Results 1 to 10 of 22
Enjoy an ad free experience by logging in. Not a member yet? Register.
Given the choice, what software would you really like developed?
So, if you could ask a developer for one piece of software (reasonable, not too big), with or without a GUI, what would it be and what functionality would it offer? I don't have the capacity for huge projects, such as office software, or browsers, or anything like that, so it would have to be something small to start out with.
I am going to be looking into the answers here because I am wanting to start a project, but I have no clue what I want to do, so who better to ask than the users?
The choice language is C++, and the gui if chosen would be in QT, to be as cross platform as possible and it is the toolkit I prefer for GUI.
So, if you could have any program you wanted, what would it be?
Code is Poetry
Depend on which category of software you want to write :
easiest are administrative ;
moderate are game / related ;
hardest are mathematical / financial.
Anyway, you said you want a simple application. Why not a text editor (wow, I am so original ...). Or a basic e-mail program.
I want to do something useful, and there are a million text editors out there, the same with email programs (and really, does anyone not use webmail?). I use webmail for my email, and I use sublime text for my text editor.
So, the question stands, it isn't what I want to write, I am just trying to get a feel for what the community wants. Not anything really big, but what would make your life easier on Linux?
Also, in answering this question, it may spark some discussion on the smaller things that are missing that most devs overlook or just don't care about.
PS. I have written some admin stuff for Gentoo, but everyone just complained that it did something that another tool for some random other distors could possibly do, so I keep it private now.
Code is Poetry
- Join Date
- Jul 2008
Personally since I am in the trades. Any software that interfaces with tuning and trouble-shooting autos and motorcycles, cad programs. Stuff that working stiffs like I use.
Office apps are covered by Linux very well while the trades stuff is pretty much Windows and Mac only and DRM and costs you would not believe.
The software like
MyTune is an easy to use Windows program that takes data logged from your bike and generates a tune tailored specifically for your bike. It was developed by a home tuner for home tuners. You don't need to have a deep understanding of the Harley EFI system to use MyTune, it takes care of all the hard work for you.
Edit: I realize what I am asking may be too hard to do. Special cables and Special onboard computer boxes that control sensors and swtiches on the car or bike
may be too hard to hack for some. Plus. Computer boxes change out every few years on a model change for car or bike. But. You asked.I refuse to let fear and fear of others rule my life. It puts my humanity at risk.
Accepting Death is the only way to stay alive.
But yes, keep the ideas coming!
Code is Poetry
I would like to see a no BS Netflix player. Silverlight is supposed to work, and I know I haven't put much effort into getting it to work, but it doesn't just install and work, at least on my system.Registered Linux user #526930
Code is Poetry
1. Egress firewall based on iptables with cli or gui interface (suggest develop a decent cli and add a gui later).
Administration of firewall owner/program combination based which the user authorises new outgoing connections on a case by case bases.
Default to admin and authorising connections through a firewall_admin linux user account. Allow each user to launch an application like firefox, claws-mail, links etc and decide if they want to give it internet access. GUI with option to behaves in a similar way to zonealarm in windows for single instance authorization - with notification on toolbar or user selected mechanism. Only open ports for user/group/program combination rather than open http port for everything by anyone.
2. data vault with data_vault_admin (possibly one per user) chroot user only capable of compare/diff/viewing file contents and cp or rsync files to/from the data vault. Old/replacement file md5sum checks to confirm vault information integrity before deleting temporary file worked on by user application such as libreoffice. It should ensure a file placed in the vault is not corrupted before deleting any temporary file which the regular user works with. Option to archive old vault content rather than replace. Options to backup per file/per folder/per user/entire vault.
A timer. But not just any timer. I work in a call center. We are allowed to load custom software on to our systems "within reason". We now have several different small programs in use throughout our department that started as somebody's pet project or short cut. Almost all of my tools are browser based and interface with a remote server. The remote systems are, IMHO, rather poorly implemented and slower than sin. My stats are tracked many different ways, including on my computer and on a phone. One of the ways I am tracked is the amount of time that it takes for me to finish a ticket once I am off of the phone. There are ~12 different DBs that I draw data from to finish a ticket. I am given three minutes per ticket to wrap the ticket and either resolve it or escalate it then drop back in to the queue. But, sometimes, simple things like attaching a small screen shot to a ticket can take 30-45 seconds. And if it is a complex case with lots of errors involved I can easily be looking at needing to attach four or five screen shots; easily a 2-3 minute wait just for one step. I could give you another ~16-18 examples like that. And all the time that I spend sitting there watching the doughnut of death is being counted against my times / stats; which I don't think is particularly fair. Especially when a complex ticket can easily take 10 minutes to complete and ~ 6-7 minutes of that is me waiting on the system to attach that file and populate this search field. So, I would just like a nice little simple timer that tracked how much accumulated time I spend throughout the day waiting on my tools / system to respond. That way when I get pulled in for the "you're a bad kid / you're stats are all fouled up" talk I could reply: "Ah, Ha! You're not taking in to account the 67 minutes I spent today waiting for our rather poorly implemented systems to respond to my queries!" And in the long run it would benefit more than just me. The bosses would want to audit the code. If it was clean and kept good metrics then they would push it out to other agents in my center and if it proved useful enough they would push it out to other call centers. I'm not going to lie and say that my motives are all altruistic; my stats effect my pay rate (bonuses) and promotion opportunities. (Nor would I take credit for code that is not mine. And at some point I'd hope there'd be a way in there for you to make some money or at least get a really good job with my company. ) But we all agree, even the bosses agree, that a lot of times the client tries to hold us to some pretty ridiculous standards considering the significant amount of lag inherent to the tools that they give us to work with. That being the case, if we had a better way to track all of that accumulated lag across time for individual agents, departments and centers we could then use that information to request that the client either improve the responsiveness of the tools that they provide us with, and /or increase the number of agents that we have as we are having a very hard time meeting the current standards. The net effect would be to both make the agents lives a little less stressful because we would not be sweating like mad while impossibly trying to beat the clock and improve our overall ability to deliver customer service, which would make our 150K users days a little bit better as well.
Last edited by Steven_G; 11-09-2013 at 06:23 AM.
That sounds easy, but it is extremely deceptive. First, to record time accurately, the timer would have to interact with each specific program (which would require more than just documentation) to check its "wait" time, because while you are waiting, the program is still chugging along doing stuff internally (so no real downtime). Not only that, but it would require network lag checking and timing accumulation, and that can be very inconsistent.
Honestly, the best way to prove what you want done is with a video camera.
Not saying it is impossible, but without the programs and a lot of internal code, it is very improbable.
Not to mention there really isn't a fast way to access database stuff, so that would bottleneck you every time.
Programming can do a lot of things, but some just don't realize the amount of effort that goes into something as small as a timer like is being talked about. The software that is to be timed is probably closed source, which means no chance of interfacing with it really, and there would be very little chance of actually timing anything within it because it is working while you aren't, unless you timed the amount of work it is doing, which would be hard to do anyway (without the source that is). I really wish I could help with this, but it looks to be beyond my scope and it would possibly have to be implemented by the software makers themselves.
Great ideas guys, more!
Code is Poetry