Re: put put
Just for the record: It was a server side problem. The server required connection: keep alive in the http headers to fully download the pdf file. Thanks for the help though. All the best, Malte___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: put put
Ok, I don´t get it. It does not seem a general problem, which makes it even stranger. At least I know now it does not have to do with the 4.0 engine. Also happens with a 3.5 standalone. The scenario is as follows. rev client communicates with a server. It fails most of the time on one XP and at least one Win 7 machines. A whole bunch of other XP machines and Vista machines are known to work. If it fails, it is the download process that does (thanks for the suggestion Dave). I get all kinds of fun errors. First: error 10038 on socket, which if I understand it correctly means, I am not communicating with a socket at all. Next error: No headers sent. Next the result is empty, but only half of the data is received (exactly the half). This is s weird. The only scenario where I could reproduce it, without having to open a teamviewer session is a machine that hosts both server and client. Now it gets strange. If I connect from a remote machine to said server (while the same IP is used) all works fine on the remote machine. If I connect the "broken" machine to a remote server, it also works. also there are at least 10 other machines (both Macs and PCs) where the same scenario (one machine hosts server and client) just works perfectly... I have no idea... Just standing here shrugging. I will now try to load URL first and then write from the cachedUrls. If that´s no worky I give up suspecting that the wee free men got hold of the machine and just left some critters in there. Tis not me. I start doubting it is rev, though I can access the documents with a webbrowser. Overcaffeinated. Clueless... *shrug* Thanks for your time. Malte___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: put put
Malte, I think there may be a few minor changes in the version shipping with Rev 4. (But nothing that obviously changes previous working.) In your code, I'd suggest you split the following line: > put URL tPath into URL ("binFile:"& tTempPath) into two parts, and add error checking after each part. put URL tPath into tData if the result is not empty then answer the result -- or whatever else put tData into URL ("binFile:"& tTempPath) if the result is not empty then answer the result -- or whatever end if end if In that way you can see if the problem concerns the download or saving to file. Cheers Dave On 8 Dec 2009, at 15:18, Malte Pfaff-Brill wrote: > I think something is fishy with libURL in Rev 4. Some script that used to > work in previous versions now breaks with the current engine. > > local tPath, tTempPath >put getServerUrl()&"test.pdf" into tPath >put the tempname & ".pdf" into tTempPath >libURLSetStatusCallback "dataupdate",the long id of sb 1 of stack > "progressupdate" >send "init" to stack "progressUpdate" >set the filetype to "prvwPDF " >put URL tPath into URL ("binFile:"& tTempPath) >libURLSetStatusCallback >close stack "progressUpdate" >launch document tTempPath > > This works sometimes(on Windows XP), most of the time it does not and we end > up with half downloaded files or 0 byte files. Rather strange, because until > now it used to work. Has anyone else seen strangeness with put? Any > workarounds we could try? > > Cheers, > > Malte > ___ > use-revolution mailing list > use-revolution@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your subscription > preferences: > http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: put put
Malte, do you still have the dps of version 4? Try building with dp-3 or dp-4 but not with gm-1 see if it goes away... Cheers andre On Tue, Dec 8, 2009 at 1:18 PM, Malte Pfaff-Brill wrote: > I think something is fishy with libURL in Rev 4. Some script that used to > work in previous versions now breaks with the current engine. > > local tPath, tTempPath >put getServerUrl()&"test.pdf" into tPath >put the tempname & ".pdf" into tTempPath >libURLSetStatusCallback "dataupdate",the long id of sb 1 of stack > "progressupdate" >send "init" to stack "progressUpdate" >set the filetype to "prvwPDF " >put URL tPath into URL ("binFile:"& tTempPath) >libURLSetStatusCallback >close stack "progressUpdate" >launch document tTempPath > > This works sometimes(on Windows XP), most of the time it does not and we > end up with half downloaded files or 0 byte files. Rather strange, because > until now it used to work. Has anyone else seen strangeness with put? Any > workarounds we could try? > > Cheers, > > Malte > ___ > use-revolution mailing list > use-revolution@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your > subscription preferences: > http://lists.runrev.com/mailman/listinfo/use-revolution > -- http://www.andregarzia.com All We Do Is Code. ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: put put
I think something is fishy with libURL in Rev 4. Some script that used to work in previous versions now breaks with the current engine. local tPath, tTempPath put getServerUrl()&"test.pdf" into tPath put the tempname & ".pdf" into tTempPath libURLSetStatusCallback "dataupdate",the long id of sb 1 of stack "progressupdate" send "init" to stack "progressUpdate" set the filetype to "prvwPDF " put URL tPath into URL ("binFile:"& tTempPath) libURLSetStatusCallback close stack "progressUpdate" launch document tTempPath This works sometimes(on Windows XP), most of the time it does not and we end up with half downloaded files or 0 byte files. Rather strange, because until now it used to work. Has anyone else seen strangeness with put? Any workarounds we could try? Cheers, Malte ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Put Put
