Re: Something that seems to have been missed in the discussion..

On 24 Apr 2007, at 19:35, Peter Houppermans wrote:

...

... I would just make the (IMHO entirely logical) observation that the original code thus stands a good chance of harbouring significantly more problems, they just haven't been discovered yet. Let's call it the iceberg effect.

This leads to an interesting choice, either:

- trust your accounts and their integrity to a team that evidently works hard to make it work as it should, with pretty defined goals and with evidence that what they have set themselves out to do is indeed happening (and you have the option of going the 1.1 or 1.2 route - I'll go 1.2 as I start afresh anyway) or
- use the original code (i.e. SL) and keep wondering what other problems you'll never hear about. Personally, that's not quite the sort of beta test I'd like to be exposed to.

Well said.

Admittedly the schedule of my migration to L-SMB was advanced by unrelated issues, but this is fundamentally why that migration was on the table.

When I realised the implications of the original SL security vulnerability, that (someone as ill-educated as) I could understand the risk inherent in the bug AND yet the developer obviously did not ... well, at that point SL became no longer a serious contender for my use.

I guess I would rather suffer bugs from developers committed to fixing them than just sweep it all under the table, as is obviously occurring in SL installations. If your data is stored in a schema that is fundamentally broken, then what hope do you have?

I should add that I have had very excellent one-to-one, hands-on and real-time support from Chris Travers on IRC. He was able to answer my questions at each step of the migration process and as problems arose - when we had finished he (initially) rejected my offers to pay for his time. I find this to be in stark contrast to the SL mailing list where I would often receive no reply to a question that seemed surely very easy to answer.

Stroller.