larethian wrote:I'm afraid you guys may be wasting your time without even getting an approval from the big boss to present a proposal. Of course you can always do it and start a new translation group on your own (in fact, this will be my suggestion).
As I believe I told Simon (i'm not 100% sure I did), I plan on using the system for my own purposes even if TLG doesn't decide to use this new system, so as far as I'm concerned no time is being wasted.
larethian wrote:
1. How many phases are you going to divide the project into?
Still undecided. I have started setting out the requirements for stage one on the Github Rep, and will figure out how many more will be needed as feature requests come in.
larethian wrote:
2. How many man-hours will be needed?
Again, very hard to tell before the feature list has been completed, but quite a few I do believe. Add in that an hour spent by someone on the weekend as part of their free time is not equal to an hour spent by a full-time paid coder, and the measurement of Man-hour becomes ambiguous... but that's another discussion, and all the people involved fall into the community free time coders category..
larethian wrote:
3. Will you have a software test plan and QA procedures? (for a SW of this scale and where security is important, you would need one)
As you said, we need them, and we will have the Milestone releases hosted somewhere (I'm still debating between one of my Virtual Servers and Heroku), and possibly even host the rolling version somewhere as well.
larethian wrote:
4. For all the things you want to do, will there be an impact in QoS?
I'd prefer not, but at this stage I am not sure... I have sites running on top of the same framework that load in reasonable time (~200ms to load completely with caching in effect), so I am not foreseeing any great QoS drop.
larethian wrote:
5. Are you going to open source the code?
Yes, MIT style
larethian wrote:
6. Maintenance, who will be responsible?
Well, as I said, I plan to use it myself so at the very least I will maintain it. As it is open source the community is free to fork it at any time if they feel my direction is not the same as where BT wants to go. I will comment my code, and the development documentation which we are using to sort out our ideas will be released so that anyone that wishes to work on the code will get a better idea of where we were coming from.
larethian wrote:
7. Deployment, what kind of infrastructure will be needed? (note that TLG has been hosting baka-tsuki on his own servers from his own pocket and have been against us supporting him with server and infrastructure costs)
Python 2.7, and Django, though you can install Django in a python virtualenv with pip. You could host this on the current server, or just as easily host this on a PaaS could system, such as Heroku, where I sometimes place my Personal Projects when I want a friend to take a look and recommend changes... It's flexible.
larethian wrote:
Currently, this is only 'talk'. If you are serious about wanting to do this, before we even talk about prototypes, you should come up with a sound proposal and plan that can impress some of the people here (not me, since I'm no expert), especially the big boss.
My original intention was to present it after I was finished, but Simon wanted to get some input from the community. As confirmation from from TLG and the other community leaders is not the be all and end all as far as I'm concerned I will most likely not wait for a sign of interest from them before starting. I am open to suggestions from them however, as this will be software available to them, should they want it.
I plan on adding more details to the original post, including some links, but I'll wait, as I'm to tired at the moment.