Andre. It wasn't me. I just butted in. Craig ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Put Put
Craig, put should be blocking, actually, I know it is blocking! I've been blocked by it many times. I don't know why you're experiencing that behaviour at all but I am trying to work around it. I don't know how familiar you are with liburl callbacks. You can basically tell it to execute a command and when that command completes execute something else. In your case you could use liburlftpupload call and pass the callback you want to execute when it complete. Bonus point this is asynchronous so there's no blocking and it doesn't need to have one. You still face the previous request not completed if the thing hangs out forever but I do think it will queue the transactions so that multiple calls just fill a queue but I can recall if that works with ftp calls. Andre On Mon, Dec 7, 2009 at 4:59 PM, wrote: > > In a message dated 12/7/09 1:54:15 PM, an...@andregarzia.com writes: > > Andre. > > I am always trying to change my habits, to use "send in time" instead of > wait. But here I intended it to be used well before the handler actually > does > anything, likely right after the "case" statement. A simple block of errant > keypresses, and no danger, as you pointed out, of anything running amok. > > I get it, though. > > > Craig, > > > > never use a wait call while there's a network transaction going on. > LibURL > > does not like wait commands. > > > > Andre > > > > ___ > use-revolution mailing list > use-revolution@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your > subscription preferences: > http://lists.runrev.com/mailman/listinfo/use-revolution > -- http://www.andregarzia.com All We Do Is Code. ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Put Put
In a message dated 12/7/09 1:54:15 PM, an...@andregarzia.com writes: Andre. I am always trying to change my habits, to use "send in time" instead of wait. But here I intended it to be used well before the handler actually does anything, likely right after the "case" statement. A simple block of errant keypresses, and no danger, as you pointed out, of anything running amok. I get it, though. > Craig, > > never use a wait call while there's a network transaction going on. LibURL > does not like wait commands. > > Andre > ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Put Put
Craig, never use a wait call while there's a network transaction going on. LibURL does not like wait commands. Andre On Mon, Dec 7, 2009 at 4:47 PM, wrote: > I am a newbie here, but would inserting a "wait for messages" prevent the > errant keypress? > > But is that slightly different than what you really want, which is to > ignore double keypresses? In HC I would have said: > > ... > wait 15 > if the mouseClick then exit yourHandler > ... > > This works in rev, too. > > Craig Newman > ___ > use-revolution mailing list > use-revolution@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your > subscription preferences: > http://lists.runrev.com/mailman/listinfo/use-revolution > -- http://www.andregarzia.com All We Do Is Code. ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Put Put
I am a newbie here, but would inserting a "wait for messages" prevent the errant keypress? But is that slightly different than what you really want, which is to ignore double keypresses? In HC I would have said: ... wait 15 if the mouseClick then exit yourHandler ... This works in rev, too. Craig Newman ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Put Put
Dan, this is me guessing here, but why don't you use a script local variable like local lRunning on doSave if lRunning is true then send "doSave" to me in 200 millisecs end if put true into lRunning put "ftp://admin:passw...@123.456.789.012/UploadTest.txt"; into ftpPath --use a valid path put "This is a test" into url ftpPath if the result <> "" then put false into lRunning answer the result end if put false into lRunning end doSave With such script you can be sure that no doSaves can work at the same time, of course, you can enter a loop if doSave never returns... On Mon, Dec 7, 2009 at 4:34 PM, Dan Friedman wrote: > Hiya... > > Isn't the "put" command blocking? I find that if I call a 'put' to write > to a server twice in fast succession, the second 'put' is handled BEFORE the > first one is complete. For example: > > I made a new stack with this script: > > on doSave > put "ftp://admin:passw...@123.456.789.012/UploadTest.txt"; into ftpPath > --use a valid path > put "This is a test" into url ftpPath > if the result <> "" then >answer the result > end if > end doSave > > Then, I added a menu with a Command-S "Save" option with this script: > > on menuPick pWhich > switch pWhich > case "Save" >doSave >break > case "Quit" >--Insert script for Quit menu item here >break > end switch > end menuPick > > Then, type Command-S twice - quickly. POOF! You get "error Previous > request not complete". Shouldn't the first Save complete before the second > one is even dealt with? What am I doing wrong? > > Thanks in advance, > -Dan___ > use-revolution mailing list > use-revolution@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your > subscription preferences: > http://lists.runrev.com/mailman/listinfo/use-revolution > -- http://www.andregarzia.com All We Do Is Code. ___ use-revolution mailing list use-revolution@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution