Aquaculture software

[jaybro]

OH sorry I forgot to mention... quick note. I have an internet facing web server with a seperate sql server available if you need space/resources for this project. I can set you up a remote desktop/etc as well if needed for local access to these resources. Anything I can do, let me know.

[mwp]

Nicole, FWIW I'm using an Excel spreadsheet to track my tanks - charts things like water parameters, etc...I mark rows by days and time and also have a section for comments (plus all the columns for water parameters). It's downside is that it's hard to track a specific spawning event or look it up; the upside is perhaps all the graphing from the data (i.e. lets me notice trends in calcium levels, temp etc...)

FWIW,

Matt

[NicoleC]

Jay, I have a domain and SQL Server I use as a sandbox, but I'm thinking long term. A web app requires a long term infrastructure that I am not willing to commit to.

On the use case issue, as long as livestock can be split and merged into tanks, that will work from a model based around a livestock entity. Those who want tank detail will be able to access it, those who don't can just use generic tank info to track parameters.

Rough DB structure coming up (for the technically inclined.)

[jaybro]

I was offering it as a developmental solution, not a production solution. That being said, it sounds like you have a developmental solution all ready taken care of!

[NicoleC]

Okay, here's the DB model. Not all the fields are in place, of course, but it shows the relationships. Highly normalized recursive designs will give you a headache to try to visualize, but it's the only structure that I think will work for ALL the items.

BTW, on the note about calendar items, that doesn't mean ONLY milestones can be made to go on a calendar. Just that you can set items to show up on calendars by default.

There are a few tables missing already -- for example, a nutrition table where you can store a list of food items you commonly use so they populate a dropdrop box for quick selection (or something like that), but since they don't relate to the other tables, I left them off.

Enjoy.

[spk]

Nicole,

Sorry for the late reply. With regard to the LocationID. I was thinking along the lines of the ability to share data between people. You would need to have a way to identify the source of information.

Say for instance, you were populating your db with data, and perhaps I was populating mine, we could then share our data with each other and upon inport a new ID is created that allows us to have two seperate sets of data that we can look at.

It basically means that we can see what others are doing and keep it seperate with out having multiple DB.

Perhaps the other way is to allow the app to open a different data base rather than go through the whole hassel of structung the db to allow for multiple id's Yes that sounds a lot easier.


Your DB structure looks great.

Steve

[NicoleC]

When you say "share" data -- what exactly do you want to share?

[spk]

Experiences, Feeding regimes, success rates...... etc.

But it would be far easier for me to send to some one who wants it a copy of the database rather then the problems of an extract, import routine

Steve

[NicoleC]

Sending a copy of a whole database will include all your tanks, customization, costs of supplies, etc. I wouldn't want that, would you?

Export/import doesn't need to be a hassle if designed properly. But I think sharing electronic data should be tabled until there's a program to put data IN.

The most common item I can think of to share are a data file that summarizes a fish, for example if I buy a fish from Morgman and want to breed it, I might want to attach a record of hatch date, etc. instead of retyping it all.

Otherwise, I might want to share a report of a batch's progress or some other static report, which will really be to complex to send from one stand alone application to another. Of course, the export could be in Excel or a text file or even XML, and that could be used by almost any other program that wants to deal with parsing out the info -- for example if someone wanted to host a site that shared/compared breeding reports.

[Coraldynamics]

Nicole, would having a webserver you don't have to maintain help you out "in case you tire of the project" then the info would not be lost. I can provide the webspace via my non-commercial website (reefobsession.com) It currently houses the Clownfish Surveys and Eric Borneman's Elegance Coral Project stuff.

It does have access to a SQL database server (surveys use it)

Thoughts, concerns? I was planning on moving the surveys to another server but have been putting it off. So I can continue to put it off for a while.

[NicoleC]

Well, guys, Stephen has made a geneous offer, so I will put it to a vote. As you may have figured out, my preference is for a stand-alone app. I like web apps, but I don't think they are a good choice in this situation. I will probably not use any web app for this and will build a simpler desktop version for my own use, but I am willing to spend some dev time on a community project.

Here's the pros and cons:

Standalone app PROS
*Development time is faster. I can have a functioning beta version up in about a month. ETA 3 months for a preliminary, reduced function web version to be in beta.
*Your data is private, and you share only what you choose and with whom
*Your have access to your data on your own computer anytime. With a web app, you must be connected to the internet to add, change or view any records
*You can continue to use the software indefinately, even if support is withdrawn for the web site or software changes
*You can choose not to upgrade if you don't want to


Web app PROS
*Easier to share development, if person/people will commit to helping
*Immune to local computer crashes; backups are made on the server level (one hopes; you won't know for sure)
*Works on any computer with an internet connection and compatible browser
*Upgrades are immediate and easier, including changes to backend DB design, although it will require downtime
*Live data can be made available for public view

[Morgman]

Nicole-

I think something that would be interesting about your previous idea would be have some type of linage certificate much like the AKC does for dogs.

