LC7 & Unicode

[Richard Gaskin]

TEDennis wrote:

> Kay C Lan: Well, I'm glad you got that off your chest.  Feel better?
>
> It seems you have an emotional tie to somebody or something at RunRev.  It
> could simply be that your years of Revolution/LiveCode usage has created a
> strong loyalty.  Whatever ...  It's likely your view is somewhat biased
> towards the positive.

One man's "bias toward the positive" is another man's "bias toward the 
productive". :)

I thought Kay's post lent a welcome balance to the discussion.

But before we rush into "Richard's an apologist fanboy!" (the former is 
not the case, but I'll gladly admit the latter), let's see if we can 
avoid ad hominem altogether and instead look at Kay's comments on their 
own merit.

For example, the reference to iOS 8 is a good and fair one, IMO:

The transition from LiveCode v6 to v7 was in many ways similar in scope 
to Apple's migration from OS 9 to OS X, yet even with the relatively 
modest scope of change in iOS 8 we saw one of the wealthiest and most 
powerful multinationals in the world stumble, and have to release 8.0.1 
shortly after - and then it was discovered that 8.0.1 had critical 
issues, and rushed 8.0.2 out right after.

This isn't to suggest that Apple fell down on the job.  It's just a 
healthy reminder that sometimes software engineering is less trivial 
than we might prefer.


> re: Apple wouldn't let such a performance hit get past the design phase
> would they - unless of course you call it Lion, Mountain Lion, Mavericks or
> Yosemite.
>
> Your comparison of the relative complexities of LC's application development
> platform and Apple's full blown operating system is almost comical.  But,
> that discussion is beyond the scope of this forum.

Such a perspective may actually be a point in LiveCode's favor:

LiveCode is far more complex than just about any consumer app; it really 
does have much more in common with an OS or virtual machine.  It's a 
sort of meta-OS, allowing us to write apps to a single API which then 
get translated across a range of OS APIs that few developers of even 
simple consumer apps dare to attempt.

It would seem LiveCode's been doing a pretty good job if we find 
ourselves with a perspective that lets us take this level of effort for 
granted.

If we look at just one small corner of that, the pulsing default button 
on OS X, we'd see a level of effort deeper than I care to write here but 
suffice to say nothing in the OS APIs supports what LiveCode allows its 
users to do with that.

And that's just one very small corner of one of seven platforms LiveCode 
deploys to.

Of course our participation in improving the quality of the tools we 
rely on is essential, and it benefits no one to whitewash areas of 
actual concern.

But I won't hold it against Kay or anyone else if they enjoy using 
LiveCode.  I certainly wouldn't feel compelled to dismiss their thoughts 
as "emotional", any more than I would dismiss complaints on such grounds.

To move forward productively it's helpful to look at what's working and 
not working based on our own direct experience.

This may be a good example:

> I am, and will continue to be, a supporter of this company and its goals.
> But, that doesn't mean I have to sit by quietly and let issues that affected
> an entire user community go by with nary a comment.

"Entire user community" is pretty big, and neither Kay nor anyone else 
has suggested no one report bugs.

Since I migrated my work to v7 I find I'm working at least as 
productively as under v6.x, in some ways more so given the effort Fraser 
put in with GDK integration and that I spend nearly half of my time 
these days on Ubuntu (most of the rest on Mac and a bit on Windows, FWIW).

But I can't presume to speak for everyone, and more importantly I don't 
believe any single one of us can.

I know firsthand many who are enjoying v7.  Sure, like all previous 
versions, and like nearly all software ever written, it has bugs.  But 
the degree to which bugs specific to v7 are preventing anyone from using 
it has not been clearly defined, and seems likely to be as frequently 
the result of social memetics than actual firsthand experience.

In fact, I think Kay's observation about human nature here is relevant:

"And then of course the human trait not to believe what we've been told."

This is not to dismiss actual bugs, but it does help remind us that 
readers of this list enjoy a great many repetitive posts about a 
relatively small number of issues from an even smaller number of posters.

When we look at the community's discussion as a whole, the number of 
actual issues is far small than the many posts about them.

So let's find issues and report them.  And when we have trouble pinning 
down a recipe, lets explore the issue here and work together to find the 
recipe that will make it reproducible for the team, and thus fixable.

But let's also try to maintain a perspective of actual scope here, 
relying more often on what we've seen for ourselves rather than what we 
heard from someone who who heard it from someone else.


Let me put my time where my mouth is, and see if I can help steward your 
concern here toward a fix:

 > Unfortunately, 7.0 caused a problem with "eof" being reported when
 > the last record in the file was read, instead of on the next read.

Would the next read contain data, or be empty?

If the latter, it may not be a bug, but a fix for a bug that we'd just 
come to accept.

If the former, let's report it.

Have you submitted it?  If not, do you have a sample script I might use 
to verify the issue and submit it for you?

-- 
  Richard Gaskin
  Fourth World Systems
  Software Design and Development for the Desktop, Mobile, and the Web
  ____________________________________________________________________
  Ambassador at FourthWorld.com                http://www.FourthWorld.com