|
I just read an article here on CP where the author says "For many years I've been building a library themed "things that should be built into the .NET framework, but aren’t"."
I too have been compiling things I use regularly into my own framework. Everything from logging, registry stuff, app security, WPF things, Data related stuff, ZIP, messaging, etc.). In my WPF library I even subclassed the standad WPF controls and then in my WPF apps I use my version of the controls.
For the most part, when I find myself using code repeatedly I start thinking about where in my framework to put it so that I can re-use it again later.
So I'm curious... Do you do this? Create a library/framework of reusable code? If you, how do you put it all together? If not, how do you handle using/reusing common code?
If it's not broken, fix it until it is.
Everything makes sense in someone's mind.
Ya can't fix stupid.
|
|
|
|
|
I do for my embedded software/firmware, makes life a lot simpler.
New version: WinHeist Version 2.2.2 Beta I told my psychiatrist that I was hearing voices in my head. He said you don't have a psychiatrist!
|
|
|
|
|
Kevin Marois wrote: Do you do this?
Constantly
We're embarking on a full rewrite in part because lots of what we've done over the years has finally been implemented in the BCL. Though not as well as we did it, natch'
cheers
Chris Maunder
|
|
|
|
|
So if the BCL now contains a function that you created in the past do you then go back & refactor to use the BCL version?
If it's not broken, fix it until it is.
Everything makes sense in someone's mind.
Ya can't fix stupid.
|
|
|
|
|
Sometimes, but not as often as we should. In some cases we've grown the system around our methods in a way that's different enough that using the BCL methods (or entire patterns) that are now available would mean more work than it's worth.
Hence the rewrite. It's actually rare that a rewrite actually ever makes sense, but in this case it does.
cheers
Chris Maunder
|
|
|
|
|
I can see the pro's and con's of both. For one, you KNOW you've written. It's tested and working, whereas new framework code could be troublesome. While on the flip side having framework code would mean less source you've got to maintain.
If it's not broken, fix it until it is.
Everything makes sense in someone's mind.
Ya can't fix stupid.
|
|
|
|
|
I don't. The only things I reuse are tiny snippets that I write from memory.
|
|
|
|
|
But that's got to mean an increase in testing time. I find that the benefit of creating my own library/framework is that once a component is written and tested I can use it freely in other apps with a high degree of confidence that it's going to work.
If it's not broken, fix it until it is.
Everything makes sense in someone's mind.
Ya can't fix stupid.
|
|
|
|
|
Perhaps, obviously I'm not really in a position to make a comparison.
But I don't expect so, since it's not that I'm just refusing to reuse code, the problem is there isn't any opportunity for it. Except for those trivial snippets (such as, idk, inverting a permutation or something, that comes up occasionally), but I don't test those anyway. (inb4 public lynching) That's not where the bugs are, and the context they appear in will probably be tested anyway.
You have some bigger reusable parts so it makes sense that it would make a difference in that case
|
|
|
|
|
Same here. No external library, just code snippets.
|
|
|
|
|
Kevin Marois wrote: So I'm curious... Do you do this? Create a library/framework of reusable code? I had a small library for WinForms, but have been doing mostly web-applications for the last years. Not really building a new library, since it would be outdated before it is started.
The only interesting thing in the library was a virtualized datagridview that accepts an IDataReader for quick re-use.
Bastard Programmer from Hell
If you can't read my code, try converting it here[^]
|
|
|
|
|
I think just about everyone does, don't they?
I've got personal libraries for every language I've ever worked in.
I wanna be a eunuchs developer! Pass me a bread knife!
|
|
|
|
|
Harold doesn't...
I think that any serious developer would understand the need for their own collection of code.
For the last 15 year I've been doing .Net. Before that I was a FoxPro developer. I has a very large collection of resuable code that I later developed into a framework that i tried to market as FramePro. Never caught on as a selling point but I used it in many apps.
If it's not broken, fix it until it is.
Everything makes sense in someone's mind.
Ya can't fix stupid.
|
|
|
|
|
Kevin Marois wrote: I later developed into a framework that i tried to market as FramePro The one thing I have resisted is trying to "market" my framework. I do leave it behind at every contract I have worked on. I know of at least 5 developers who have adopted it (ie snaffled the code from previous contracts) but I have never been tempted to publicise it.
I agree that most serious developers create a suite of tools/snippets and concepts that become part of their infrastructure, but to me it is a very personal thing suited to the way a developer works and it's transferablility is quite limited.