In other words, a type of paper(or digital) that gives customers or end users some idea(as accurate as one can be) to where their fish came from and its generations of CB etc. Then ten or 20 years down the road some breeders may have some pretty deep lineages of generations of CB intermixed with wild caught species. So when they get their clowns they can say hey look the great grandparents of this clown came from ORA broodstock in the 90s that was breed with this xx clowns for 15 years etc etc.

I know this is obviously a hard thing to do in the real world.

[NicoleC]

I won't be hard within the application itself and the DB structure supports it as outline above to infinite generations.

However, some sort of centralized registration organization would be required for it all to work together. I can just see it now... my female clown is "Nicole's Tangerine Janice" and the male is "Nemo's Adventure Jimmy."

Right now it's not critical, but someday when species' start getting banned from wild capture, it would certainly help produce healthy pairings to know how closely related and inbred certain fish are, or to use line breeding to produce new colors/patterns.

[rsman]

I know "here he goes being a problem AGAIN" well why not both

though for this scale i suspect that at least a few of the con's for the web one will not really be a problem. i also see some of the con's for the desktop version not being a problem either

will I use it, donno, might, ive got a LOT of records to convert, but i just might consider it, how much will I make public, probibly all of it as i dont need to hide my secrets from my neighbors



i dont understand you say a month to release a beta does that mean you have some kinda life or job ????


as I run quickly and far for cover

if your into doubling up the efforts I wont mind doing the web work to have a beta out in a month ish, assuming we can come up with a data structure that we can agree on , and features we can impliment in both in a beta release.

as for privacy I see it as a 2 part thing, users who used the web one would have to trust those that set it up that the information they desired to keep private was kept private, but a simple check for private box seems to work for me ???

the structure I see starts with a 2 part key a date and a system description, the date would be the date it started, then as tanks/filters .... were changed they would be added in another table [tanks] and [filters] with descriptions and dates as the PK then from there you have a [breedingfish] with date and tank as the PK if you moved the fish then a new record would be started using the origional fish and a new tank, then other tables for things like larva, rearing, growout and food sources and such.

this establishes a timeline, the timeline itself would have to be handled in SW not the DB but will be easy to impliment in either. you just start at the top "system" and filter down to whatever was connected to it. selecting all of the filters connected to system, then all the tanks, a "end date" field would signial that the tank was to stop at some point on the timeline.

i started a visio thing on it, but while i dont have a life i am stuck at work so my play time is limited.

[NicoleC]

Do the Visio thing when you have a chance, Richard. What I visualize from your text doesn't support a lot of basic items and involves a lot of redundant data, and someone already wrote that application and called it "Microsoft Project," -- so maybe I'm failing to adequately interpret your vision.

[rsman]

hehehe

actually have project open right now, didnt draw the connection, but hey why not

this isnt going to be a HUGE database no matter how you work it a little redundancy here or there isnt going to kill much but im supposed to be working so I can go home before midnight even though this is way more fun

[NicoleC]

My concerns about redundancy are multifacetted.

First, I'm a DB developer first and a programmer second, and normalization is kinda like opium to a DB developer. I can't help it; I get the jitters without it, and if I'm gonna break normalization I want a darn good reason for it.

Second, as a programmer, non-normalization seems easier to work with at first, but it only takes one logical but previously unconsidered request to make you have to go back and restructure the database and causes more work in the long run.

Third, I detest duplicate data entry. Once a person enters something, there should be no reason they should ever have to enter it again. Anything else is pathetic design. It's a waste of time and effort and this is why some people hate computers.

Finally (okay maybe not finally, but finally for now), repetitive actions stink. If you have to perform the same actions over and over to use a program, that's poor design. The program should handle repetitive tasks for you.

Not saying your schema will require that, but that's what I'm seeing so far. So do your diagram when you have a chance and shut me up.

Now, it's gotten awfully quiet in here. Ignore the geek talk if you don't grok it; this is the normal phase where developers get all geeky and have fantastic but irrational dreams about a product. Trust me, the gritty reality will kick in soon.

Requests and comments from regular users are much needed!

[rsman]

well I'm a programmer first and DBA LAST ish !!!


