World: r3wp
[Core] Discuss core issues
| older newer | first last |
| Henrik 13-Apr-2007 [7497] | trying form-error.r gives: none is missing its obj-a argument OK. :-) It would mean that the function I'm passing directly is missing an argument. I was confused by the none! part. |
| Rebolek 13-Apr-2007 [7498] | Can anybody explain to me, why this thorws 'false even if the page exists? thanks. >> exists? http://www.aving.co.kr/newproduct/default.asp?p1=newproduct&mode=read&table_id=English&num=79 connecting to: www.aving.co.kr == false |
| Graham 13-Apr-2007 [7499] | >> exists? http://www.aving.co.kr/newproduct/default.asp?p1=newproduct&mode=read&table_id=English&num=79 URL Parse: none none www.aving.co.kr none newproduct/ default.asp?p1=newproduct&mode=read&table_id=English&num=79 Net-log: ["Opening" "tcp" "for" "HTTP"] connecting to: www.aving.co.kr Net-log: {HEAD /newproduct/default.asp?p1=newproduct&mode=read&table_id=English&num=79 HTTP/1.0 Accept: */* Connection: close User-Agent: REBOL View 1.3.2.3.1 Host: www.aving.co.kr } Net-log: "HTTP/1.1 500 Internal Server Error" == false >> |
| Oldes 13-Apr-2007 [7500x2] | using trace/net - Net-log: "HTTP/1.1 500 Internal Server Error" that's what the page returns |
ech... too late | |
| Graham 13-Apr-2007 [7502] | hehe |
| Oldes 13-Apr-2007 [7503x3] | it returs 500 response even to my Firefox, but it's displayed anyway.. interesting. |
it's goog vay how to hide content from bots:) | |
(a good way) | |
| Rebolek 13-Apr-2007 [7506] | so is there another way how to check if URL is valid? |
| Graham 13-Apr-2007 [7507x2] | read it ? |
no .. don't know | |
| Sunanda 13-Apr-2007 [7509] | You might be able to fiddle with the http scheme for this condition..... If you do this: zz: open http://www.aving.co.kr probe zz find mold zz "response-actions:" You'll see there is a tale of acceptable reponse codes. So (just an idea) patch that to include 500, and then check if the in-buffer field has any <html> content |
| Oldes 13-Apr-2007 [7510x3] | but that doesn't mean it's valid... the content of the buffer after 500 response is usualy the message that there was an error on the server and that you can contact admins |
if servers like the one above hase content, they are not going in the standard way | |
(hase = has) | |
| Sunanda 13-Apr-2007 [7513] | Good point -- he has to check with some care what's there. Usually, it's a text string, so no HTML. I'm only sugesting he patch the port for *this* website, not all sites; and some eyeballing should help him decide what is a real 500 and what is a faked one.....Maybe, anyway. |
| Oldes 13-Apr-2007 [7514] | anyway... the exists? should probably return true on url which has internal error as the url exists but has problem at this moment. |
| Gabriele 13-Apr-2007 [7515x3] | hard to say, maybe the 500 would turn to a 404 if the error went away... |
since one expects to be able to READ something tha EXISTS?, i think returning FALSE there is the best thing to do - since READ would error out. | |
tha = that | |
| Maxim 15-Apr-2007 [7518] | I just had a language flash... what if we used the comma as a stop word for variable argument enabled funcs... ex: sum 1 2 3 4 , sum would be defined as sum: func [**] [ result: 0 foreach value ** [result: result + item] result] |
| Steeve 15-Apr-2007 [7519] | you don't like sum [ 1 2 3 4] syntax anymore ? ;-) |
| Maxim 15-Apr-2007 [7520] | hehe |
| Steeve 15-Apr-2007 [7521x3] | but i agree for one thing, the comma should be loaded as a valid token in rebol |
so, we shoold free to use it as we want | |
*should be | |
| Maxim 15-Apr-2007 [7524] | darn I hate object oriented development... I've come to a point where I need multiple inheritence in rebol (which I can hack) but some methods overlap and I need both for the object to be compliant... again, I can hack most of this ... but the fact that I just came to this point reminds me of the down sides of using objects (in any language)... |
| Steeve 15-Apr-2007 [7525] | yep agree too |
| Maxim 15-Apr-2007 [7526] | overall a big can of worms :-( the slightest little mistake in the future (forgetting a detail in a year from now) could lead to a 2 day bug fest... even if documented, some side-effects can't be pre-identified. |
| Steeve 15-Apr-2007 [7527x2] | i think it's a problem when any type of source becomme huge |
Rebol don't solve this problem | |
| Maxim 15-Apr-2007 [7529x2] | I'm now thinking of flattening the whole object inheritence tree and adding switches, to enable some options. |
Even in languages which provide some support for this issue... in the end it becomes an issue anyways... every time you want to change something, you end up asking your self more questions because of the "features" of the object paradigm, than the actual problem you are trying to solve. | |
| Steeve 15-Apr-2007 [7531x5] | yep it's an endless issue in our karma |
we are too perfectionnist :-) | |
it's the main raison why gurus can't produce business release of their idea | |
raison = reason | |
there is some exceptions for sure :-) | |
| Micha 16-Apr-2007 [7536] | what use "native" function ? |
| Oldes 16-Apr-2007 [7537] | Native functions are functions that are hardwired into the kernel of REBOL. So you cannot see source of such a functions. http://en.wikibooks.org/wiki/REBOL_Programming/Internals |
| Gregg 16-Apr-2007 [7538] | If you mean the NATIVE func that you can see in REBOL, I think that's used only to document native funcs; to create a REBOL spec for use with HELP. |
| Oldes 16-Apr-2007 [7539] | ach.. interesting.. I missed this function:) |
| btiffin 16-Apr-2007 [7540] | This is just my opinion, but all you guru's... Don't lead Carl to add early termination to function argument scans. Some of us aren't as smart or as thorough as you... :) |
| Anton 17-Apr-2007 [7541] | How do you know that it isn't for your own good, then ? ;) |
| btiffin 17-Apr-2007 [7542] | You are correct sir. But I had to try. :) |
| Maxim 17-Apr-2007 [7543x4] | any one ever encountered an error in REBOL's binding? yesterday I stumbled for hours on a bug in REBOL's interpretation... really. |
for some reason, it seems that binding does not handle sub objects very well. strangely, it seems that in a circumstance I cannot determine (and cannot reproduce) when you have a series of nested objects at some point, even if you remake all parts of the inner objects, you still end up with functions which call previous bindings, thus calling the wrong functions! I had to explicitely call the paths from the supplied outermost object back to the current object in order to reach my own local function: ex: within the body of a function in objc, I had to call this: instance/ctx/other-function instead of only 'other-function. yet they are defined in the same make ! | |
oops objc = ctx | |
I tried recreating the error and somehow it happens only in a circumstance which I can't determine... but its a strange issue cause its the first time I hit this in all my years as a reboler. | |
| older newer | first last |