Hey Chris / whoever has the power to change things
-
Any response on this: http://www.codeproject.com/Messages/5030350/Reputation-for-downloads-of-article-source-code.aspx[^] ?
We can't count downloads offsite. We could trap clicks to offsite hosting of code, but we can't then tell whether they've merely visited a page or actually downloaded the code.
cheers Chris Maunder
-
We can't count downloads offsite. We could trap clicks to offsite hosting of code, but we can't then tell whether they've merely visited a page or actually downloaded the code.
cheers Chris Maunder
Yeah - obviously system is easy to trick... but so is current one - i.e. I could just put up link to empty zip (or text file with link to GitHub) and say it source code download. So, I am wondering, could we find a way to improve this? Like if you are against counting clicks to repository (even though it's the same thing - intent is to browser source code) could I link to zip that downloads repository: https://github.com/lepipele/csharpWebCam/archive/master.zip[^] I definitely think that linking Git repository is improvement over linking zip with source code and people shouldn't be penalized for using it. Thanks for listening and let me know what is decided - I could help with implementing this if you are short on dev resources.
-
Yeah - obviously system is easy to trick... but so is current one - i.e. I could just put up link to empty zip (or text file with link to GitHub) and say it source code download. So, I am wondering, could we find a way to improve this? Like if you are against counting clicks to repository (even though it's the same thing - intent is to browser source code) could I link to zip that downloads repository: https://github.com/lepipele/csharpWebCam/archive/master.zip[^] I definitely think that linking Git repository is improvement over linking zip with source code and people shouldn't be penalized for using it. Thanks for listening and let me know what is decided - I could help with implementing this if you are short on dev resources.
We (speaking as a community member rather than as a CP staff member) definitely prefer projects to have at least a version of your code here on CodeProject rather than out in CodePlex. Part of the reason for this is that it's incredibly easy for a project and an article to get out of sync if they are hosted in different locations. At least, if we have a version here, we have a project that is a snapshot taken at a point in time. By all means, link out to your CodePlex/Github/etc article, but have a version here.
-
We (speaking as a community member rather than as a CP staff member) definitely prefer projects to have at least a version of your code here on CodeProject rather than out in CodePlex. Part of the reason for this is that it's incredibly easy for a project and an article to get out of sync if they are hosted in different locations. At least, if we have a version here, we have a project that is a snapshot taken at a point in time. By all means, link out to your CodePlex/Github/etc article, but have a version here.
-
I mean - you can view this problem from different perspectives. But the core of the problem is if I leave version of code here on CodeProject people will download that version of code and then CAN'T CONTRIBUTE. Exact reason why Git was invented.
There's no reason to shout. Of course they can contribute. That's why I said provide a link to your CodePlex/Github/whatever version. How hard is it for you to put a comment on your article that people should go to your repo to get the latest version if they are looking to contribute? Try looking at the problem from this perspective - suppose someone chooses one of the many, many, many online repositories and then just provides a link to that from their article. Now, take this further and imagine what happens when Google takes this source provider over and decides two years down the line to close it because it no longer aligns with their core business. Now where do people go to get the code?
-
There's no reason to shout. Of course they can contribute. That's why I said provide a link to your CodePlex/Github/whatever version. How hard is it for you to put a comment on your article that people should go to your repo to get the latest version if they are looking to contribute? Try looking at the problem from this perspective - suppose someone chooses one of the many, many, many online repositories and then just provides a link to that from their article. Now, take this further and imagine what happens when Google takes this source provider over and decides two years down the line to close it because it no longer aligns with their core business. Now where do people go to get the code?
-
Yeah - obviously system is easy to trick... but so is current one - i.e. I could just put up link to empty zip (or text file with link to GitHub) and say it source code download. So, I am wondering, could we find a way to improve this? Like if you are against counting clicks to repository (even though it's the same thing - intent is to browser source code) could I link to zip that downloads repository: https://github.com/lepipele/csharpWebCam/archive/master.zip[^] I definitely think that linking Git repository is improvement over linking zip with source code and people shouldn't be penalized for using it. Thanks for listening and let me know what is decided - I could help with implementing this if you are short on dev resources.
As I said, recording a click outside to another site is easy, but tracking what happens on that site is impossible. We simply can't tell what's going on outside of CodeProject. At a deeper level, reputation points are for what people do on CodeProject, not what they do outside of CodeProject. If we support GitHub then we also have to support BitBucket, and CodePlex, and SourceForge, and then what about gists, or Samples on JSFiddle? It goes on an on. I understand where you are coming from and it's a nice idea, but for sanity's sake we'll keep it simple.
cheers Chris Maunder
-
As I said, recording a click outside to another site is easy, but tracking what happens on that site is impossible. We simply can't tell what's going on outside of CodeProject. At a deeper level, reputation points are for what people do on CodeProject, not what they do outside of CodeProject. If we support GitHub then we also have to support BitBucket, and CodePlex, and SourceForge, and then what about gists, or Samples on JSFiddle? It goes on an on. I understand where you are coming from and it's a nice idea, but for sanity's sake we'll keep it simple.
cheers Chris Maunder
Well, I disagree with some points - but as I said to Pete - I'll stop wasting everyone's time and figure out what I can do on my own on this topic. Maybe add text file to zip with source code or something... will see what works best with what is available to me. Thanks for taking time to respond and I appreciate that you said you understand the idea - who knows maybe something will be done about this down the road...
-
There's no reason to shout. Of course they can contribute. That's why I said provide a link to your CodePlex/Github/whatever version. How hard is it for you to put a comment on your article that people should go to your repo to get the latest version if they are looking to contribute? Try looking at the problem from this perspective - suppose someone chooses one of the many, many, many online repositories and then just provides a link to that from their article. Now, take this further and imagine what happens when Google takes this source provider over and decides two years down the line to close it because it no longer aligns with their core business. Now where do people go to get the code?
:thumbsup: Not only because of companies might take over the hosting provider. There are many code hosting sites out there and the author himself might someday want (or has) to change the site he's hosting his code on (e.g. not too long ago a hosting provider shut down after he got hacked). If the article doesn't get updated there isn't a working link to the code any longer.