Minkata is back open for testing! Age State problem fixed.

[cjkelly1]

True, a constant value no longer was sufficient.

I believe the use of pathconf(..., _PC_NAME_MAX) was a reasonable (and intended) portable answer to that problem. The only platform I found so far that didn’t support it was WIN32.

_R

A final comment on readdir_r(). In a'moaca's commit message, it says:

The biggest change is removing the use of readdir_r which is deprecated.
We didn't really need it, because each thread opens the directory itself.
According to the readdir_r man page on Linux, "However, in modern
implementations (including the glibc implementation), concurrent calls to
readdir(3) that specify different directory streams are thread-safe."

I am curious if this is the case under Solaris. I would consider Solaris to be a "modern implementation", since Oracle is still releasing updates to the OS.

[rarified]

....

The biggest change is removing the use of readdir_r which is deprecated.
We didn't really need it, because each thread opens the directory itself.
According to the readdir_r man page on Linux, "However, in modern
implementations (including the glibc implementation), concurrent calls to
readdir(3) that specify different directory streams are thread-safe."

I am curious if this is the case under Solaris. I would consider Solaris to be a "modern implementation", since Oracle is still releasing updates to the OS.

The Solaris man page is silent on the thread safety of readdir(). But in going to the POSIX standard documents updated last year, it explicitly says readdir() need not be thread safe. Only readdir_r() is called out in the thread safe category. I would have more confidence in the underlying POSIX standard than a specific implementation’s documentation.

It seems the primary concern with readdir_r() is the designation as deprecated. Given that the POSIX standard still calls for readdir_r() and has no indication it will be removed in current revisions, I’m not too concerned about retaining it. I doubt implementations such as Linux, Solaris, or *BSD will risk becoming uncompliant with POSIX.

_R

[Treehugger]

Quick update... I've built a new server with some more debugging code. It shouldn't change things, but given past history, I'm not one to predict

Keep wandering folks, it's great to hear the last version finally fixed the state bug!

_R

Just to report that I created a new avvie and played through, and all seems fine

[rarified]

Thank you, Treehugger! Thorough as always

_R

[Jaytee]

same as Treehugger deleted avie started new one and all works good for me nice job

[PodHopper]

All is looking good so far, finished the first ages, got the cleft book, relto pages and most of the sparklies. I only found one glitch so far: in Teledahn I got into a climbing-down-the-ladder-loop. (much like I've experienced in Moula before.) It was the ladder below the starting point. It's easy to jump out of the loop, so I'm not really troubled by this.

I noticed that walking up and down stairs is going faster than in Moula or the other shards. Although I believe that's great when you have to climb the great stairs for example, it is not looking natural. That's because the avatar is now running slower over horizontal ground than it is in running up the stairs. It is more noticable when the stairs has only few steps.

But I have to admit, every time I do Ahnonay, I find myself wishing for a boost or turbo button for my avatar...