Wow--last week's blog entry on the future of COMMON certainly got a lot of responses. This week we're hoping to continue that conversation because we think it is an important one --and we don't mean the volunteer remuneration part--we mean the whole picture.
We'd like to start this part of the discussion by asking a very simple question: What does the IBM i Community need?
We think it might be better to take COMMON--at least as we know it today--out of the picture initially. Why? Because we find that many people seem to have a romanticized view of COMMON from "the good old days" and assert its need to exist simply because it always has. Surely the criteria by which we think about COMMON's future path should be whether (and how well) it fulfills the needs of the community. Hence our decision to start with this question.
Before we ask for your input on that question, we should perhaps identify what we view as the three branches of the community:
- The first branch is IBM--the maker and supporter of the system and software platform on which we run our applications.
- The second is Vendors--Vendors in this context includes ISVs, education suppliers, hardware vendors, consultants, media, etc. One could make a strong argument that this branch of the community has more to lose than any of the others on this list should the IBM i and/or a clearly identifiable community of IBM i users fade into the sunset.
- Last, but by no means least, come the users of the system--those of you who run your businesses with the help of the applications that run on i.
As soon as we split it up like this it became obvious to us that the needs of the three branches are really quite different. The only thing they really have in common (no pun intended!) is a love of the platform and a desire to see it continue into the future.
So putting COMMON aside for the moment, what do each of these branches of the IBM i Community need?
Let us know what you think.
We need a solid direction that shows how the POWER6 boxes will compete against the Microsoft of today and its various servers and applications moving forward. If we don't take the bull by the horns hopefully the system doesn't end up like GM's Pontiac Division. This roadmap should show how the IBM i apps stack up against those of the competitors, TCO and application $$ comparison savings might not hurt. Example: What is the cost savings of running MYSQL under IBM i 'VS' Microsoft SQL on other hardware.
Vendors need to show how there products running on I stack up against the others. Savings, TCO, etc. How about products that show hardware and software conversions from Microsoft to System i, always seem to read it the other way around. Product comparisons, what are the comparable products to sharepoint and other microsoft products.
Native GUI, let's face it we have had allot of alternatives to do all kinds of GUI things, but a lot of the time the learning curve and the cost to implement always seem to win out and then it gets shelved or mothballed. It might be save to say that to convert away from PDM to WDSC might have come easiear if PMD were to be eliminated all together. I don't want to order machines with twinax and modem cards etc. The cost analysis of keeping it the same old way 'VS' going the new and improved way never seems to quite make it across the directors or managers desk. If I can using XML or FTP to replace dial up I should be exploring it.
Posted by: Jack Kingsley | May 12, 2009 at 12:56 PM
More than anything, we need to bring in new customers to the platform. It's perfectly normal for customers to switch from one product to another in any industry, and people leaving the i for another platform is perfectly normal. The problem is, we're not bringing in enough new customers to offset what we're losing, and so the customer base is eroding.
Virtually every vendor I've talked to in the IBM i space is either supporting multiple platforms or is giving serious thought to it. By multiple platforms, I mean supporting Windows, Linux or AIX in addition to IBM i. They are doing this because there isn't a large enough customer base in the IBM i space to generate enough revenue. The vendors desperately need to keep bringing in more customers.
So what is really missing? Why is this happening? Marketing! Vendors need to tell customers why running their product on i is advantageous. Is it better because it's running on i? If so, WHY?
IBM also needs to market the system. They need to tell people what it's advantages are over AIX and Linux. And not just to people who specifically ask the question, but to people who have no idea.
Most people don't even know this platform exists, or if they do, they know of it as AS/400, something people used to use years ago. Remember those? The name changes have actually had the opposite effect from what was intended, instead of making people see it as a new more vibrant product, they've obscured the system to the point where people don't even know it still exists.
Last night, I was at a friend's house, and another friend, who I've known for 17 years was talking about computer programming. He works in IT, but not on the IBM i platform. Someone (not me) mentioned RPG programming, and he said "ewww.. that's positively evil". He remembered RPG from school, and remembered how awkward it was and how hard it was to do things. To him, RPG was RPG/400. He had no clue there was anything else.
Marketing. Someone needs to get the word out, not just to the existing faithful subscribers, but to everyone else. That is most desperately what the i community needs!
Posted by: Scott Klement | May 12, 2009 at 01:26 PM
Recall the words given to Dorothy at the end of her journey to the strange land of Oz? There's no place like home, there's no place like home, there's no place like home...
