Saturday, January 16, 2010

The Ethics of Blogging about Bugs.

As a certified IBM/Lotus Business Partner I believe that I have a duty to report bugs to IBM in order to help them fix the product. Generally I'm quite happy with the speed of the response I get from IBM/Lotus even though I don't always get the fix for which I was hoping. Whether or not the bug gets fixed, the event normally provides some Blog Fuel for slow-news days.

Late last year I had a response from an IBMer to a bug post on my blog which said, in the friendliest possible manner, that it was better to report bugs directly to IBM than to just blog about them. Normally I do both - but in this case I had just blogged about the issue and hadn't raised a bug report. Leaving that particular bug aside for the moment, I want to consider whether blogging about bugs is a positive contribution to the Yellowverse. On the Down side ...
  • the risk of triggering a beat-up by Microsoft PR - "Lotus bloggers snub IBM software!!!"
  • uh... [open to suggestions here]
On the Up side we see:
  • the possibility of getting a workaround from some otherwise unknown Yellow Bleeder who has a better understanding of the problem than I do.
  • the possibility of getting a permanent bug fix from some otherwise unknown Yellow Bleeder who has a better understanding of the problem than IBM does.
  • a public warning to other Users that 'Here Be Dragons' and to tread carefully.
  • gratuitous gratification at holding up IBM's dirty underwear? Maybe ;)
  • a Great Reversal of the traditional vendor lock on information where the supplier can continue to deny the occurrence of the bug and there is absolutely *Nothing* that the end user can do to get more information.
  • a new ability for the individual to publicly prod IBM with a pointed stick and keep Big Blue focused on fixing bugs quickly, or as Gavin said in his reply - "Somehow, blogging seems to make IBM just that little more responsive."
Allen was right in his reply to my blog last year - I should have reported the bug and I didn't (because I considered the bug to be trivial and there was an obvious workaround). Leaving aside my failure in that particular incident I'm interested in other people's opinions regarding non-emotive blogging of software bugs. Is it a good thing to do?


Peter Presnell said...

I am all for sharing bug in open. If Lotus has a strong brand it should have nothing to fear from MS about making its bug public. I have clients that provide me with no direct access to IBM so reporting a bug through "official" channels is often more trouble than its worth. I only wish IBM would be more open about all the bugs it already knows about so I wouldn't waste so much of my own time trying to diagnose.

Tim Lorge said...

I think the ethics issue here isn't really yours but more the IBMer who asked you to keep it dark. Unless you are under an NDA for beta software or there is something in the partner agreement, you have no legal, moral or ethical obligation to do that.

In fact, ethos actually demands that you reveal it. Now, one could debate form and the volume of vitriol but, that is something else.

It is one thing for an IBMer to keep something quiet. They have an obligation to do that because of their relationship with IBM. There are internal documents that must remain internal for a variety of reasons. Which is fine.

That being the case, it is your blog! You write what you damn well please! IBM doesn't have to like it. We are their customer. We pay their salaries. IBM has an obligation to be forthright and open about issues. They have to respond to us.

Should you have opened a PMR?

Maybe. There is no guarantee that support would see the issue as you do or, quite possibly, they leave you hanging for months on end. As Peter mentioned, what if you have no support with IBM? How then do you report it?

Now, about IBM asking you to keep it quiet? THAT I have a real issue with.

We as customers have no obligation whatsoever to keep dark knowledge of problems that IBM, or any other vendor for that matter, may have. EVERY bit of software running has bugs in it somewhere. Whether those bugs manifest themselves in one's environment is another story.

To be fair, I've not had this problem with IBM.

In the end, it is ethical for us to use whatever means necessary to resolve our issues. Some may not like the blogosphere but that is their issue. Hmmm .... perhaps they should blog about that?

Graham Dodge said...

Let me reiterate:
IBM did NOT ask me to keep quiet about the bug - they merely asked that I should report it to them through official channels so they could fix it. They were absolutely right in asking me to do that, and this blog post is just exploring the general theme of blogging about bugs rather than pointing a finger at IBM.

BTW I should also have clarified that I see blogging about security bugs as a no-no until you have given the vendor all reasonable opportunities to fix the problem.

Mika Heinonen said...

You can't report many bugs to IBM, since they ask you for a reproducable scenario. That is in many cases impossible, when you have a complex system.

The simple solution for customers is to say to IBM: Just examine your source code, and fix the bugs.

In practice, that is also a unrealistic request.

What remains, is that customer needs to stay on Domino 7.0.3FP1 and hope that someday, after posting enough NSD's to IBM, any newer version of Domino would become as stable.

Bill Hanson said...

This post could not be more timely for me.

I finally got around to upgrading to 8.5.1, and now I am unable to see the field list for a Domino Document data source.

At first, I thought it had something to do with my environment, so I setup a new laptop and installed 8.5.1 in a clean, stand-alone environment (no Internet, no domino server, no id).

The bug is the same in all of my tests (3 different machines & 3 different ids now). The bug is reproducible every time!

My questions are...

1) If I can reproduce this every time, why can't I find anything about it on the web? There must be others out there experiencing the same problem, no?

2) Since I don't have IBM support, how do I let IBM know about this bug?

Graham Dodge said...

I'm just back from two weeks holiday so my apologies for the late reply.

I can't reproduce your problem here, but I believe that it is happening there for you.

I suggest that you contact a local IBM/Lotus Business Partner in Florida and look to set up a long-term relationship with them. That DOESN'T mean that you pay them money every time you pick up the phone but it does give you a conduit into IBM.

Most Business Partners that I know are happy to establish and maintain that kind of relationship as long as you don't abuse the friendship and you pay for your annual Lotus license subscription through them.


Bill Hanson said...

Hi Graham,
Thanks for responding. That's probably what I'll do.

On a side note, I have finally found others that have encountered this bug. It's being discussed on several forums, now so it should make a future FP.

In case you're wondering, here's how you can reproduce the indication on your own machine:

1) Create a simple xpage for testing this bug.

2) Add a document data source that points to the Group form in your address book.

3) You should not see any fields in the data palette. This is the bug.

4) To fix it, open the Group form in designer, make a trivial change and save the form.

5) Switch back to your test xpage, and presto! The fields are there!

A few observations regarding the bug:

The bug only happens when the first line of a source form is hidden.

It only happens for legacy forms. If you create a form using 8.5.1 and hide the first line, no bug.

Sometimes, all you have to do is save the source form once in 8.5.1 to get rid of the bug, but not always.