This is related to the question I posted here, but I hadn't gotten much visibility for that question so I wanted to ask in a more general way. I have a Qt 4.7 project that utilizes QFtp functionality. Until very recently we were using this with an FTP server that was vsftpd. Everything worked fine with it then. However, several days ago we moved the server to a new computer. All the contents are identical, but now it uses pure-ftpd instead of vsftpd. Since the move, none of my QFtp code works properly. Is there any known problems that arise when trying to use QFtp with this type of FTP server? I can't find anything helpful online, and it's rather frustrating not being able to find anything wrong with the code and yet having it not work. If anyone knows anything about this and could please share, I'd appreciate it a lot. Thanks!
So I think I just figured something out... I had it run QFtp::list to go through the ftp and retrieve directories, then use list() again on those to retrieve the files in those subdirectories. Our subdirectories to get files out of had spaces in the name, eg "My Directory". Apparently, the vsftpd we were using before could handle this with no problem, but the pure-ftpd can't handle spaces in the directory names. When I switch it to something like "MyDirectory" or "My_Directory", the pure-ftpd works fine. I couldn't find anything online about this difference, but apparently it's there, because that fixed the issue I was having.
Related
I created a very customized leaflet map on a Bitrix website (they forced me to, not my choice). Now other coworkers who are basically "afraid" of code need to be able to add markers to that. I already created a C++ program where they can simply enter all the details they want (what category, whats the popupcontent etc.) and it spits out the geoJSON code for the marker for them to copy and paste into the website.
To make it even more easy for them I am wondering if there is a way to basically have my program connect to the internet, go to the backend of my website and, after asking for login, adds the code to the respective .js file that contains only the marker code.
I have been googling the problem but unfortunately couldnt find any other related posts.
Okay I finally found the I guess easiest way, I will force my colleagues to install python and write a little thingy to concatenate the code and upload it using Selenium. Thanks for your help guys!
I'm struggling with getting the Xbox Live sign-in working for my game. Everything else is ready to go and this is now the final blocker.
Looking at the MSDN documentation page - here, it looks like it should be trivial. However, I'm not using CX/CPP, merely CPP / winrt and therefore, I'd expect to use Microsoft.Xbox.Live.SDK.WinRT.UWP rather than Microsoft.Xbox.Live.Sdk.UWP.
In my project, I've added the link, but it's then not clear what I'm meant to do next. I was expecting to then have the C++ headers be created automatically but I can't see anything being built. Does anyone either:
Know the equivalent code to use to do sign in in the CPP/WinRT world?
Know that it can't be done / I should use a different approach?
Any help gratefully received. Apologies if this is Cpp/winrt 101. I've been using the templates from MS Blogs / Chuck W and so everything else has 'just worked' and so I'm probably missing something obvious.
Thanks
Steve
I'm sorry if this question sounds a little childish, but I'm very new to C++ and console applications.
I recently wrote a HTTP1.1-Client to down/upload files. Everything works fine. The Problem is that there are quite a lot of logs to be output on the terminal and there is also the content of the remote host that should be print to the terminal. At the end of the application some of the output is not well formed - sometimes the order ist not right and sometimes the messages are just broken apart.
The application does not use multithreading and uses the libraries SPDLOG and FMT::FORMAT to output the data. The code is way to big to be put in here but if you'd anyway like to see it i can as well post it here.
I'd be thankful for every suggestion about what I might have done wrong and what the problem might be. Thanks!
I’m hoping to port an application to open vms using HP C++7.3, but at the minute just trying out a few examples to establish if it's a potential way forward but a bit stuck. Being completely new to open vms it's being a lot of fun so far.
For starters I've not got very far building boost as a whole, after some googling I found a few mentions of people having dcl scripts to build boost but no actual scripts. I don’t suppose anyone has any or can pint me to some. I ‘only’ need asio and threading.
Currently I’m just trying to compile the asio chat examples. And as I said I’m completely new to VMS and the compiler, I managed to get a few things to compile but now get a “could not open source file "sys/select.h"’, a bit of googling suggests I may need to have an HP porting library installed or maybe I’m just missing an prepocessor directive to include the correct code, can’t see anything obvious in socket_types.hpp to set?
Any pointers in the right general direction would be gratefully appreciated.
No answers since the questio is too open ended and in a realm I know little about.
Just some thoughts.
Select can be provided by the OpenVMS TCPIP services (aka UCX), or other TCP stack.
At the dcl prompt try: $ HELP TCPIP_Services Programming_Interfaces Socket_API
Async programming on OpenVMS is typically 'event driven'.
Google for SYS$QIO and AST
There is a public access OpenVMS system which holds 'NOTES FILES' (pre-cursor to web-forums)
Try TELNET to EISNER.DECUS.ORG and you'll get sign up instructions.
Once you are in find your way to the PORTING_TO_VMS conference
Once in the NOTES conference you can issue the command DIRECTORY or DIR/TIT=xxx
Unfortunately there is not topic for select, but you can find some interesting
reading using NOTES> SEARCH "select(" followed by repeated SEARCH commands)
( $ NOTES ... OPEN PORT ... SEARCH "select(" ... SEARCH [] ... SEARCH ... )
The PERL folks did an admirable job getting open stuff to work on OpenVMS.
Good luck!
boost library on OpenVMS can be used directly from headers.
I remember one thing to make sure was that BOOST logical is defined concealed, so the relative-includes work.
_____ I have a Django website on my Linux server.
Goal : i want to make web jquery shell/bash of my Linux server , where my site is located , so i could directly from web site pass commands to linux and get all the outputs ,etc.
Questions
Question 1: I know about new subprocess module and popen class , haven't tried it yet , but surely will but right now i experimented with os module. So why os.popen('ls -l') works , but os.open('ls -l >> somefile.txt') doesn't.
Questions 2: i know that when you use screen command somewhere in some file everything that is done is written to a file . Any idea where and what file it is ? Or which file contains everything that is output to bash ?
I am pretty sure that once i come home and try some more :D i will have much more concrete questions but for now that is all... Any suggestions or pointers where i should go , what i should read are appreciated. Thx in advance.
Do you plan to change SSH to web interface of terminal? I would not do that... You have amazing stuff called SSH and which allows you to perform even complex commands securely. I don't believe that you will manage to finish your "webterminal" with the same functionality. But you never know :) Think about transferring files though it, also security (how you will encode the input you will send, or you will just send it as a plain text? Of course https could help here also). I think you will have a lot of fun :)
P.S. I know this post isn't very helpful, but just wanted to make sure if you're not trying to reinvent the wheel...
Well i found an answer myself.
Didn't magically code it yet , but soon will .
So as we all know we have a Linux command screen. This command can make logs of everything that is happening in your screen , by pushing "ctrl-A H" , who uses it knows. But the problem is, that logs are not written immediately , so it doesn't help much . But screen command has also another interesting command called hardcopy , which basically forcefully dumps all screen content copy to a file , where from you'll be able to get result as soon as you'll type it.
I maybe be not accurate and right about explanations of commands , but generally it works like this .
Security
I am in no way Linux expert , and haven't done it yet , but as much as i understand person is logged as user in a screen session , unless he knows root login and password or is super mega hacker which can obtain it, there is basically nothing he can do to destroy the system.