Sunday, September 23, 2007

Alpha!

Code complete


shapeimage_1-3.jpg

Ok, at 11.45pm on Sunday, September 23 2007 I can finally declare Alpha. I’m using the old Oracle notion of basically the code is written there are just one or two bugs to get out of the system. But as you can see, we have a real FlexClone savings report based on real data captures from a real Oracle database and uploaded and processed through the GUI. Shocking I know. What do I know about code?

Unfortunately there are some tasks still left to do, but in balance we are looking good! Time to get a good nights sleep!

Oustanding Tasks
Outstanding Tasks
Completed Tasks
Complete Tasks

Wednesday, September 12, 2007

Red Rock cafe is the place its at

Product demo's in Mountain View


shapeimage_1-2.jpgSo I met an old friend in Mountain View for dinner and a coffee. I have know Roger B for many years back from my Oracle days. He’s been V.P of Engineering or higher roles at various organizations (WebVan, Apple etc.). He’s doing something related to managing AdSense words at the moment for a startup in Mountain View at present (forgot who).

So after a very nice Thai dinner (thanks Roger) I gave him a demo in Red Rock coffee on Castro Street. So, he had little idea of what the product was about or what I had been doing, but after the initial “very nice ajax GUI” and I nights sleep I got the following back :

Value: Yes, but right now it's still a feature (with good ip). A question to ask would be that now that you can do this, what are you going to do with the information? If it is calling the snap api, I think that needs to become more front and center of the pitch.


Model: I think it makes sense. I think it's important to have some way of keeping track of where the db files go, so that you don't reindex files unnesserily and also that you can surprise a dba by having returns show up immediately if you've seen the file before. This could be another feature onto itself... Finally, not sure if you can scan files incrementally (i.e. only the delta). This could be done by caching the inodes of the blocks, and only scanning the new blocks. This assumes that blocks don't move on the filesystem of course.


Show: I think Michael would be a prime candidate...


Launch: Get a marketig guy to develop a pitch :-) Really, I think there are several ways to position this, and I think you should spar with someone and explore at least let's say 3 of them..


HTH,
-Roger


So this made me feel great. If what I had was complete BS then Roger would have called me on it. Funny enough, the guy at the next table listened in and gave some useful suggestions on “nio” and scanning large files. So kudos to Jon Frisby a.k.a MrJoy from www.mrjoy.com for big ears! In case you want to “hang out” too, here’s the map for Red Rock Cafe.


View Larger Map

Amazon Web Service Startup Shindig

Amazon EC2 and S3


shapeimage_1-1.jpg
So I have been tracking Amazon Web Services for some time. So, being signed up I got an invite to a event at Stamford University (those who know me, I know, don’t get me started). It got me thinking about the whole decision about virtualization. It makes a great deal of sense for my application since it allows me to add compute resources as I needed them. However, I now have to manage (and learn) about a whole bunch of things I really don’t want to know about, from DNS to Cicso IOS... you get the picture.

So I thought wow, EC2 plus S3 makes a great deal of sense. Allows me to focus on my application and Amazon to manage the infra-structure for me. The cost is not that bad either, I reckon on about $20 / day to start with which is a lot less than building it myself. It also means that I get some location transparency and can manage on the road, which seem a big thing at the moment.

It just all means that the applications needs a little re-architecture!

Saturday, September 1, 2007

Launch day decided

D-Day is Oracle OPen World


shapeimage_1.jpg
So, there we have it a launch day is decided. Initially when the first lines of code were written on Memorial day weekend in 2007 the first date I came up with was 1st August 2007. Well writing today I guess I have missed that mark by some way. In reality this first date was not motivated by anything else than picking a date.

Given the work at hand is targeted initially at Oracle, it started to make a great deal of sense to launch in time for OOW. There will be 10,000 people attending and its a prime time for some gorilla marketing. Hey, I’m a bootstrap so I don’t have a penny to spend on this sort of stuff so lets make it as cheap as possible.

So, just 10 weeks left to get everything done. The task completed list has not grown :-(