Well, notwithstanding the new runtime environments deployed under IBM i (Java, J2EE Application Servers, PASE, and PHP), there's no place like native languages and the native runtime environment.
It kind of shocks me to see the amount of content at EGL Cafe dedicated to the migration of IBM i applications off the platform. One IBM blog in particular advises that IBM i hardware may be too expensive, but that EGL applications can be deployed under Intel.
Rather than reaching out to .Net developers and possibly attracting new customers to the platform, IBM is reaching out to RPG developers with EGL, Java, and PHP, and ultimately the migration of applications off the platform.
With that type of leadership from IBM, is it any wonder that so many other vendors are promoting and offering cross-platform solutions - or in other words, inviting us to gather in the strange land of Oz?
Is it any wonder that so many other vendors have built businesses on the attraction of migration off the platform. Some of their migration messages are so subtle, that people perceive them as platform supporters - and their success ensures advertising revenue for traditional IBM i publications.
IBM's server consolidation strategy isn't that effective - as more organizations are using VMware to partition Intel servers, and managing mixed workloads thereby.
IBM's server homogenization strategy has taken a toll on IBM i too, as so many of its distinctive qualities have been watered down. IBM's most prominent message is that their road-map is cross-platform, that their software will run on any platform. Prospective customers get mixed messages about IBM i being both distinct and the same as as other platforms.
Like Dorothy, prospective customers as well as IBM i faithful need to hear that there's no place like home, there's no place like home, there's no place like home...
But who will share that message? It seems to me that IBM's rudder is set and not likely to change. It seems to me that the message must come from minority stakeholders - IT departments, end-users and small ISVs, if anyone at all.
Actions speak louder than words. My objective is to continue developing new software, written in RPG, interfacing with browsers, and attracting new customers to the platform, by focusing on the core architectural principles that led to the platform's rich legacy. I've been to Oz, for a visit. But there's no place like home.
Posted by: Nathan Andelin | May 12, 2009 at 04:11 PM
"If I can using XML or FTP to replace dial up I should be exploring it."
Not to embarrass you, but I wonder if that is one of the problems with the platform - that technical people on the platform don't know left from right when confronted with the "next generation" of technologies.
Note that FTP is a network protocol used to exchange files over TCP. XML is a syntax used to delimit data. You might send an XML file over FTP. "Dial-up" is a form of internet access using a modem that is connected to an ISP (internet service provider). You can use FTP to send an XML file over dial-up.
---What the community needs---
In my opinion what we need are solid apps to run on the platform (i.e. business apps like inventory, order entry, shipping, etc). Apps are solid because they make use of the different ground floor features like logging and job control (i.e. Java/PHP don't really give us this because they run on top of another "virtual machine", but RPG/CL/C/COBOL most definitely do). If the application building tools that IBM gives us don't make integrated us of the operating system then we start to lose a lot of the benefit of running on the IBM i.
If IBM would give us the capability to create solid graphical applications using native languages I think that would go a long way. Note that the infrastructure for such a "application framework" would have to be integrated and productive without requiring developers/programmers to become techno weenies to produce business applications (as is the case with most GUI technologies today on the IBM i). It never ceases to amaze me at all of the things that we DON'T need to be concerned about that other developers do. For example, we never have to worry about our connection to the DB from RPG.
People on other platforms are used to putting up with a lot of un-productive practices and I think if you eliminated those people would start jumping ship to the IBM i from other platforms (along the lines of what Klement is saying).
Aaron Bartell
http://mowyourlawn.com
Posted by: Aaron Bartell | May 13, 2009 at 07:45 AM
Not much more than a 'me too' post here. I agree with Nathan and Aaron, why is everyone seemingly pushing/pulling us from what makes the i what it is?
I'm a businessman first and a computer guy second. I want to spend my days doing things for the business. I don't want to be constantly monitoring and tweaking electronic equipment. I want to USE that equipment for business purposes. The i makes that possible.
Posted by: Jeff Crosby | May 14, 2009 at 06:56 AM
Interesting thoughts Nathan (Andelin). I hold the same belief, because in the end many of the newer frameworks and platforms being adopted will eventually lose their luster once people have used them for 10yrs. Will some succeed? You bet. But the reality will be that what we have with the IBM i is much better than what others have (minus the UI - but that is being worked on).