Never underestimate the power of human stupidity
RAH
|
|
|
|
|
Mark_Wallace wrote: I've got personal libraries for every language I've ever worked in.
Same here. I still have the 20+ classes I wrote for VB4-6, mostly ones that wrapped the Windows API (not that they're of any current value). OTOH, my current Word normal.dotm file has macros that were written in the Word 95 days; they still work just fine.
I appreciate Harold's POV in that a lot of what I write is not situated for re-use, although the ideas behind it are. However, I find I do write a lot of things that are re-usable: encapsulating export of grids to Excel (including formatting the resulting workbook), extensions to the treeview and listview, encapsulating Windows file searches, and encapsulating database functions.
Accessing WinAPI functions from VB was a PITA and it made sense to encapsulate the functions so they were easy to use. This "encapsulate hard to use functions" mindset of my VB days bleeds forward -- I look for things I'll probably re-use and automatically think about creating a re-usable class.
|
|
|
|
|
Oh yes.
When I write code, I write it with an eye to "generalization" so that it can be moved into a library - and I have have several just for C#, each containing different but related material - once it's shown to work and can be properly tested.
If you don't keep libraries of "good code" then you will be re-inventing the wheel far, far, too much.
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
|
|
|
|
|
Agreed
If it's not broken, fix it until it is.
Everything makes sense in someone's mind.
Ya can't fix stupid.
|
|
|
|
|
I'm in the same boat. My "goodies" library lives inside a single DLL that doesn't have any external dependency, so I can very easily add it to any project.
For years I went out of my way to have it target .NET 2.0 (what I've considered to be the lowest common denominator), but I've "upgraded" it to target .NET 4 last year or so only because 2/3.5 are no longer installed by default on the newer OSes. But, it still doesn't rely on anything from .NET 4; I could retarget it for .NET 2 and it would still compile.
I don't use interfaces nearly as much as I should, but I make sure function signatures don't change from one version to the next...so in theory whenever I add/fix something in the DLL, I can drop in the replacement file and all my apps should continue to work.
If it gets more complex than that, then it doesn't belong in my utilities library.
|
|
|
|
|
|
Yes I do.
For Android (some .aar libraries) and C#.
Couldn't do without it.
Sometimes you have to find the edge of "how closely are these and those classes related to each other" and decide, whether to put them in one library together (which means more memory usage / download size, especially on mobile platforms) or to split them (which means, more references, maybe more builds, a bit more organizational work in the project structures).
In most cases I prefer to do more, but smaller libraries, than to do one single dinosaur who "eats everything".
It adds flexibility on the cost of a bit more time to set up a new project. But the setup is done once... the memory is used every time the app starts.
Exception is my WinForms library, but the past showed, that indeed most of the classes in there are used in each of my winforms apps.
I strictly divide code into "Business logic" - which means, code that "does the job of the app" and "everything else".
As long as I have not made the "everything else" code as generalized as needed to be library-ready I consider it a prototype.
I have finished platform (non-business) code of my app when I have succesfully outsourced it to a library project. not a minute sooner.
|
|
|
|
|
Yes.
I have a couple of libraries (being primary a SharePoint Developer), I’ve got a Common, Winforms, ASP.NET and SharePoint libraries (including unit tests) which I’ve built up over the years of consulting.
|
|
|
|
|
Anything that can be generalized is added to the library.
Some of it even ends up as articles. But the backlog seems to be growing, mundane things such as family, work and procrastination, tends to get in the way.
|
|
|
|
|
I'm doing it even for a higher level stuff like search dialogs, connectivity, data handling etc. I'm lazy as f***, so I reuse every single bit of code.
There is only one Vera Farmiga and Salma Hayek is her prophet!
Advertise here – minimum three posts per day are guaranteed.
|
|
|
|
|
I do, and as soon as i figure out it needs to be enlarged or split i do so to keep it on point. Though it's not always easy since all dependend projects could be affected with the change.
Rules for the FOSW ![ ^]
if(this.signature != "")
{
MessageBox.Show("This is my signature: " + Environment.NewLine + signature);
}
else
{
MessageBox.Show("404-Signature not found");
}
|
|
|
|
|
I have a tiny one for automating view and viewmodel comms, and lots of share code in base classes in my framework. So far I still just Copy-Paste the first steps from the previous project. It is a framework, but only within each new project and not yet ready as a proper stand-alone framework.
|
|
|
|