Our db reminds me of old Converse All-Stars wrapped in duct tape.

March 27, 2006 6:58 AM

We're outgrowing our 10 year old Progress database. We need some recommendations.

BACKGROUND
I'm the graphic designer at a distributing company. I was hired 3 years ago to create a catalog of the 5700 items they carry. I immediately set to work on some sort of automated way and achieved spectacular results taking a text dump from their System Critical DB, importing it into Access and scripting InDesign from Access VBA. Along the way, I learned a bit about databases - enough to see how much more the System Critical one could be doing. One of the problems we run into is that updating my pretty database involves doing a text dump on all updated tables and importing, dif querying, appending.

CURRENT STATE
Here's what I know about their system: Progress 6.3 - Level 5 Pro. with dumb terminals throughout the office. We were in the middle of upgrading to Progress 10 when the reseller -Afflink I think? - went out of business so we just have the hardware (not sure what that hardware is). It doesn't seem they have discussed what the plan is (stick with Progress 10 or go to someone else) but I got the go ahead to do some research as I'm much more technically inclined.

I see the company not doing anywhere near as well as it can because it's working harder, not smarter. I believe part of this is the unintuitive formats (quereying is considered writing a program - so only some actually do it). I would like to present them with other options - or at least know more about what we're going towards.

I realize that changing over a System Critical...er...system is a bad idea, especially when some of the workers verge on luddites and only know the system they know...but I really feel (and everybody agrees with me in theory) that we are working at WAY below our potential due to this outdated technology.

The idea of a db driven website (and eventually E-Commerce) requires either using my Access db and doing ASP pages or dumping to mySQL once a week and doing PHP. Running certain graphical reports requires people to list things on Excel for me to import and then generate from. Sure I could build an Access form but then I'm supporting Access on half the computers and our competing databases are growing against each other...and further apart

I realize my questions are scary simple for this kind of system. But really, these guys are business people, the dba has been doing this for years and really just wants to keep doing the same (can't blame him) so it's up to me to have the pie in the sky ideas that can possibly lead us forward (we need it).

MY WISHLIST:
-Barcode Scanners
-Connect a website to the db securely
-Connect to Access front end so I can run catalog generating code
-Have a gui so we can create interfaces to easily wrangle together a bunch of items for monthly flyers (see catalog code)
-Allow people to more easily create reports they need so they can see the system more dynamically and therefore make better choices.
-Can be connected to from the outside without dialing in.


ACTUAL QUESTIONS:
Will any/some/all of these wishlist items be possible if we upgrade to Progress10? (all the actual info is hidden behind marketspeak)

Is Progress a good technology to stick with?

Are there other systems out there that have good reputations? That are easy to learn and to find new functions? Will have clear and obvious benefits?

Are there package solutions customizable enough that we probably won't have to pay for a custom built system (since we don't even use built in features to their fullest, we're probably not that special)

Would hiring a consultant be the best idea in this situation?

Just how expensive are Oracle, or IBM's solutions? The more explicit the better, because I have no clue what's average for this sort of industry.

I can't help but feel we're trying to reinvent the wheel - or at least, we're sick of the cubes at the end of our axles, so if you can guys...point me towards Goodyear.

posted by Brainy to computers & internet (13 comments total)

The problem with Progress isn't that you're outgrowing it, but that it is a bit cryptic and arcane to deal with. None of this is helped by the fact that almost no good books on the database from an author outside the Progress corporation exist.

Newer versions of Progress do provide ODBC and SQL access to data. Progress also has the ability to integrate with Crystal Reports. These things could be used with Access, eCommerce, and graphical report generation. You could also look at running some reports as batch jobs at night and having them available via email or in a shared directory.

I think Progress is a decent technology to continue to use. I think it would take a lot more money to switch away from Progress than it would take to upgrade Progress and customize it to better suit the company's needs. I would upgrade and work out any additional modules/pieces you may need to buy.

Get yourself to some Progress training classes and insert yourself between end users and the DBA. Work with users to define the reports they need and then with the DBA to provide solutions.

Progress isn't flashy or trendy as far as database technology goes, but it is mature and powerful enough to do everything you're asking for--except the part where you want a VPN.
posted by joelr at 8:32 AM on March 27, 2006

I'll second what joelr says, with the addendum that getting the progress-->web links was (a coupla years ago, anyway) a confusing and difficult task.

I have a friend who uses Progress for a largeish organisation. They have ...

-Barcode Scanners (Yes)
-Connect a website to the db securely (yes)
-Connect to Access front end so I can run catalog generating code (yes, but no need - all this code can be done in progress)
-Have a gui so we can create interfaces to easily wrangle together a bunch of items for monthly flyers (yes)
-Can be connected to from the outside without dialing in. (I'm not sure what you mean by this.)

With progress, you should be able to keep the old systems, and add the extra things on with progress 10. This way, you'll get both the flexibility you need and you'll keep the luddites happy.

One question - how many users do you have on the system?
posted by seanyboy at 8:54 AM on March 27, 2006

This is a very detailed question; it does sound like you need a consultant.

Relevant information you didn't include: rough budgetary constraints; size of the company; industry area; number of people using the database.

I've never heard of Progress, but I'm not a DBA. You may need to look for help in a more specialized forum. Maybe a more appropriate question for you would be "How do I find a consultant" or "How do I begin researching the answers to this problem". I don't have answers for either, I'm afraid.
posted by amtho at 9:24 AM on March 27, 2006

should have previewed.
posted by amtho at 9:27 AM on March 27, 2006

Thanks for the responses already. I'll add this just because I can for any future respondees. See how lost I am? I don't even know how to ask what I want to ask!

