.com/categor... ! Feel free to use Google Trans widget and leave comments !

Regards,
Vincent

Re: What Would Concern Me About Android if I Worked for Google

stephen o'grady — Sat, 05 Nov 2011 15:01:27 -0000

Happy to update the piece: can you point me to a tablet optimized application that isn't in the featured tablet applications category?

I find it curious that, assuming your assertion is correct, that Google wouldn't market the higher number of tablet applications that you assume, given the skepticism around Android's success in tablets

Re: What Would Concern Me About Android if I Worked for Google

Andrepleblanc — Fri, 04 Nov 2011 23:52:03 -0000

'featured tablet applications' is not all tablet optimized apps, its only the featured ones put there either because theyre particularly good,or because someone paid to have them listed there, its by no means a complete listing of tablet optimized apps, nor does it claim to be, as evidenced by the use of the word 'featured'. shoddy journalism

Re: What Would Concern Me About Android if I Worked for Google

stephen o'grady — Thu, 03 Nov 2011 13:14:48 -0000

"In Android, applications do not need to be tablet-specific. In fact, that's considered bad form. You want your apps to work across the whole range of devices. When the XOOM launched, there were many more than 16 applications that worked well on tablets."

I think we'd agree that - with the exception of in between devices like the discontinued Dell Streak - handsets and tablets are different form factors. My Xoom, as an example, has substantially more pixels available than my Nexus One. An application designed for one, therefore, may not be designed for another. Consider the difference between the iPhone and iPad Twitter or MLB At Bat applications, as an example: they are each built specifically for the device.

So while the native Android Twitter client may work on my Xoom, it doesn't work well because it wasn't designed for it.

As for the 16 applications, that's directly from the Android Markets tablet category.

"I cannot find anything on market.android.com that makes a claim of a number of tablet applications. Suffice it to say, if you actually look on a tablet, you will see countless apps, well more than 150. Again, just because a developer had the brains to create an app that works on phones *and* tablets doesn't mean their apps don't exist. I have no idea how you can even tell from the Market site (or on-tablet app) whether an app is "tablet-specific", and I suspect I've used more apps on more Android tablets than has the author of this post."

As of yesterday, the "Featured Tablet Applications" tab on the Android market had 166 items in it, up from the 16 previously mentioned.

Are there more applications that will run on a tablet? Certainly. But as a user of the device, the experience of using blown up handset applications is suboptimal; and the "zoom" feature introduced in Honeycomb was essentially unusable.

Ice Cream Sandwich will solve some of these problems by making it easier to resize user interface elements dynamically, but the fact is that that, by itself, won't solve the problem. When the device is sufficiently distinct, it's important to design for that.

Let's say, however, for the sake of argument that there are 100X more tablet specific applications than I'm allowing for using the Android Market metric. That makes the number 16,600. That gives the tablet roughly 12% of the current iPad application catalog, all of which are designed for that platform.

Re: What Would Concern Me About Android if I Worked for Google

stephen o'grady — Thu, 03 Nov 2011 13:01:08 -0000

UI polish and design isn't mentioned here for two reasons:

1) I think it generally falls into the category of "good enough," and in some cases betters the iPhone (e.g. notifications).

2) It's not a systemic problem in the sense that patents, as an example, are. UI design is entirely within their ability to solve; the same cannot be said of the other examples on this list, for the most part.