I was talking with Elvis Budimlic (Centerfield Technology) and he made the comment that some day we might see Microsoft touting an integrated/seamless/native database access methodology, and wouldn't that just beat all considering we have had that for years yet it isn't marketed as an AWESOME feature. Just think of all the JDBC/ODBC/connection-pooling/etc problems we DON'T have because everything is integrated.
There's no place like home...
Posted by: Aaron Bartell | May 14, 2009 at 11:27 AM
I think to a lot of young programmers the idea of an OS made up of objects with an integrated database is a novel and welcome idea – some might even say ‘new and innovative.’ So I’m going to echo Scott’s comments above and say that instead of trying to bring RPGers into PHP, Java and EGL worlds, which leads us off the platform, IBM should be evangelizing the strengths of the platform among the PHP and Java folks, to get them on the platform.
The problem is, if I am a Java or PHP programmer and I (somehow) hear about this system from IBM, there is no way for me to kick the tires, explore and see what it is all about. The only way for me to develop on the I is if I get a job with a company that has one, and that company is probably populated with guys who don’t know PHP from XML from AJAX.
Maybe IBM should take one of the IBM i machines in Rochester and open it up for anybody to sign up for a free account. Maybe you get a certain amount of storage, one or two libraries, access to the compilers, limited access to a web server, etc. They could also have a web site with tutorials and sample applications. Maybe they would have an intro to i5/OS, intro to RPGLE, intro to Java development on the i, intro to PHP development on the i, etc. This would be a great resource for anyone interested to get their feet wet without having to get a job with a company that has one. And *gasp* do you think it might increase the number of applications available on the platform if an independent developer could at least start developing something without having to buy or rent time on an i?
The message could be ‘Look, we have this system that is more secure and stable than anything out there. It runs Java and PHP better than anything out there. It has a database integrated into the OS. You don’t need a DBA. Oh, and by the way, we have this language called RPGLE that is the easiest language for implementing server side business logic, bar none - check it out. If you are a Java programmer, you will really appreciate how easy it is to perform sophisticated file driven business logic.’
We need to get to the point where solutions providers (web development, application development firms, consultants) are saying to their clients “If you want the least amount of headaches, you want to run this on an IBM i, not SQL Server or Linux. It’s just a lot easier.” One way to get there would be to get it in the hands of anybody who wants to try it out and let the world ‘rediscover’ the platform.
Posted by: Chris Vollstadt | May 14, 2009 at 11:47 AM
Chris, the YiPs (www.YoungiProfessionals.com) are looking to have something similar to what you are suggesting (i.e. have a machine available for people to try out new things and do open source development on). Note we are in the very early stages of what we would like to do and we don't even know how feasible it will be, but basically we learned of a way you can split up an LPAR into even more logical "environments" to keep developers on the same LPAR from stepping on each other. We haven't implemented the idea on the machine we have on the internet because the core YiPs have been busy this spring but now that some schedules are opening back up we hope to pursue this again.
I couldn't agree more that this is a MAJOR hurdle to new never-seen-an-IBM-i-before adoption. As I also understand, it sounds like the Virtual Loaner Program is also looking to move forward with an initiative to help with the situation you describe. I don't have a lot of input into the VLP process, but I can say that it is probably one of the coolest services I have ever made use of from IBM. Basically it allows me to setup a new IBM i LPAR in about 10 minutes via the browser, and then two hours later I get an email saying the LPAR is ready to be used. Such a cool program that could really put a dent in the issues of people not being able to try new things out. If you want to learn more about the VLP you can go here: https://www-304.ibm.com/jct09002c/partnerworld/wps/pub/systems/vlp/index
Aaron Bartell
http://mowyourlawn.com
Posted by: Aaron Bartell | May 14, 2009 at 03:33 PM
Having recently returned to a System i developer role, here's what on my wish list:
A low cost IDE. Very excited to be working with Eclipse-based IDEs (WDSC v7, ZendStudio for Eclipse, i Edition). Not surprised to learn that IBM has once again decided to tinker with developer licensing at a premium price.
A current Subversion for i- even if it's an AIX based version running in PASE. Softlanding doesn't seem interested in porting anything past v1.4. (And maybe I'd take a crack at it if comprehensive PASE documentation was publicly available). Failing that, I'd settle for a Rational Team Concert for i Express-C version (if a 3-seat teaser product works for linux/windows, why not for very limited System i pilot projects, too?).
