Vincent Parrett

Yes it's a custom written dotnetnuke module (written in c#) - the part that interacts with stripe is quite small. You need to refrence a stripe javascript file on the page, that is the part that actually handles the card number and submits it to stripe, that gives you back a token that you then use to submit a payment request. Our c# code never see's the credit card details, that is all handled client side (in the browser using the stripe api js). See https://stripe.com/docs/payments/accept-a-payment-charges

No it wasn't that much, I think the fees were around AUD$250 (on the original purchase) and the refund around AUD$3000 - not in the office right now to check.

The key is "There are no fees to refund a charge, but the fees from the original charge aren’t returned."

https://stripe.com/docs/refunds#:~:text=There are no fees to,a refund cannot be canceled.

I can highly recommend stripe - https://www.stripe.com - very reliable and they pay weekly (you can choose the payment frequency). My only gripe is that they do not refund the fees when making a refund (apparently due to visa not refunding them) - had a case a few weeks ago where a customer accidently ordered the wrong number of licenses and the partial refund cost me several hundred dollars in fees! I have tried many different payment processors over the years, in my experience, the one's aligned with banks tend to have the worst tech and support.

We are delighted to announce the Release of Continua CI 1.9.2. We have added the following new features: Export and Import: You can now export one or more configurations to a file and import them back from the file into Continua CI. Requeuing Stages: Requeue a failing stage without restarting the build. Multiple Daily Cleanup Rules: Each type of build by-product can now have a different shelf life. https://www.finalbuilder.com/resources/blogs/introducing-the-release-of-continua-ci-version-192 Continua CI is a low cost, easy to use Continuous Integration Server which includes first class support for Delphi (using FinalBuilder or MSBuild) and version control integration with Git, Mercurial, Subversion and more. https://www.finalbuilder.com/resources/blogs/building-delphi-projects-with-continua-ci

Yep, that's the main reason we added the feature for, we find uploading to chocolatey fails occasionally - it's so nice to be able to just try again - we've been dogfooding this feature for a while here and I've had to use it a few times now. Yes, you can certainly edit the exported file. It will be checked for validity when imported - that happens before the change is committed (the import is done in a transaction), so if there are errors in the file nothing bad should happen (this is why this feature took so long to develop!).

FinalBuilder is a fully featured automated build tool, which supports Delphi 3 to 10.4, along with C++Builder 4 or later. FinalBuilder makes it simple to automate your entire build process, from compiling your Delphi projects to compiling and uploading installers, creating ISO's. There are over 600 built in actions, with support for Git, Mercurial, Perforce, Subversion, TFS and many other version control systems. Unlike xml or batch file based systems, with FinalBuilder you can easily debug your build process, with breakpoints, step over, step into etc. Of course FinalBuilder also integrates with Continua CI - our continuous integration server product, and with other CI servers such as Jenkins. Thousands of Software Developers rely on FinalBuilder to automate the build, test and release process. If you are not using FinalBuilder to automate your builds, you are missing out. Download a fully functional 30 day trial version today.

I just had a quick look and it doesn't inspire a lot of confidence. I'm about to evaluate this library - https://www.devart.com/sbridge/ for SFTP (and SSH, FTPS, FTP) - currently using Secureblackbox (.net version) but looking for something faster and more reliable.

This looks to be just what I needed (javascript engine with com/IDispatch support), but like others, I'm torn on whether to use chakra core given MS have pretty much abandoned it (apart from security fixes). The 'community' doesn't seem to be achieving much, that may just be a timing thing (covid). I'm definitely going to take a look at this though, I need something to replace ActiveScripting - I've looked at a bunch of libraries (yours included) over the last year and none had com/IDispatch support, or if they did they were missing other features.

I do this to remap IDE Insight (Ctrl + .) since I use Ctrl+. for the procedure list (old habit from coderush days) - however the IDE has a habit of resetting it while I'm working so I find myself having remap it again several times a day. I thought about writing a plugin to just fire a timer to periodically check if it had reset and then change it again 😉

Another issue. Right now there is a single setting for Math operators, unfortunately that leads to ICompiler = interface function GetSearchPaths : IList<string>; getting formatted as ICompiler = interface function GetSearchPaths : IList < string >; which is rather unfortunate. If GExperts was on github I would fork it, try to fix and submit a pull request.. not sure how that works with sourceforge?

You know what would be nice for the formatter - an option to format on save. Not sure how hard this would be to implement but I'd love to see this.

