Hawaii Needs a Cabinet Level CTO & Google Apps

Comment

Comment by Russell Castagnaro on February 16, 2010 at 10:22am

Maya,

I mean no internet access allowed at all, not just sometimes. They can only access the NGN. No public email either. Nada. Gears wouldn't help.

Comment by Ken Mayer on February 16, 2010 at 10:22am

I think both are useful, but I like the idea you proposed in your initial post - "Maybe a live TechHui gathering would be appropriate. There might be people from the state and DOE who would be willing to discuss the issues. I'd be happy to participate." If you participate along with someone in a similar position at the DoE, I think it could be very productive. It would allow private industry members to learn about your efforts and vice versa.

Unconferenz 2010 is coming up Saturday, February 27th. (hint, hint, nudge, nudge, wink, wink)

Comment by Russell Castagnaro on February 16, 2010 at 10:21am

Plenty of others have already said that legislating a specific technology is a bad idea, and I agree. I will talk about the state CIO/CTO positions that are being discussed in the Legislature now.

Hawaii has a federated IT approach and this approach has many advantages and disadvantages.
PRO:
Agencies accountable for own services
Agencies who are most publicly accountable typically have the best services.
Slower moving agencies don't slow down faster ones.

CON:
Agencies don't share purchasing power (there are 5 different Cold storage systems that I am aware of).
Solutions created for one agency are not easy to share.
Redundant personnel


There are relatively advanced agencies such as the DCCA or tha AG's Office and much less technical agencies like Agriculture (I don't need to mention others here). The reason that the more technically advanced agencies were able to progress is that they had a high level champion (i.e. Director, Administrator) who felt a responsibility to improve the services that Agencies provides.
You'll notice that Government workers don't get paid more for successful deployments. There are very few positive rewards. They do get pulled up in front of the legislature, the unions, etc.. when things go wrong. Its a scary thing to be an innovator in government right now.

Will a State CIO or CTO appointed by the LG or the Governor have the same level of accountability that the IT directors in the agencies have? I know that a good CTO would be better than the current situation, but a mediocre CTO might not be.

My personal take is that a combined approach where there is a federated system with many shared services that are centralized would work best. There have to be tangible (read budgetary) incentives to leverage the centralized services and they should probably be optional or else there is no incentive to have solid core competency in the centralized services. Unfortunately, it also seems like the hardest (politically) approach to implement. I am afraid that centralizing our existing state government IT will be an expensive prospect.

Regardless, I am interested in seeing how this all plays out. I think that we will all benefit in the long run from the IT changes that will probably be enacted this year.

Comment by Mika Leuck on February 16, 2010 at 10:12am

Hi Russell - They can use Google mail offline with Gears or whatever mail client they are currently using. Google provides POP and IMAP access.

Comment by Russell Castagnaro on February 16, 2010 at 9:44am

You know, there are MANY state employees who do not have access to the Internet from work. Gmail would not work for them.

Comment by Daniel Leuck on February 14, 2010 at 12:33pm

Cliff Frost: I think this discussion is going in too many directions to really work well on a BB. Conversation would likely be more effective.

I think both are useful, but I like the idea you proposed in your initial post - "Maybe a live TechHui gathering would be appropriate. There might be people from the state and DOE who would be willing to discuss the issues. I'd be happy to participate." If you participate along with someone in a similar position at the DoE, I think it could be very productive. It would allow private industry members to learn about your efforts and vice versa.

Cliff Frost: In general, I don't think having a legislative body mandate particular technical solutions to particular problems is a good idea. Actually, I think it is a horrible idea. ;-)

I agree. Note that in my post I suggested this would be a good first move for the state CTO, not the legislature.

Cliff Frost: Again in general, having a state CTO or CIO seems like a pretty good idea, if that person is strong and reasonably insulated from the legislature--it won't work if it's a political patronage position.

I absolutely agree. The CTO should only be responsible for showing progress against coarse grain objectives in a quarterly or yearly report. We want IT experts making IT decisions.

I remain perplexed by the idea of writing a petition calling for a specific technology solution (Google mail and apps) without having first made an effort to find out why these solutions haven't been implemented yet. From the outside of anything it's easy to criticize.

While I agree its easy to criticize from the outside, note that many of the people on this side of the fence are also experienced IT professionals and are familiar with the state's various IT systems by virtue of the fact we service government departments and build systems that integrate with them. Our small company alone services three groups at UH, one at KCC, HTDC, OTTED, the DoE, etc. Many of us have negotiating experience with the companies you enumerated. Their are plenty of credentialed people on both the government and commercial side. As you suggested, we would probably all benefit from getting together to share ideas.

Cliff Frost: I would almost never favor outsourcing an effort to outsource. When one outsources an activity one still has to manage that activity.

Again, I'm not clear on your terminology here, or how it relates to the CTO or SaaS discussion. I'm familiar with outsourcing and the need to manage it. Managing software outsourcing jobs at large organizations was my job for years before starting Ikayzo.