[geek talk goes in here]

I dont see a new fish record just because the fish has been moved, I see seperate tables linking fish to tanks,



[end geek talk]

so how does everyone else label their tanks, I had(well some still are there ) mine labeled in systems FW1 FW2 for fresh water 1 and 2, then tanks in numbers so FW1-1 was the first tank, for FW1 that was really a critter keeper but plumbed in with an overflow and return so ill call it a tank

then SW1 which was clowns all clowns all breeding SW1-1 was a pair of perculas SW1-2 was a pair of clarki's and so on .....

when i refered to my notes id refere to the system being seperated by different notepads, then the tanks in my only semi-orginized state id mix up the tanks by just recording the tank number date and aprox time (sometimes that was dark or light or sometimes it was an hour, for some things i recorded minutes when it comes to specific events often in cultures like

in the folder SC
2 11:30 - 11:40 :
     ran air at an increased rate hoping to clean up bottom
     of culture air rate fast enough to cause foam, no cell
     damage noted culture not cleaned
     *** at next change clean it!

SC is my salt water cultures 2 was nanochloropsus in a 30 gal setup, ran a long time, then crashed big time, but it became clear it was going to i just ran out of time to fix it first.

but how else is everyone recording such events ????

is paper and pen the norm (its how i did it and yep im a geek )

[NicoleC]

[geek talk goes in here]

I dont see a new fish record just because the fish has been moved, I see seperate tables linking fish to tanks,

[end geek talk]

Yes, in my diagram that's the tblLivestockTankLink table. You deactivate the row in the linking table, then creat a new row that is active. Physical tanks don't change; livestock doesn't change, they just get moved around, and the linking talbe builds a history.

As for a tank numbering system... I don't have enough tanks to number

[rsman]

As for a tank numbering system... I don't have enough tanks to number
which is why i asked how everyone else did it, i mean its just as good to have the tank number be a string, but do you assign some name/number to each tank??

[Coraldynamics]

Right now I am in the same boat as Nicole, not enough to number however with Bangaii, Seahorses, and new clown pairs coming online in the next month or so, I will have the need.

What I envisioned are:

Hatch Tanks (non plumbed)
H-1 thru whatever

GrowOut Tanks (Plumbed)
GO-1 thru whatever

Broodstock Tanks (Plumbed)
BS-1 thru whatever

Kinda get the idea...

I don't see a need to separate them further like B-GO1 (Bangaii Grow Out) as they could and will be used for any of the species being bred.

It's early and I ramble with coffee yet....

[FMarini]

you guys have lost me totally.......
whatever it is you guys are doing, it sounds great. Keep up the good work.
signed
Confused

[Dman]

my tanks go like this:
P1A-P1R (Parent system 1, tanks A thru R)
P2A-P2G (Parent system 2, tanks A thru G)
P3A-P3H (Parent system 3, tanks A thru H)

H1A-H1CC (Hatch system 1, tanks A thru CC)
H2A-H2J (Hatch system 2, tanks A thru J)

G1A-G1E (Grow out sytem 1, tanks A thru E)

I had started running an excel spread sheet to track hatch times and days, salinity, temperature, pH, and daily and cummulative losses. But once I had more than 4 pairs hatcing it got old REAL fast, not to mention that it started to get cumbersome as spreadsheet are only two dimensional and when you need to track multiple pairs over long periods coupled with cummulative losses a third dimension is really handy.

I have saved some of my earlier work if anyone is interested (provided I can find it) and found that entering the same value for things like temperature and pH got to be a deterrent.

I have absolutely NO DBA, C?? or programming experience and can therefore only argue logically.

I'l chime in with my thoughts on the whole web/SA debate when I have a bit more time.

[JodiLynn]

Wow. Didn't realize to be a breeder you needed to be a computer programmer. I'm beginning to see the connections though.

I've only taken one computer programming class in my college career and it was Java. So I'm not any help.

It would be neat if it would construct graphs, like Quicken does with your money. But maybe that's more complicated.

Whatever the end product is, I would buy it and use it! Or subscribe to a website to help support it! Seems like a very useful product yall are developing.

[jnowell]

I'm still leaning towards the stand alone app (and I'm a web developer). I think if you build in a simple import / export function that the common (non geek) can use, we could make it work easily.

How about a sticky thread here where we can export our data and post a link to the file. That way, in a few months when I decide I want to breed YWG's, and I see that Nicole's data is online, and she is getting 80% success rates I can download, import, and learn!

The interface will likely be cluttered if you have easy access to every species or breeder listed, I would usually just want to see MY data. Then when/if the need arises, I could download someone elses data for review.

In a chart or graph from a web app, a newbies data would be weighted the same as more experienced breeders. I know I wouldn't want anyone trying to emulate my first attempt, nor would I really want everyone to see how many I killed But if I had one hatch where I got 90% to survive, I could export just that data, and offer it up as a very good method.

< geektalk>
if ("Jason" writes it himself)
then ["Jason" gets everything he wants]

else ("Version and work" = Nicole)
then ["Jason" happily uses what he is given]
end sub

< /geektalk>