I'm just trying to format the code for a project.. the GExperts formatter does a much better job than the built in one (in 10.4 at least) - one thing that bugs me (with both) is how it indents guid attributes on interfaces so ICompiler = interface ['{4A56BA53-6ACD-4A5D-8D55-B921D6CDC8A0}'] becomes ICompiler = interface ['{4A56BA53-6ACD-4A5D-8D55-B921D6CDC8A0}'] There doesn't seem to be any way to control this. Also some comment formatting gets messed up const //options for IACList2 ACLO_NONE = 0; // don't enumerate anything ACLO_CURRENTDIR = 1; // enumerate current directory ACLO_MYCOMPUTER = 2; // enumerate MyComputer ACLO_DESKTOP = 4; // enumerate Desktop Folder ACLO_FAVORITES = 8; // enumerate Favorites Folder ACLO_FILESYSONLY = 16; // enumerate only the file system gets changed to const //options for IACList2 ACLO_NONE = 0; // don't enumerate anything ACLO_CURRENTDIR = 1; // enumerate current directory ACLO_MYCOMPUTER = 2; // enumerate MyComputer ACLO_DESKTOP = 4; // enumerate Desktop Folder ACLO_FAVORITES = 8; // enumerate Favorites Folder ACLO_FILESYSONLY = 16; // enumerate only the file system I turned on Align var/const statements at position 30 and it comes close const //options for IACList2 ACLO_NONE = 0; // don't enumerate anything ACLO_CURRENTDIR = 1; // enumerate current directory ACLO_MYCOMPUTER = 2; // enumerate MyComputer ACLO_DESKTOP = 4; // enumerate Desktop Folder ACLO_FAVORITES = 8; // enumerate Favorites Folder ACLO_FILESYSONLY = 16; // enumerate only the file system use At position will not always work. I tend to use tabs to align them manually after the longest line in the block. So in the original (pre format) version there are tabs before the = and before the // - not sure if this is possible in the formatter without doing some block analysis.

Anyone know how to detect specific update version in compiler defines. I'm already doing {$IF CompilerVersion >= 25.0} but I need to detect compiling under 10.2.2 vs 10.2.3 etc

Thanks for the input folks. This has all become a moot point as I realised I can't have 10.2.0/10.2.1/10.2.2/10.2.3 installed at the same time, so it's likely I'll just support the latest patch version for each compiler version. That said, @Uwe Raabe shared (privately) how he handles this (tools api change) in MMX, that will allow people to compile the project with older patch versions themselves if needed.

Thanks. I have it working, but it does introduce a positioning issue with the include statement, can only be included after the uses clause and that uses clause must include the unit with the identifier. Not a huge deal in my case but could be a bit of a gotcha if you were using that same include file to ifdef the uses clause itself!

I suspected as much. It's a change in the tools api I needed to detect, which was added in 10.2.2 . In any case I can't have different patch versions installed to compile against anyway so I guess.

404 - did you forget to make the repo public?

Lots of unhappy Atlassian customers https://community.atlassian.com/t5/Atlassian-Cloud-Migration/To-all-Atlassian-server-champions-we-want-to-hear-from-you/qaq-p/1500873 And yes, it seems like Bamboo is just being quietly shelved, they don't mention it anywhere on their roadmaps or cloud migration pages. My guess is bitbucket piplelines is the intended migration path. I'll take this opportunity to offer a different migration option - https://www.finalbuilder.com/continua-ci - currently windows only, but we're working towards linux & mac support (agents first, server later).

Yes, this is true for all Atlassian products. We're moving away from Confluence because whilst it has support for producing documentation sites, it has no support for versioning.. unless you buy an expensive plugin - which totally takes over and makes a complete mess of things when you install on existing spaces, and and even worse mess (had to restore from backup) when uninstalling it. TBH, I fell out of love with confluence a long time ago.. we'll probably just switch to some sort of static site generator. They announced yesterday they are moving to cloud only and will retire the server/dc products over the next few years - that's just another reason to stop using their products imho.

My take is what he really wants in code completion that actually works? It bugs the hell out of me that when I'm typing an assignment statement and the left side is an enum, that code completion offers all sorts of crap that is not relevant. If the type isn't found, tell me that.

Are all the cards the same height with the same layout? If so then something like this might be faster - https://github.com/VSoftTechnologies/VSoft.VirtualListView - it's working well for me in the application I'm using it for I pre-calculate the layout for the rows when the control resizes, so the painting is fast. It does take a lot more effort to write the layout and painting code but it's very smooth. The sample app doesn't show the pre-calc idea, you can see that here https://github.com/DelphiPackageManager/DPM/blob/master/Source/IDE/DPM.IDE.EditorViewFrame.pas I wrote this control for a specific purpose, but I have tried to keep it reusable for other things, it might still be a little rough around the edges.

Umm, you might want to edit your reply so it doesn't look like I said it! What is the api you are calling? If an api you are calling is blocking, and doesn't take an event handle or have some way of cancelling then you cannot safely kill it. Killing a thread on windows is dangerous, https://docs.microsoft.com/en-us/windows/win32/api/processthreadsapi/nf-processthreadsapi-terminatethread

CancellationTokens are what you need. Parallel.Async as a second optional parameter which is an IOmniTaskConfig - you can create that using the Parallel.TaskConfig class function - then you need to create IOmniCancellationToken and pass that into the taskConfig.CancelWith - when you want to cancel the async, call cancellationtoken,Signal. Your procedure can also take an IOmniTask parameter which will give you access to the Cancellation token in your function, so you can gracefully exit eg Parallel.Async( procedure (const task : IOmniTask) begin // Call a Winapi Blocking API which take 5000 ms to complete if task.CancellationToken.IsSignalled then exit. end ); If the blocking api you call takes a handle or array of handles then you can use task.CancellationToken.Handle - eg in calls to WaitForSingleObject or WaitForMultipleObjects HTH