I first want to say that I am so in awe with how you all are pioneering with OpenBD. Before I really got into OpenBD, I was a bit disturbed with information about you all viewed the CFML engine initiative where your priorities were concerned. There were some things that I read and there were also a lot of rumors going around that I had not verified.
Either way, with action speaking louder than words, you all have created a believer out of me. It is the only truly free CFML engine that takes scalability as a serious priority (evident by the existing wealth of features). It is the only truly free CFML engine that truly leverages and keeps ahead competitively in creating tags to maximize the use of the different cloud computing services (S3, SQS, SimpleDB, etc.) It is the only CFML engine that runs on Google App Engine. It is the only CFML engine that has an automated Amazon AMI / VMWare builder for easy cloud deployment with just a few clicks. Did I mention I am a fan of this software and that this is the only truly free CFML engine that is 100% free! Did I mention that the free support is great as well? okay, okay, ok. I could go on and on... WELL I was just thinking about something for the last few days and then just ignored it. Yesterday, I was speaking with a buddy of mine and decided to mention it. Why? Well, here are my 4 reasons: 1) This idea manifested, would simply cause any cloud service (who accomplishes it) to catapult to number one instantly. 2) I thought about why OpenBD has propelled to its greatness so quickly and even, IMHO, beyond those who've been around a lot longer in the game. Watching you guys, I can see that there is NOT a dysfunctional committee in place to slow down or inhibit good progress. 3) You all actually respond quickly even from the TOP guys on down - no one there is ABOVE THEE. We are able to get answers directly from the developers of the server. 4) I believe that if any company could -or- would already be IN POSITION to do it, it would be you all! With that said, here goes: I build applications and nowadays most are designed with the cloud in mind from the ground up. The vision / idea is- what if it was as simple to deploy a fully cloud-scalable application as we are used to with a single server ftp upload? What if the developer never had to think about necessarily coding for for a specific cloud platform but just upload code trough ftp and the "ColdFusion as a service" (CFAAS) platform would just scale as the server-load demands? I know that hostmysite is supposedly doing something like that for Adobe ColdFusion. But I would like to know that my app has defenses against ddos and can utilize GAE and/or AWS as needed (maybe not all in version 1.0 necessarily.) I just believe that whoever can bring scalability down to the masses (without complexity or rewrites) will make a killing financially and create incredible die-hard fans and that goes for any server-side platform (not just CF.) Now, I know about stax, that's really incredible what they are doing. However from my limited experience so far, the system is set up to round-robin which is not true clustering / load balancing. Yes, GAE deals with the scalability for you but you have to code much differently for the bigtable data storage (without use of a relational database.) And yes, we've seen that on huge sites (NO - REALLY HIGH TRAFFIC SITES) like twitter and others that the traditional relational database setup just did not work well in scaling. What gave me a little different opinion of the relational database's ability to truly scale was a webinar I saw recently from terracotta & amazon. If my memory serves me right, in summary, they used 3 small amazon servers that could scale to 100k users pounding on it. This was only possible when they enabled ehcache to automatically & periodically pass the data captured in memory to the db. The db load never seemed to move much from idle during this process. On the other end, they had people pounding on the same three servers only disabling ehcache and the db seemed to show 90%+ usage more consistently. It really took a lot of know-how for them to deploy that system and would make the average CF developer nuts. See, there are lots of servers, components, etc. that you can put together on amazon to scale and there are other pieces that can be tied into GAE that brings you closer to doing development as you are already used to (somewhat). There will probably always be things you have to code for clustering-scalability. CFML has been know throughout history for being the easiest / simplest web server language. How about taking it to the next level to where it is also the easiest server-side language on the planet to scale in the cloud without really changing your code. Better yet, maybe there could even be a scanner plugin (or something in your new debugger) that will automatically scan for code that need to be changed in order for it to properly scale in your CFAAS. I am by no means a cloud expert but would it be possible to create a SERVICE like that that can simply scale as needed without the complexities? Sorry for being long-winded. Just excited knowing it's going to happen with "at least" one of the cloud providers in the near future but I would love to see it with my preferred server first. Should we call it OBDAAS? Pronounced... Well, whatever! -- Open BlueDragon Public Mailing List http://www.openbluedragon.org/ http://twitter.com/OpenBlueDragon mailing list - http://groups.google.com/group/openbd?hl=en !! save a network - please trim replies before posting !! To unsubscribe from this group, send email to openbd+unsubscribegooglegroups.com or reply to this email with the words "REMOVE ME" as the subject.
