Nae neeps and tatties?

Just get rid of the stupid locked door. The passengers have known since sept-11 that they must deal with hijackers from the cabin. The locked door just makes it harder for anyone (crew or pax) to assist in dealing with threats from within the cockpit.

How many suicidal hijackings have there been? 3 - and the passengers had already realised on the day and changed their behaviour as #3 was in progress. The best defence against these hijacks was already in place by September 12th - the passengers.

How many suicidal pilot crashes have there been of jet airliners? At least 4.This one, Egypt Air 990, Mozambique Airlines TM470, Silk Air Flight 185, since 1999.

There's just no good defence against a suicidal pilot. Bear in mind that Egypt Air 990 went from level flight at FLA330 to the ocean in 43s, even thought the captain was back in the cockpit within 12s (no locked door then) and behind the controls no later than 27s.

Also bear in mind cockpits have at least one crash axe (for crashes and getting at electrical fires behind panels).

IMO the best thing to do would be to get rid of the locked doors. Whatever chance the other pilot has, it would surely be increased if other crew and/or passengers could easily and quickly get in to assist with any man-handling needed.

The cockpit already has at least one axe. They may even be more, e.g. a hand-axe, if not two (one by each pilot's seat), and a large axe stowed behind.

That's great. The cockpit is required to have at least one axe btw.

Gender is irrelevant. Egypt Air 990 crashed *without* locked cockpit doors. The captain was back in the cockpit within 12s of the co-pilot initiating a descent. He was making control inputs within 27s. However, he didn't start to suspect the cause of the problem might be the co-pilot until between 30s to 33s. The aircraft hit the sea at about 43s.

Every second may be vitally precious in these situations. Locked cockpit doors, even with over-rides, will waste potentially extremely-critical time.

Axes are mandatory in the cockpit. Nothing to do with protection, but for hacking your way through debris or bent doors after a survivable crash (AFAIK). More axes may be stowed in other parts of the aircraft for the cabin crew to use.

Cockpits have mandatory fire axes.

Except they didn't re-implement the sub-system framework and data-structure APIs themselves. The lawsuit alleges that they took some code from Linux (e.g. radix tree, timer API stuff). Even if VMWare /had/ reimplemented those APIs from scratch, then there is still the issue that they /also/ have appropriated the code GPL-only drivers (as alleged by the lawsuit at least) for ESXi.

Also, will you provide that list of drivers? If you won't, I have to wonder if you're favourably predisposed to VMWare in some way.

Good point. I should have used "had" instead of "have": "they had the choice of not using the Linux code they don't own".

Ta!

BTW, a vendor that wrote a Linux driver could give a different, non-GPL licence to that driver code, of course. However, that still leaves the issue that Linux drivers are written within a framework of core Linux code (driver sub-system specific frameworks and further more generic services and data-structures). The driver vendor can not give a non-GPL licence for that core code they didn't write.

VMWare are alleged to have copied such core code too. Further, they are alleged to also have used GPL driver code (e.g. Hellwig's SCSI). So VMWare, according to the allegations, have borrowed GPL code on /both/ sides of the "line" between drivers and their supporting code. Without fulfilling the conditions required by the GPL for legal, licensed use of GPL code...

The Software Conservancy FAQ has a diagramme giving an abstract of what they allege has been copied:

http://sfconservancy.org/linux...

With respect to the drivers, it seems they've copied SCSI, USB and network drivers. Christoph Hellwig holding copyright to at least some of the SCSI drivers concerned, in addition to core code VMWare are alleged to have copied to implement required APIs for the drivers.

It seems you could give us a full list from an ESXi installation, if you wished, rather than just a selected driver (selected why?).

Uh, no, not at all.

The GPL says "You're free to go build on what's mine, but that makes yours ours".

If you don't like that aspect of the GPL, then you're perfectly free to not use that code and go build your own instead.

My understanding is that your description is inaccurate.

Yes, they have implemented a number of Linux APIs in their own code. Additionally, they have sucked in bits of GPL Linux code that implemented bits of those APIs (i.e. NOT reimplemented, as WINE does). This is to allow ESXi to be able to re-use drivers from Linux, as you say. However, they didn't stop there, from what I understand. They also have ESXi use Linux GPL drivers.

My understanding, from what I've read, is that ESXi didn't just re-implement Linux APIs. ESXi also heavily depends on GPL licensed Linux code, both in the partial-reimplementation of Linux APIs, and in sucking in Linux GPL drivers. The issue is this direct re-use of GPL code, and ESXi's heavy dependence on that GPL code. That dependence likely makes ESXi a derived work of the Linux GPL code, and as such it - in its *entirety* - must be distributed in accordance with the GPL.

Alternatively, VMWare are quite free to not use code they didn't write and don't own, if they don't like the licence conditions.

This is very different from what WINE does, and to characterise the situation as like WINE seems to be quite inaccurate.

Which means their own code depends heavily on the Linux code they've borrowed. Which, to my layman's understanding, likely means the resulting combined work is a derived work of the Linux code concerned, at least under US norms. Which means they would need to follow the GPL for *all* the code concerned. Alternatively, if they don't like that, they have the choice of not using the Linux code they don't own.

My understanding is based on advice from US corporate counsel that I have dealt with. Corporate counsel gave us in engineering a rule of thumb that adding such dependencies on others' code likely introduced licensing issues (i.e. created a derived work).