World: r3wp

Doc, it works, with -vvvvvv and even when I've run it as root first. 
It can't create all its logs then, but at least it doesn't crash. 
Thanks again!

Doc, trying to start cheyenne   on win7 with UAC enabled, I get this 
error message: 

You need administrator rights to switch to Windows Services mode 
!
   

thing is I've never tried to switch to windows Service mode!!!


a part from the warning popping up the rebol console, the app seems 
to work ok.

this is with all releases I've tried, including the v142 release.

the warning popping up the rebol console
 Are you running it from sources?

yes

Looking at the sources, I've found a code path leading to this error 
msg where the internal /quiet refinement is not honored. Fixing that.

stupid question, can the source version run as a service, or only 
the binary one?

IIRC, you can run it from sources too. Instead of putting cheyenne.exe 
as the service target, it will put rebol.exe.

But you might have issues with paths.

Max: fix released.

thks.

11/5-23:48:50.843304-[HTTPd] Trying phase url-to-filename ( mod-static 
)

11/5-23:48:50.843599-## Error in [uniserve] : On-received call failed 
with error: make object! [
    code: 312
    type: 'script
    id: 'cannot-use
    arg1: 'path
    arg2: 'none!
    arg3: none
    near: [cfg/root-dir req/in/path req/in/target]
    where: 'rejoin
] !

any idea what can this be?
im not sure how to interpret the error message

(and why can't we see a line number for the error, btw?)

Is your configuration complete? Seems like it may be missing the 
root dir

hm... i thought i wrote a message before the previous one or rather 
together with it... :/  or i just imagined... possible. i was tired.

ah, i wrote it to my collegue then forgot to write here too

so the issue was:

i had no default vhost defined and i was accessing a non-defined 
vhost.


but kaj, how can u read what u've just said from this error message?

those args are completely fucked up too. there is a lit-word of a 
datatype's name but without an exclamation mark, then comes none! 
as lit-word, then comes none as ... i don't know for sure, as a value?

It says cannot use path near cfg/root-dir, so that seems to me to 
mean that the configuration lacks a root-dir :-)

>> cfg: none
== none
>> cfg/root-dir
** Script Error: Cannot use path on none! value
** Near: cfg/root-dir

like this?

Almost. cfg is probably defined, but not the root-dir in it

>> cfg: []                
== []
>> cfg/root-dir           
** Script Error: Invalid path value: root-dir
** Near: cfg/root-dir

it would be like this, wouldn't it?

Ah, maybe cfg is completely lacking, then

but why am i not getting such a nice and descriptive error?
btw, i was trying to catch this error, but no luck:
>> e: catch [cfg/root-dir]
** Script Error: Invalid path value: root-dir
** Near: cfg/root-dir
>> e: try [cfg/root-dir]  
** Script Error: Invalid path value: root-dir
** Near: cfg/root-dir

how can i get this as an error object, do i can explore?

I think you're focusing too much on the details. I never noticed 
all those ' and ! :-)

If you would FORM that error, it may be a bit more like English. 
Or at least in R3

:) i've been told im too meticulous or what

I notice that most programmers have a tendency to loose sight of 
the wood for the trees

btw, the thing above worked. 'e got its value. i was just confused 
by the output.
didn't know that an error! is displayed like that in the console

It's a good skill to be able to sort out details. Most people can't. 
But you need to start with the overview to see the meaning and decide 
where to focus your efforts

i think i have a better understanding of most of the things compared 
to my collegues, so i usually excel in debugging and i think i thank 
this ability of mine being curious about such details, so i can understand 
things a lot deeper than the average fellows

Probably

im just fucking curious, nah. can't help w that :)

and thx for putting me on track :)

Me too, but it's times like these that differences with other people's 
brains pop up that I can't really put my finger on, either :-)

I always considered myself detail oriented, but there's still a difference 
with my peers

By the way, I also removed my default host, but I don't get errors 
when I go to localhost

but why am i not getting such a nice and descriptive error?

 Because I am so used to read disarmed errors, that I don't require 
 pretty printing them. I guess I should change that.

I get an error when accessing a not defined vhost, but not the same 
as yours (it stops on missing 'default section here).

(and why can't we see a line number for the error, btw?)
 See my answer in ~Chit Chat group.

i was running on r136 when i got this error...

13/5-8:10:46.9889960-[HTTPd] Connection closed
13/5-8:11:31.0776360-[HTTPd] Connection closed
13/5-8:12:11.7342000-[HTTPd] Connection closed
13/5-8:12:53.2066060-[HTTPd] Connection closed
13/5-8:13:33.7584240-[HTTPd] Connection closed
13/5-8:14:41.3166000-[HTTPd] Connection closed


im seeing such messages on the cheyenne console in the past few hours.

any idea what can it be? is it normal to have such a message every 
minute?

probably when the keep alive is being released.

but every minute for hours?
im not getting any other messages...

and actually - theoretically - there is no browser window open showing 
anything from the site this cheyenne is supposed to serve

are http://localhost/docs/rfq/and http://localhost/docs/rfqconsidered 
the same resources?

http://localhost/docs/rfqredirects to   http://localhost/docs/rfq/

i was not talking about any specific implementation but about the 
theory of URIs / URLs

of course in the context of apache and the context of mapping a filesystem 
to urls, i know it's the common practice

but should i give different results if it's a webservice?
i had a use case for this ~2days ago, but i cant recall it atm

the rfc doesn't say how URIs are supposed to be analysed.    its 
up to each server to decide such issues.

ah, versioning:
GET obj -- would return the latest version
GET obj/ -- would return the list of version ids
GET obj/version -- would return that specific version