Very pleased about the improvements in API documentation. Hoping to see newer editions of some RedBook classics (Didn't Know you could do that with RPG, Porting UNIX Applications Using AS/400 PASE)
Posted by: Jack Callahan | May 14, 2009 at 09:52 PM
>A low cost IDE. Very excited to be working with Eclipse-based IDEs (WDSC v7, ZendStudio for Eclipse, i Edition).
Definitely not a feature rich replacement of WDSC or RDi, but have you checked out RPGNextGen (www.rpgnextgen.com)? I use Linux on my desktop and make use of this almost half the time as of late (with the other half still in WDSC).
>A current Subversion for i- even if it's an AIX based version running in PASE.
I would also like to see this. Based on my minimal past experience of porting C based tools to the IBM i, I believe all we would need to do is determine what needs to change in the build script so it works on the IBM i vs. Windows/Linux. If you are truly running it in PASE I would imagine much less changes would need to occur. I ported a Subversion client to the IBM i last year (http://mowyourlawn.com/SystemiSVN.html) - but obviously you are wanting the server portion.
I am also actively looking into Rational Team Concert for i. I wish they would just give each shop one free license to it so I can have plenty of time to test it out without worrying about the 60-day trial. Not only that but it then can become the defacto standard for source management (and whatever else it does) for the community, and guess what everyone will prefer when they go to the next shop? They will prefer the tool they have worked with the most.
>Very pleased about the improvements in API documentation.
Curious to know what you mean by this? Maybe I missed some improvements? I would LOVE to see IBM take a similar approach to what PHP does: http://us.php.net/manual/en/function.checkdate.php. Note they allow community involvement at the bottom of each page for "best practices" or "watch out for..." type stuff.
Aaron Bartell
http://mowyourlawn.com
Posted by: Aaron Bartell | May 15, 2009 at 10:59 AM
What the '400 camp needs is customers, and lots of em. But there is only one way to get them, and thats to have the apps that customers want to buy. Why? Because companies do not buy systems, they buy apps.
If I run a warehouse as part of my business, I need a warehouse app. If I already have a system, whether thats windows, unix, linux or i, its likely that I will look for something that will run on a system I already have, but mainly I will be looking to get the most functionality for the least cost. If I have an I and the best affordable app runs on windows, I buy the app and get the windows skills in.
But to get the apps on the system we need the 2 things. We need the vendors to step up and make as400 apps as good in all aspects as those available on other systems. We know that customers do not want to look at green screen apps these days. There is absolutely no reason for a vendor to be selling a green screen app today, with all the modernisations options available.
But we as developers need to raise our game too. There are still a lot of developers that believe that RPG and CL are the only things they need to know. If you arent encouraging your boss to modernise apps to bring them up to whats available elsewhere and increase functionality and usability, its likely you will be out of a job when that slick salesman in the shiny suit sells him a whizzbang product with all the bells and whistles (less functionality but a nice front end!) but runs on a windows box.
I also think we all need to get out of this free mindset. Lots of developers on system I have got so used to getting everything and the kitchen sink as development tools for nothing from IBM, that they dismiss any tool that has a cost associated with it. Program generators and lots of other interesting tools have been available for years at low cost on I, but many dismiss them and stick to coding things by hand or not at all (still updating that config file using DFU when you should have written a WW screen for it!).
What im getting at here is if we dont spend money on tools and software for the I, there is no encouragement for the vendors to create them. We cant get something for nothing, and at the end of the day, we are not hobby programmers learning new skills. We are veteran developers working for large organisations. There is nothing wrong with encouraging your boss to pay a few thousand dollars for a tool if that tool is going to to give you bang for the buck for your organisation in terms of more productivity or more functionality. We need to get away from this cheap / free mindset in the as400 world, as it is not doing us any favours. If vendors arent making any money they will soon move to a platform where they are, which I suspect is happening slowly but surely.
So developers and vendors need to raise their game. I think IBM has done lots to make the i equal or better of any system out there, and we need to start using the tools available.
Just my 2p
Posted by: as400 guy | May 19, 2009 at 03:13 AM
Aaron,
I can't wait to see what the yips come up with in terms of a box on which one can do OSS development. It would be great if you include CGIDEV2 and PHP environments as well. I wonder how many more yips we could create if any college student could just sign on and start coding?
Jack,
You might want to check out the RPGNextGen editor here: http://www.rpgnextgen.com/
Apparently, it is a free RPGLE plug-in to the open source Eclipse distrib. I haven't tried it yet, but I'm going to this week (on my Mac!).
Chris Vollstadt
Posted by: Chris Vollstadt | May 19, 2009 at 09:29 AM