Budget...I really have no idea. Take the low end, make that the middle of the road.

Relevant Size of company: 12 db users in house. 20? salespeople not using db, faxing orders in. Would like to get them online if possible.

Industry is Housewares.

joelr & seanyboy are making me feel better overall. Sean, would you know where those solutions came from?
posted by Brainy at 9:37 AM on March 27, 2006

This is a very detailed question; it does sound like you need a consultant.

I'll second that sentiment.

It sounds like your company needs a tech strategy, not just new technology. There are a lot of unanswered questions about your business process- how is the data getting into this Progress db? What else is being done with this data?

A good place for you to start is by dragging in a Progress sales guy- explain to them that you're midway through a conversion, and you're mad that you sunk this money into hardware that you can't use since your reseller went belly up. Explain that you're using this as an opportunity to re-evaluate that decision in light of new circumstances (e.g. this annoying task you have to do). They'll send over a consultant, pronto, guaranteed.

Spell out your needs (drag the DBA guy along), and have him tell you how Progress can or can't do what you want, and what it's going to cost to get there.

BTW- Your dba tell you that he's been "doing this for years and really just wants to keep doing the same" is probably because he's a Progress DBA who's worried about his skill set if you switch platforms. Be prepared to dig in against this if you start to think that Progress may not be the right strategic decision for you. I've seen companies be hobbled by an "indispensable" tech guy who keeps them wed to bad technology.

To answer one quick question- Oracle/DB2 (IBM) are terrific products, but totally overkill for what you need.
posted by mkultra at 9:58 AM on March 27, 2006

Connecting a website uses webspeed
I'll find out about the barcode scanner, and get back to you.

Given the small number of users, the low amount of data you have, and the fact that you're currently shuffling things in and out of Access anyway, I'd hazard a guess that Progress is overkill for what you really need. Converting to the latest version of Progress may also be too expensive.

If you can find a cheap piece of off-the-shelf software that does 90% of what you need, I'd seriously consider using that. Your Management may disagree.

It does look as though some kind of consultancy may be in order. Of course, most consultants will be ...
- trying to sell a specific solution
- be very expensive
- not know what they're talking about.
- fuck things up and then run away.
So good luck with that.
Personally, I'd allocate myself a couple of weeks into looking at it, and then, if I'm stuck, I'd look at external help.

OK. So, I'll need a bit more information...
I'll need some more rough information for the System Critical side of things...
- How many different screens are used in this system? If you can break these down into Data Entry, Query and Report screens, that would be good.
- How many entry fields does the most complicated screen have?
- How many tables in the database?
- How many different sections in the system. (e.g. "Orders" would be a section, and "Stock Control" would be another section.

My feeling at the moment is that you've either got to scrap everything and start again, or you've got to incrementally add your extra functionality onto the current system. With a small system (20 screens or less), then option (a) may be the way to go. More, and I'd look at a first step of adding ODBC and webspeed connectivity. With the ODBC, you may be able to use off the shelf barcoders to inject the database into your core system.

If the system is really tiny, then you could probably get someone to knock one up for a few thousand dollars. If not, then it would make sense to look for an existing package that did everything you need.

Finally - One great thing about progress is that the code is usually shipped with any system. If you can find a progress forum, they may be able to help you hack together a link to hand-held devices.

As always, take what I say with a pinch of salt. This is a complicated subject, and you'll be suprised at how varied opinion is on something like this which should simple.
posted by seanyboy at 10:05 AM on March 27, 2006

mkultra - that's a good stragegy. And the tech guy is pretty indespensible - he's the oldest son (family run business) and it's hard to argue with the success they've had running Progress in the past...if we can stick with it and it's extensible enough to give us what we need we can all be happy.
posted by Brainy at 10:10 AM on March 27, 2006

Progress Forum
Barcoding
posted by seanyboy at 10:13 AM on March 27, 2006

I think mkultra has hit it. My only concern is that for the size of the database / user base, you're going to be paying too much by upgrading the system you currently have.

There's no harm in getting sales guys over though.
As long as you can resist their evil charms, that's a good idea.
posted by seanyboy at 10:17 AM on March 27, 2006

seanyboy

Just talked to the dba...and got nothing. He seemed a bit busy at the time, but the general feeling was that it had been so long since we set it up (15 years?) that table counts were forgotten - and screens could be counted into the hundreds.

webspeed - now this was a teeny bit what I was afraid of. I can't get ahold of any real examples from that site but it all seems so...proprietary. They mention webobjects, which SEEMS like its just ASP without the massive support. If it's just different terminology that's fine, but if things generate their own (invalid) html, it will be quite frustrating.

I believe this system was knocked up for a couple thousand dollars to begin with.
posted by Brainy at 10:33 AM on March 27, 2006

If your dba doesn't know roughly how many tables there are, then woah...
You're correct. webspeed is proprietary.

I can't believe that screens can be counted in the hundreds. Maybe we're confusing terminologies here. By screens, I mean the base template of what actually appears on the screen. A list of Orders would be one screen. The lines on an order may be another.

without knowing more about the actual system, the politics of the current situation and the companies new and existing requirements, anything I say from now on would be absolute guesswork. I would say (and this is a guess) - get a progress guy in, cost up the conversion to progress 10, and then approach the problem again when you have up to date technology.

sorry I can't be more helpful.
posted by seanyboy at 10:49 AM on March 27, 2006

Sean, you've been VERY helpful. I realize the intricacies of the situation can only lead to vague suggestions but you've given me a lot to think about...and a lot to present. Now I can go out and try and learn more about Progress knowing that it's certainly capable of doing what we want it to, if we know how to do it.

Thanks.
posted by Brainy at 11:07 AM on March 27, 2006