Plone vs Hippo: The Followup

Earlier this evening, I took part in the CMSConnected webnar, Plone vs. Hippo. I was representing Plone and up against Arje Kahn (@arjecahn) from Hippo.

The format of the event was fifteen short questions put to each of us by CMS Connected hosts Veronica Cooper from CHEK TV and Scott Liewehr from Outsell's Gilbane Services and CM Pros. There were also questions put forth by the webinar guests and those following along the stream on twitter.

Even before the event, the banter had started between Arje and I...

...I knew this was going to be a fun event :)

The debate itself went very well, even the technical gremlins managed to stay away, which considering I was using hotel room internet access, and Scott was also away from his usual site all went well.

The 15 questions went all very well, with both of us taking turns asking each question in turn.

The debate should be online shortly for those that were not able to watch it live, but in the meantime I'd like to expand on a few points raised by both my opposition and the audience.

Edit: The recordings of the show are now online on Youtube broken into 15 minute chunks

One of the questions that threw me a bit was the one about costs. This is always a bit of an open ended question, and Scott asked 'how much would a 3,000 hour project cost with your system'. I've never really thought of projects based upon hours like that, and whilst that is how we quote for work, we start with the functionality. Indeed I was not alone in that thought... Jon Marks... who is well known for saying what is on his mind ;) quipped in with:

I guess my confusion came about due to the fact that "How much is a 3,000 hour project" is actually so simple a question in the Open Source world that I was thrown a bit. With Open Source you a paying for development time, and not license fees, and so the answer is simply 3000 x <hourly rate>. However in the Plone world you can choose from many companies worldwide to do the implementation so the hourly rate will vary greatly. A larger company that can give you things like Prince 2 project management will cost more than a 1-2 man band. Same as a company in Northern Europe will likely charge more per hour than a company in South America.

I can only speak specifics about Netsight, and our standard hourly rate is £85 / hour, so a 3,000 hour project would cost you about quarter of a million pounds. We'd probably give a discount on a project that size, but you get an idea of the order of magnitude.

The problem being that what you haven't addressed in the question is what you can implement in 3,000 hours. That is the point that Jon so eloquently makes. This is both a function of the team doing the implementation and their experience; and a function of the tools being used. A very big generalisation, but Open Source content management platforms and frameworks tend to lend themselves to more rapid, agile development. That isn't strictly a property of the tools, but often more to do with the mentality and approaches of the teams using the tools.

I mistakenly referred to Irina as a 'CMS Consultant' which was a mistake on my behalf, she is a CMS Analyst not a consultant. However again, she asks for the cost of a 'mid sized' project. Now I'm used to vague 'how long is a piece of string' questions from clients, but from an analyst was a surprise. Yes you want to know if you have a budget of £1,000 then you are not going to get OpenText in, and if you have a million pound budget you might be expecting more than what Wordpress could offer you, but still it seems an odd question to me.

Another question that surprised me quite a bit was Arje's question at the end. We had both been told the final question would be a chance for us to ask each other a question. Arje asked me in a nutshell "How is Plone going to cope in the current world of multi-channel delivery when it just deals with pages?". It seems that Arje believed that Plone just deals with the notion of 'Pages'. In fact Plone deals with content 'objects' not pages. Plone uses an object database to store its content in and the object publisher in Plone serves up an object and that object renders its view (normally via a template to produce HTML). Plone comes with a number of content types by default (Folder, Page, Collection, Image, File, News Item, etc) and whilst there is a 'Page' content type, all that happens to be is an object that has a 'body text' field in which people enter content using a visual editor.

You can create your own content types in Plone, and define the fields you want, and the types of those fields. You can also define the templates that are used to render that type of object. You could have a different template to render objects differently depending on the context in which they are accessed. For instance you could have different views for desktop HTML, mobile HTML, XML, json, etc. Content types could whatever you want, such as an 'asset', 'support ticket', 'office', 'person'. Basically, any type of content that you might have in a business. For example, Netsight's own intranet has content types for 'Contact', 'Company', 'Quote', 'Project', etc. If you render a 'Quote' you get a fully formatted PDF file based upon aggregating all of the sub-objects within a quote, such as 'Section', 'Cover Page', 'Costs'.

This separation of content and presentation allows you to easily deliver content to a multitude of channels in specific ways. If you want to actually have different versions of content (e.g. I want to write a different version of content for mobile and desktop) then I can use Plone's multi-lingual capabilities to do this. You set up a 'language' for mobile and for desktop. You can move those around to different areas of the site and have tailored IA to match.

So is this really a problem in Plone, no I don't think so. Is there a perception and/or marketing issue with this aspect of Plone... perhaps. I guess we have been doing this for so long in Plone that we sometimes forget or don't realise that we take this functionality for granted, and not all content management systems work the same way. Indeed there are systems out there that deal specifically in 'pages'. Maybe we need to make a more concerted effort to explain this in Plone marketing material.

My final question to Arje was "how do you think we Open Source CMSs can collaborate in the CMS marketplace to make Open Source an even stronger option for clients in the Enterprise sector". Scott intercepted the question and wanted to remove the 'Open Source' part of the question as thinking that was not so relevant and just ask what each project as  whole can learn from each other.

As I've just spent the past two days at a Government IT Procurement expo in London (SmartGov) and been talking with quite a few people in the IT industry here in the UK there is an issue with the procurement of Open Source solutions in the public sector in the UK. Maybe this is not such an issue in the USA or Netherlands, but here in the UK it is. I would love Open Source to not be a factor at all. Arje said he wishes the day all CMSs were Open Source, and I have expressed that sentiment more than once. In my mind, I just really don't see how a business model based upon keeping something secret can continue.

Earlier today at SmartGov Gerry Gavigan, Chairman of the Open Source Consortium, made a fantastic analogy of Open Source to hairdressing (a refreshing, non-car based IT analogy). His point was that just because you can cut your own hair doesn't mean you necessarily want to, but you have the ability to do so, the knowledge and tools are there. You have visibility of the process and the supplier can't 'hide' costs. You have a number of suppliers available and you can go get a cheap cut at the barbers, or pay more for a cut at a salon with a cup of cappuccino offered to you. It is a mature and well understood market: hairdresser's business is entirely dependant on customer satisfaction and reputation, not that they are the only people who have access to scissors.

I was not really trying to imply an 'us vs them' sentiment with Open Source vs Proprietary with my question, but more the fact that Open Source products have a greater propensity to collaborate as there isn't any 'secret' to be protected.

Anyway, it was a fantastic webinar, and thanks to Scott Lewehr, Gary Eisenstein, Veronica Cooper and everyone else behind the scenes, and of course Arje Cahn, my very worthy 'opponent' in the ring...