Cumulus development

[steve]

Edit (25th March): Please ignore this thread. The idea I had for a 'facelift' 1.9 instead of Cumulus 2 is turning into a complete rewrite. I hadn't realised that the only thing that "Delphi Prism" has in common with Delphi, is that for some reason they put "Delphi" into the name. None of the advantages I saw (building on the stability etc) actually exist as most of the code would be new anyway.

So, I have to apologise again, and announce that I am going to continue with the version of Cumulus 2 which is comprsised of 3 parts: A Windows service, which does most of the work, a control application for configuring the service and a 'viewer' for displaying the data. The viewer will look pretty much like the existing Cumulus 2 alpha.

Steve


--------------------------------------------------------------------------


I am part way through restructuring Cumulus 2 into 3 components: 1. A Windows service, which does most of the work. 2. A control application for configuring the service. 3. A 'viewer' for displaying the data.

The complete rewrite which was the original Cumulus 2 was a huge undertaking for a part-time project. The restructured version is even more so. And I am using so much technology that is new (not just to me) that I think the whole thing is just too much for me at the moment. It's interesting learning and using all this new stuff, and if it was only going to run on my computer, it would be OK. But I can't see any way that I am going to be able to support it for use by anyone else. It's hard enough supporting 'old technology' Cumulus, given the many and varied systems that it is used on. We were already seeing situations with Cumulus 2 where it just failed in odd unexplainable ways for some people, in the depths of third party components.

So... What I have decided to do is to start work on a 'new look' version of Cumulus 1. Call it version 1.9 or whatever. The intention is that this will be file-compatible with Cumulus 1.8; it will offer the same facilities (initially) but the 'look and feel' will be more like Cumulus 2 alpha - more 'up to date', if you like. It will have graphs, gauges etc which are similar to those in Cumulus 2. Once I have done the first new version, with the new graph components etc I will be able to start adding some of the features which have been asked for - making things more configurable etc.

An interesting side effect of the way I am doing this is that it might run on Linux. And it might run on Mac OS X. It won't look like a Linux or a Mac OS X application, it will look like a Windows application. But using something called 'mono', the Windows executable should be able to run on those operating systems.

I have already started this and it's going quite well. I am actually enjoying what I am doing, whereas with Cumulus 2 I just had this increasing feeling of impending doom.

My sincere apologies to anyone who was waiting for Cumulus 2. Hopefully you won't be too disappointed with 1.9 (or whatever) instead. Cumulus 1 is reasonably robust and reliable for most people, and I think it's important to build on this.

Steve

[nking]

Steve,

The most important thing is that you should enjoy doing it and I for one applaud all that you do for us

[Synewave]

Steve,

A very well written post.

I think you are absolutely right to to take the route you feel most comfortable with, and most importantly - enjoy it.

I don't think anyone can complain that Cumulus2 may never come, but should admire you for continuing development for Cumulus (whichever version). I and I'm sure most of this community really appreciate everything you do.

Well done.

Paul

[daj]

Cumulus 1 is reasonably robust and reliable for most people, and I think it's important to build on this.

I think this last sentence is the key here -- C1 is very stable, well used and liked. A face lift will be great but no need to change the 'engine' as is working well and is already very flexible.

I completely agree with Neil too -- This is a hobby for you, Steve, and you need to enjoy it and not feel it is weighing you down in anyway.

Brave decision, to be applauded

[geoffw]

I agree with the previous reactions to your news Steve. I've got to admit I rather 'like' C1 and now I'm getting to know it I would prefer development rather than radical change.......

Weather Station still broken I'm afraid and Maplins are being a touch slow to react to my emails. Their tech support helpline charges 50p a minute so I'm not taking that route!!!

I am starting to think I will have to start a thread on what to replace it with......