Laurence A. Lee: You're calling for a Cabinet level CTO, with the power and authority to make sweeping changes across the State -- which is an opinion I generally agree with.

I thought you might. :-) There seems to be some consensus on this point in the discussion.

Laurence A. Lee: My biggest pet peeve with Google Docs is the lack of a Real-Time Notification API.

This is a good point. I too would like to see this, but I certainly don't think its a deal breaker. There are a lot of things I would like to see, but then again, I could say this about just about any API. No API is broad enough to make every developer happy. Having worked on standardizing Java APIs for the JCP, I'm painfully familiar with this dilemma :-) In API design, I believe its important to start with a narrow but solid core (i.e. the 20% 80% of people need), and ensure you have a solid implementation before widening it. This is the approach Google is taking.

Laurence A. Lee: Finally, it's certainly convenient to advocate skipping the investigation process and rely on other organizations' research results -- especially when those results work out in your favor. I was somewhat disappointed by that statement, as we are all aware of how quickly the landscape changes.

These are very recent decisions that were carefully considered by large organizations with very similar requirements. NIH (not invented / investigated here) is easily the biggest money waster I've observed in government and commercial organizations. There is absolutely no reason to re-investigate questions that have already been solved by organizations that are willing to share their findings. The first step in this sort of exercise should be to talk to LA, DC and the private schools and companies who have already been through the investigation, planning and implementation phases.
As to "my favor" - Like you, I am a Hawaii tax payer and resident. What works to my favor is efficient government. That is why we participate in groups like the tech caucus and strive to facilitate discourse on this sort of topic in forums such as TechHui.

Comment by Cliff Frost on February 14, 2010 at 10:42am

In general, I think this discussion is going in too many directions to really work well on a BB. Conversation would likely be more effective.

But here are some general thoughts and I hope they help:

--In general, I don't think having a legislative body mandate particular technical solutions to particular problems is a good idea. Actually, I think it is a horrible idea. ;-) This is not politicians' strong point and they tend to know it--getting them deeply involved is a recipe for disaster (in my view).

--Again in general, having a state CTO or CIO seems like a pretty good idea, if that person is strong and reasonably insulated from the legislature--it won't work if it's a political patronage position. Stuffing DoE and UH under it may or may not be good ideas. I spent many years at UC Berkeley and I can tell you several specific ways that the UC could save a lot of money in IT, and I can also tell you a zillion ways the state's IT organization would have caused far more wastage. And don't get me started on what I saw the state of CA do to k-12 there. Be careful when you bring in politicians! Unintended consequences can get ugly.

--I remain perplexed by the idea of writing a petition calling for a specific technology solution (Google mail and apps) without having first made an effort to find out why these solutions haven't been implemented yet. From the outside of anything it's easy to criticize. From first-hand experience negotiating with not just Google, but also Yahoo and Microsoft, I can tell you that the details matter. All the PR happy-talk does not tell the whole story.

--Having the legislature say something like "all state email and apps should be outsourced to Google" would significantly disadvantage those of us who are trying to negotiate contract terms with Google. Google is a huge company and not easy to negotiate with.

--I would almost never favor outsourcing an effort to outsource. When one outsources an activity one still has to manage that activity.

--With respect to protecting intellectual property, how would we know if a legal entity had used legal means to steal it from a University (for example)? Have you ever heard of (or seen) a US National Security Letter? A foreign instrument could easily be a lot more secret, so how would we know when it happens? Furthermore, the Google/China set of incidents reveal how much interest there is in Intellectual Property cyber-theft, so this concern is far from theoretical.

I'm going to bow out of this bb discussion now, but would enjoy some sort of techhui in-person discussion.

Comment by Laurence A. Lee on February 14, 2010 at 8:50am

Daniel Leuck: We are talking about email, calendaring and some aspects of document management. OLTP can stay local until bandwidth improves to the point where the discussion becomes academic. It doesn't need to be all or nothing.

You're calling for a Cabinet level CTO, with the power and authority to make sweeping changes across the State -- which is an opinion I generally agree with.

Cloud Computing and Outsourcing doesn't have to be all or nothing, but we should at least formulate a Roadmap. Only then can we evaluate the underlying offerings and decide whether or not Google Docs (or Amazon S3, GitHub, or anyone else's repository) is up to the task of fulfilling the need.


As much as you advocate outsourcing to Google Apps is a prudent decision to save money, it could equally be countered with a handful of feature-rich commercial products from Microsoft, PeopleSoft/Oracle, or IBM.

Why exclude the tightly-integrated commercial offerings if they can fit the need better, and are readily available? They can just as easily provide "Cloud Computing" benefits housed in a rack at DR Fortress or Superb. After all, when a standardization initiative is driven from the Top down (via CTO directive), any competent solution will likely save money over the status quo.


