In which I discuss some issues
I have been brewing on this one for some while and some recent posts have really got me thinking and ready to write. Before we get started, for full disclosure, for those who don’t know I worked at ESRI from December of 2004 until June of 2005. I was a product engineer on the Java team so I worked on Engine and Server, and I was on the EDN team to help with community building. There are a lot of great people I worked with who care about the product they build and about how their software performs for the user. This discussion is NOT about those people, so I am not saying this about everyone at ESRI. I am going to be talking about ESRI as a company and not the individual people, though sometimes I may call out indiviual people as examples of doing the right thing.
[Update – I was unhappy with some of the stuff I said so I need to work on it some more and I may repost]
[New update – after seeing fantom post a comment I now realize I have to put some of it back. When I cut it I lost most of my links. If you see me on the street let me know I should click save on posts like that so I can read it the next day before I publish. I am conflicted about this post for the reasons I stated above. I really like and respect a good number of the people at ESRI. I just think there are some bad decisions happening and would like to see them change. So while this may read like a rant (and I guess part of it is) it also can be seen as a wish list. I need to get back some of the major links and then I will repost. ]
My thoughts started with the discussion of the new licensing terms for the ADF. When I heard the news I was pissed. I ranted and vented over IM to James and Dave and I talked to my coworkers. I was so upset about it that I had to call James and verbally complain even more. Why was I so upset…
1. This is a slap in the face to all those people who worked so hard on the ADF and put a lot, and I mean A LOT, of cool features and love into that product. This licensing definetly kills some of the motivation to use it. But in addition almost all the points brought up by Dave, James, Brian Flood and others are really valid reasons to not want to host your ADF on the same machine as your ArcGIS server. This new licensing makes it look like ESRI doesn’t understand how enterprise web development works (that is the generous interpretation). I am quite sure there were few if any people on the Server team who thought this was a good idea. They understand all the reasons why asking people to pay yet again to host the client that talks to the server is a bad idea. They know that licensing like this puts a good chunk of thier customers in the uncomfortable position of having to think about site security or relialibility versus cost. What other vendors do this kind of licensing for the ADF that talks to their own Server product? I can see asking people to license the ADF if they don’t buy ArcIMS or ArcGIS Server, but even there I think the argument is weak. Acrobat didn’t become the defacto standard for ebooks by charging for their reader. If ESRI wants to win market share give the frickin ADF away for free! Every time someone opens that code or downloads it they have to think of ESRI. OK, don’t give it away for free but then don’t license the heck out of it.
2. There was not one official word from ESRI anywhere clarifying this new stance on licensing. There was a heated thread going on and we all know that the ESRI folks read blog threads. We never see them officially answer anything there but from my logs and discussions with others, I know they read those kind of threads. Even if they don’t respond on the thread there should be somewhere they can address this issue in a formal manner. There was enough heat there to deserve a little bit of light.
3. In the end this comes down to what I see as a sad state of evolution in the ESRI interactions with customers. There was a feeling when Brian and Art were pushing the blogging and EDN added comments in the API doc that perhaps ESRI was going to start engaging in discussion directly with their users in a more open fashion. I know that the people who get to attend the UC or developer summit can talk to some of the people that matter, but that is twice a year and only a select few who can make the conferences. As numerous people have pointed out, the rise of the internet, blogging, forums, and other community building areas has enabled companies to engage in more direct and open communications with their customers. For example:
- Sun, MS, AutoDesk, and IBM allow their employees to blog. And not just corporate marketing blogs but real open and interesting blogs
- Jetbrains opens their entire bug database to users and they allow users to submit public feature requests and then allows voting on those features
- Infragistics has discussion forums where the developers and support staff actually contribute to the discussion
ESRI was moving in these directions. Of course there were voices of caution and ESRI is, in general, a cautious company, but there was movement.
What is getting me down is that the voices of fear seem to be winning out over the voices of openess. Being an open company is not without it’s risks – I mean you could get Dmitri as one of the people blogging. To make it work you actually have to tell people it is important to engage in that conversation throughout the year and then actually reward them if they chose to do it. There can be great gains that come out of the risk.
And this ties into the thread James, Dave, fantom, and others have going around about Open Source. ESRI charges quite a bit of money for their products. There software has some quality issues. With those two things going for it ESRI needs to start showing some love to the average everyday users. It is not enough that big pocketed or strategic clients get to call Jack and demand new features or bug fixes. Without a fault, every person I talk to in small to mid-size firms complains about having to use ESRI software. When I taught using ArcGIS software my students would greet me in tears because they had left a job running overnight only to be greeted with an application error. I gave a presentation the other day and ArcMap seg faulted. Thank goodness they were all ESRI users and so we all shared a laugh.
We like doing GIS work but we don’t like the way we are being treated. Brian T. says that AGX is to GE as the Zune is to the iPOD. I would go one step farther and say that ESRI is very similar to Microsoft in the way they approach products.
You might be saying what do these two threads have to do with each other. Well here it is – if ESRI was more open with their users I think there would be less animosity and less of a need to look elsewhere. I am sure people would still be looking to open source for various other reasons, but ESRI would take away one of the big reasons I think quite a few people are looking. You see, if you are open and we are in dialogue and discussion then I am going to give back. Rather than it being a relationship of producer – consumer it becomes one closer to a partnership. And this is one of the big wins for open source and it is definetly one of the reasons I am looking at MapGuide over ArcIMS. MapGuide is a community now and we are partners. There is an IRC channel, there are user groups where the developers participate, there is an open bug database, there is a wiki for documentation, and I can suggest a feature and have people vote on it.
You don’t have to open source ESRI software to leverage some of the ways they have learned to have a better relationship with their users.
Let me try to wrap up since this is long enough already. I know there are some customers who are happy with ESRI and most of them have big pockets or are strategically important. And that might be a viable strategy for ESRI but it doesn’t seem to be making the people I meet happy. In the best of all worlds you have a rock solid product with sexy appeal and you make users feel like they are part of the team and that you care about them. In the next step down you better get at least one part of that equation working. I bet if you do the latter you will get to the former.
One other observation, to all of ESRI’s competitors and to Open Source projects trying to gain traction – if ESRI pitched them to you any slower it would have to be T-Ball.
In the end I am hoping ESRI proves me wrong and things go swimmingly. Like I said in the beginning I like and respect my former colleagues at ESRI. I know they build good products and they want to do the right thing ™.
This time I promise to keep the post up…