[Charlie]

Hi Steve,
Have you thought about doing something in between? I'm thinking that capturing an ideal architecture on paper (or electronically), as well as capturing the current structure, you might see a transition plan that does not require 80% of the re-write to be done before there is anything usable.
I'm not sure how modular the code is today, but for instance data collection makes sense to be self contained given multiple flavours of gear out there. If you created a simple service to collect the data and shove in into something like a database, then used the existing code with a simplified collector to get the info from the database, you'd have the first step on the path to independent modules. This would make adding a new station type fairly straight forward.
The next step might be to have the data generated by the application put the output into the database rather than the file storage system of today. Then you can make the new display interface simply display data from the database. With the data collection and data display seperate, adding a new calculation is also dead simple, as is feeding 3rd party applications, or even ports to new operating systems.
Or you can do things in the oposite order - if the new look is highest priority, use the opportunity (since you're touching the code anyway) to create a defined interface between the displaying (look and feel) and the actual data so that when you want to move to a new structure, it's a minor interface change.
By breaking the architecture into bite sized pieces, you could turn this into an open source project if you ever get tired of it, or some of us could help you code certain blocks or functions.

Or maybe you've already been down this path and rejected it for reasons that are obvious once you get into the detail. This is truly useful stuff you've got today and I'd hate to see it be compromised by an overly ambitious undertaking. There is certainly nothing wrong with tweaking what's already there!

Whatever you decide to do, we'll help however we can.

[beteljuice]

I for one don't see a problem.

We must remember that Cumulus is a data capture software - anything else is a bonus !

The 'community' has now become large enough for others to suggest templates / code / presentation.

The greatest thing about this site is the number of Thank Yous for Steve (Deservedly). Now if everybody put their hands in their pockets as well ..........

[6719jason]

Steve,

I think this a great move I have myself, actually been thinking for a while now...How Good it would to have what you are now working on.

Full Support here Steve and Many Thanks for all your time, effort and work that you put into Cumulus and this forum.

In Short - Without Cumulus....My Site and Hobby would be not exist.

Cumulus was the only piece of Weather Station Software (And I tried over 20) That made me want to take it Further than Just Numbers and Figures.

[M1DUL]

I think this a great move aswell Full Support here Steve and Many Thanks for all your time, effort and work that you put into Cumulus and this forum.

[beeman]

agree with all the above comments - many thanks Steve.

Life's too short to do things you don't enjoy.......

[malc-c]

Sounds good to me.

I liked the visual look of C2, but the stability, reliability, plus the ease of use (simple txt or config files which can be edited if required) C1 offers. Even if all you did was to provide a skin over the top of C1 to give it a vista look like C2, but keep the mechanics, IMO would be nice.

But as others have said, you have to enjoy what you are doing, and at your own pace. I personally won't be ditching C1 for any other 3rd party alternative (is there other apart from the supplied software ???), and am prepared to wait for ver 1.9 or whatever you choose to call it.

I think you'll find that other forum members will also be supportive in whatever direction you end up taking Cumulus as we all feel you do a brilliant job producing 1st rate software for us to use with our weather stations

[Hillbilly]

Steve,

I'm pleased you've arrived at a decision which makes you feel happier and more motivated. Cumulus 2 was clearly becoming more and more complicated and you certainly shouldn't be struggling with something part-time and for free which isn't fun.

We have loved everything that Cumulus has given us, I thought it would just be something geeky for my husband but I have loved presenting our website and learning as I go. I am personally beginning to feel I am understanding more about the workings of Cumulus with respect to the website, so wouldn't be in a hurry to switch to Cumulus 2 anyway. So in fact your decision is good news for us.

We have really appreciated what you have done so far and will be privileged if you continue. In the spirit of Beteljuice's comment, I have added a small donation in appreciation of the things you've added recently and the realtime gauges which we added fairly recently.

[bigmac]

Noooooooooooooo! Ah, well. I was looking forward to being able to run Cumulus 2 on my OS WMR200. Never mind. Unless you can find the energy and time to squeeze in a tinsy weensy bit of USB support for the aforementioned WMR200 into V1.9? Please? Pretty Please? With knobs on?

[steve]

Thanks for the supportive replies, everyone. I'll just respond to this comment in particular:

Noooooooooooooo! Ah, well. I was looking forward to being able to run Cumulus 2 on my OS WMR200. Never mind. Unless you can find the energy and time to squeeze in a tinsy weensy bit of USB support for the aforementioned WMR200 into V1.9? Please? Pretty Please? With knobs on?

My plans for supporting other weather stations haven't changed, even though my plans for Cumulus 2 have. I'll still look at supporting the WMR200, whatever the next version of Cumulus happens to be called.

[bigmac]

You are a star Steve! And as a token of appreciation, I have donated a little for the cause. I urge others to do the same! I'm sure every little helps!