Google Apps is a decent set of standalone tools, but like many SaaS offerings, the back-end integration APIs are still too immature to be as useful as an in-house offering. I expect that to change "Real Soon Now (TM)", but as an Enterprise Architect, I'm left to my own devices to fill in the gaps.

My biggest pet peeve with Google Docs is the lack of a Real-Time Notification API. Not even a simple mechanism to subscribe to a document change status. I'd LOVE to have the ability to WATCH a Bill being drafted and revised -- if only to be sure some sneaky Senator didn't try to slip in an Earmark while nobody was watching.

I could do that if said document was managed on revision-control systems like GitHub. I cannot yet do that if the document was managed in Google Docs.

What was that about Google's feature-rich API? :o)


Give me a Real-Time Notification API (via Push Subscription), and I'm right there with you singing Google Docs' praises. Until then, I'm stuck wiping Google's butt with wonky cron jobs scraping RSS feeds to drive a "Recent Activity" Dashboard across many "Watched" documents.

Finally, it's certainly convenient to advocate skipping the investigation process and rely on other organizations' research results -- especially when those results work out in your favor. I was somewhat disappointed by that statement, as we are all aware of how quickly the landscape changes. Had the question of Email been asked a decade ago, and the dominant answer probably would've been Yahoo. 5 years earlier, and it might have been Lotus Notes.

Ask again in 2 years, and it could very well be Google Wave.

Comment by Daniel Leuck on February 13, 2010 at 4:46pm

JW Guillaume: I hope this idea leaks out not only to the State of Hawaii DOE but to the students themselves.

So do I. Cliff's response is encouraging and we've had follow up discussions with several other people within the UH system. We are very happy to see engagement from UH and the DoE in general.

Joel M. Leo: There's an interesting (and lengthy) discussion of Yale's switch to google mail/apps over at slashdot.

Mahalo for the link. Ken Berkun mentioned this article on our "nerd hike" today.

Cliff Frost: I don't know much, if anything, about what the state has been doing, or DOE, but I'm intimately familiar with what UH is doing. I work at UH, and have responsibility for email, and we certainly aren't ignoring the issue.

We are happy to have a senior IT person from UH join the conversation. Note that we work with many groups at UH and we know there are good IT people working within the system. I didn't suggest you were ignoring the issue. I was suggesting that with a state our size, we should standardize at the state level, or at least the DoE, and benefit from a cheaper and more efficient unified IT strategy. We work with enough state government customers that I feel comfortable making this assertion.

Cliff Frost: Um, before rushing out to sign petitions, it might be worth asking folks involved what efforts they've made in the outsourcing direction.

Can you clarify what you mean by this? By outsourcing are you referring to the use of SaaS providers for email, calendaring and docs or are you talking about outsourcing the effort to make the transition? I don't see any problem with signing a petition to let our government officials know we want to see them move in this direction and, in general, create a unified, efficient and affordable IT strategy.
The concerns you raised are very reasonable, but again, they can all be rectified with a legal agreement. As you mentioned, LA got Google to make numerous guarantees to address these issues. What I don't think we need is a lengthy and expensive investigation of areas already covered by other, larger and better funded government organizations.

In terms of intellectual property, although I agree that sensitive data should be housed in the US or at least a country with compatible laws, I'm unaware of any incident in which a US school has had its data or intellectual property compromised by a foreign entity as the result of using Google Apps. Are you? If not, its reasonable to ask if there is really significant risk exposure. If Google allowed this, it would be a PR nightmare that would undoubtedly cause them enormous financial damage.

Cliff Frost: I personally think it likely that UH will end up using gmail and google apps (and I support this direction), but the decision is non-trivial and requires a lot of consultation and care.

I'm very happy to hear this. I hope UH will make the jump, with the necessary legal guarantees, and lead the way for the DoE and the rest of our government.
I also think the move will happen, eventually at all levels of our government. The sooner it happens, the more we will save. Because of our budget crises, we could really use the savings, so I'm voting for sooner rather than later.

Cliff Frost: However, if any of you have followed the recent controversy about Google's new "Buzz" product and its introduction, you may understand why reasonable people might not trust Google's commitment to protecting its users' privacy.

This is an interesting discussion, but I don't think its very relevant to this topic. Google Buzz was rolled out over gmail. It had no affect on Google's business or school customers.

Cliff Frost: I'm happy to discuss this gmail/outsource topic at length. I've been a techie in higher ed IT for almost 30 years, and managing IT for around 15.

Thats great to hear. You've already provide a lot of valuable information in your response. I'm really happy to hear UH is looking seriously at Google Apps. We hope to see more of you and your colleagues on TechHui!

Comment by Seth Ladd on February 13, 2010 at 4:17pm

FYI a great read on cloud computing:

http://sloanreview.mit.edu/the-magazine/articles/2005/spring/46313/...

• ‹ Previous
• 1
• 2
• 3
• 4
• Next ›
• Page