Main benefit with SQLite is wide compatibility. If you want to use objects there are plenty of ORM solutions available which work with SQLite while not breaking generic file & SQL compatibility.
It is actually really cool how you can interactively move into regexes and refine your searches. I wonder why no other editor (as far as I know) have implemented something similar.
I like the way they handle this at the Sudbury Valley School (http://sudval.org). They essentially have realized that there is no objective way to measure a good teacher, so instead they take the darwinian approach. Every year all students (and staff) vote on which teachers thay want to keep, culling out the bad ones and just keeping the good ones.
They actually have had teachers so good that they have been re-elected every year for more than 40 years.
I'm glad they have good results with this system. I recall a teacher in my high school who was very unpopular with many students. He was an ex military man and what I am about to tell you will sound like a cliché: He believed in strict discipline, penalized students heavily for being slightly late or having corrections on submitted work, and so on. He actually invented a new form of punishment at the school involving hard physical labour. I don't recall anyone having a Hollywood epiphany when subjected to his brutal style, no wonderful breakthroughs or students discovering their hidden depths.
He would never have been re-elected by students. But was he a bad teacher? I don't know! I entered the school interested in the material he taught and left it still interested. Was he a bad teacher? Honestly, I have no idea. All I can say is that he was an unpopular teacher amongst many students.
Actually the source is available and quite a few people have made it run on Linux. From what I hear it is still pretty unstable on Linux, but if a few gurus grabbed hold of it, it would probably quickly get up to speed.
I think it probably could have been working on linux with outside contributors if the license[1] did not contain strict requirements such as maintaining licensing.
So the source is indeed available but it isn't really open so it hurts in outside contributions.
I wouldn't call it opensource. Only the GUI is opensource. The core functionality is binary. You could not port it to OSX or BSD (unless the BSD in question can load linux .so)
What problem do you have with the license? It doesn't require any release of source and it doesn't prevent sale of the software. It looks a lot like a BSD license.
* Any redistribution, in whole or in part, must retain full licensing functionality, without any attempt to change, obscure or in other ways circumvent its intent.
That would imply that you have to keep the binary blobs, and that you wouldn't be allowed to reverse engineer them (say, if the E-text editor guy goes out of business).
It sounds to me like reverse engineering would be fine. All it says is that redistribution must retain the license as-is with no attempt to change or obscure the license.
That's "licensing functionality", which is not necessarily just the license itself. Given that the original program is licensed ($49 or something), that seems pretty clear to me - particularly given that otherwise it's a BSD licence, so clauses 1-2 cover the case that you mention.
Obviously the only real test of the wording is in court, but it'd be enough to make some people steer clear of it.
I don't really have a problem but the licensing is a problem for other people, not to mention that it contains binary blobs (I don't know if it is windows only) for it to work which greatly reduces how much a person can see how the program works in order for them to fix it.
My initial writing was also confusing the 'simplified BSD license' used by FreeBSD (without the "no-endorsement" clause) with the full classic BSD.
And this is another reason to hate license proliferation. Remembering all the "BSD, except X" variations is an error-prone pain. Let's be coding and sharing, not lawyering.
