Tuesday, July 19, 2016

A new S1000D Blog Reader Poll

Up to Now

In my last blog I summarised what the S1000D Blog Readers Poll had recorded. It set me thinking that what had been recorded before was ok but things needed to move on.

When the poll was started most users were working with only one S1000D Issue. We have come to see from our clients that they are increasingly working with two or more Issues, both SGML and XML.

Similarly when the poll was started very few of the smaller clients were working with CSDBs. There were not many of them and those that were available were very expensive and most smaller companies just did not have the need/cash for them nor could they justify the cost.
Things have moved on and now there are several which are configured to work with S1000D. And some have come down in price.

And Now!

To reflect the change in the use of S1000D, and the experiences of the practitioners, I have created a new survey. The original list of S1000D issues is there but it is now multiple-choice. There is the original yes/no answer to Are you using a CSDB? However the question is now there asking which one is being used. And if you are and it is not one of the above then there is a Question to enable you to put in your own flavour.

I hope that all will now regard this as a new poll which reflect more of todays working environment.

Why has it taken so long?

A good question. The simple answer is that the original Poll was constructed using the well known Quibblo facilities and I did honestly try to extend/modify the original poll to meet today's requirement. I have the username and password safe and used them. But there is a Captcha field which did not work which meant that I could not get back into my account.

So I have been forced to create a new poll using SurveyMonkey. Not quite so nice an interface but at least I was able to create an account, create the poll and then put it on the Blog site.

Tuesday, May 31, 2016

The Last Count - the survey

I first started the short survey in 2010 and it is very interesting to see how usage has changed over the years. So I thought that a quick look at it right now might be useful.

April 2016

The survey results as of April 2016 are very nicely shown in graphical form below (generated by the survey site). Apologies for the fact that they spread into the right hand column - one of the limitations of the Blog Layout.

Issues of S1000D


We will look at the progression of taking on board S1000D but it is very clear that there is now a larger take up of Issue 4. I suppose with a number of US Projects using S1000D we should expect that. I do know that some of the Issue 1 Change 6 data is still in current use and being (more or less regularly) updated.

CSDB Usage


When this question was first posed the majority of people were not using a CSDB at all. However now that is changing. Of course this was a very crude yes or know answer. Perhaps we will do something a little more granular for this in the near future.

Where Responses Came From

One of the features of the survey is that we can get an outline view of where people were when they responded to the questions and how many there were. This is interesting because it, in some small way, illustrates the spread of the Specification in documentation.
Quite a spread I think that you will agree. I find the entry from 'translate.googleusercontent.com' intriguing.

S1000D Usage

Of course what the survey does not take into account is 'How many different Issues of S1000D is each person using?'

As products get more mature one question is: are projects keeping the existing documentation in the original Issue or is it all being converted to a later Issue? This is an interesting question of course because some of the earlier Issue of S1000D are SGML only. Where SGML and XML is available for the same Issue converting to XML is very straightforward with off the shelf software (with a bit of knowhow of course). Where the company originating the information is the one making the decisions that is ok.

Where the company wanting to make changes is not the originalting one that is not ok without explicit instructions - even then it is doubtful. Who is going to sign QA document which says that the converted material reflects exactly the source material? A few years ago there was an issue locally where a company decided that they would convert some legacy Data Modules into a more recent Issue. Having been involved on the periphery of the origination of the first version I   pointed out that the company did not have the right to convert, the material did not belong to them particuarly. I am not sure how popular I was at that point but it did raise the issue with them of who had the right to change or update the material contained in the Data Modules.

I hope that the larger companies are a bit better behaaved than my example. That time they were gung ho in their attitude to data ownership.