I used to be a debaser too.
Maybe a bit sad, but time moves on.
It looks like a popular blog post about the decline and fall of dBase has knocked the long-moribund database's website offline. Sic transit gloria mundi? We were rather entertained by a recent blog post on "Delphi Nightmares" mourning the passing of the online store for the dBase website: dBase: 1979-2026. When the post went up …
> systems in Clipper, including an early estate agent database, that ran on networked Amstrad PCW 8512's!
Haaaaaang on.
1. Clipper was a DOS app. The PCW ran CP/M Plus, AKA CP/M 3. You can't run Clipper on a PCW.
2. I do not know of any way to network PCWs that was ever sold. Even an RS/232 serial port was an optional extra.
Source: me. I am a PCW enthusiast. Right at this second there is an Amstrad PCW8512 in a box about 2 metres behind me. I like PCWs. I own 3. I bought this one because it has a hard disk, which is _vanishingly_ rare for PCWs, and I hope to connect it to one of my newer, higher-end PCWs, specifically my 9512+ which has a 3.5" drive -- meaning that media are not made of pure ethically-sourced Pandoran unobtainium.
I wonder if the commentard is confused with the Amstrad PC1512?
> Amstrad PC1512
Could well be. I think I networked a few of those things myself, using SAGE MainLAN.
Meeting techies around the world, I've found a lot of non-Western-Europeans don't realise what a big deal Amstrad (and its German partner Schneider) was in its day. The PC1512 was basically _the_ first affordable clone PC on this side of the Atlantic. Before it, the American ones cost well over £1000 or whatever that was in Deutschmarks etc. A vaguely decent PC was £1500 in the late 1980s, and the Amstrad was minimally usable for £400 with a single floppy.
Add 128kB of RAM and a 2nd floppy and it was a decent enough PC for the time at half the price of a US clone.
Later on the Amstrad PC2386 was the first decent _Windows_ PC when Windows 3.0 came out in 1990: 4MB of RAM, a hard disk, and VGA. And a full 80386DX, not a crippled 386 SX with a half-width memory bus.
Something that could run in 386 Enhanced Mode, out of the box! Nobody bought them, of course. And it was the first mass-market PC without 5.25" drive bays, just before CD-ROMs went mainstream and everyone wanted one.
I wish I'd picked one up when everyone was junking such things, and kept it.
Our virology lab was given an Amstrad PC1512 and a modem as part of an AIDS surveillance system. It ran a bit of proprietary software and had a GEM desktop option. I installed Protext (from Arnor) so that I could write my book reviews in my lunchbreak and carry them home to my 3.5inch floppy equipped CPC to polish and post to BSFA in a jiffy bag. Happy days.
Back then I had a PC1512 and the bigger PC1640.
The main issue with them was that the PC's PSU was in the monitor so you were restricted to the two or so CRT models (one half-compatible CGA clone, the other was an EGA monitor) these computers were available with.
They also had pretty poor EMI shielding.
Also had an Amstrad PPC (a laptop), which could operate from standard C-size batteries.
My first real personal computer was a CPC-6128. Everything else in AU was an order of magnitude more expensive. It came with both CP/M 2.2 and CP/M 3 so could run a fair bit of shareware and fortunately the small business I bought the Amstrad from was happy to convert 5¼" floppies to the peculiar 3" format.
I managed to get a second 3" floppy and put together the cable from IDC connectors so I could run the RMAC assembler and a subset C compiler under CP/M 3.
Thanks to an affordable Amstrad I acquired a lot of practical experience with low level software and tools.
My next PC was one of the first cheaper Taiwanese IBM clones using a NEC V20 processor running MSDOS (I wanted an Amiga but were too expensive.) I also wanted a Hecules card for decent mono text and graphics - CGA was a bit ordinary.
I recall about that time a friend purchased a x86 Amstrad that ran GEM which looked interesting but he ditched it for an IBM clone to run his MSDOS scientific graphical programs that wouldn't run on his Amstrad.
Back then there was a Cambrian explosion of systems and software - some in hind sight decidedly peculiar. The Olivetti PC using a Zilog Z8000 processor seemed pretty odd even at the time - it was offered for sale by a major Australian department store but I couldn't imagine much software being available or even development tools. If it had sported a MMU it might have run Unix but it didn't.
I worked in a company and little dBase adhoc applications popped up everywhere. Stock control, helpdesk tickets, production progress, QC records, sales leads and many more. One person knocked out all of these little solutions and as far as I know the business bought a single licence copy. In a box with floppy disks and big chunky book manuals. Ashton Tate didn't make much money out of us but gave huge value.
I made my first money as a programmer with dBase II. Writing an inventory system for a car dealership in our town.
While my peers had to stack supermarket shelves or laboured helping with the harvest, I was sitting in an air-conditioned room, hacking away and making more than all but the rich kids could dream of.
Clearly, my soul was tainted then, and a chose a career in IT. Now, I am bitter and cynical and yelling at Cloud.
> dbase Is still alive in Xbase++ from Alaska Software.
TBH I did not even attempt to compile a list of all available things that can work with xBase. I suspect that there were quite a lot of them.
I vaguely remember Paperback Software's VP-Info, which Wikipedia suggests was a native PC-MOS/386 app -- and there weren't many of those.
PC-MOS/386 is FOSS now.
https://github.com/roelandjansen/pcmos386v501
Sadly I've not seen much interest in it.
@Liam... Speaking of vintage tech, I was just wondering if you'd seen a programme that the BBC have uploaded to their "BBC Archive" channel on YouTube: "1992: Remembering the Computer Literacy Project | The Trojan Mouse" [29:37]
It's a rather lovely jaunt down Amnesia Lane; the episode might even contain enough nostalgia to inspire a Reg article about it...
This post has been deleted by its author
The Trojan Mouse, also available on the BBC Rewind Computer Literacy Project website.
Fascinating website, for anyone who has not visited. Complete with all the programmes and all the software in a web-based BBC Micro emulator.
M.
I remember moving a dBase II system off a PET computer to FoxBase on a PC. They were so compatible I recall there were only a few syntax changes to fix. The biggest effort was transferring source and database files and cajoling the printer into working with the PC.
That secured a lot of respect for "the soldering iron department" from the admin staff who hadn't a clue how to do it and were facing a large bill to outsource the job.
I think the 3 custom dBase-II applications I did were my first income from programming. Not that I made a lot - they were all for small local companies. The guy I worked through would do custom printer paper for each one, and I designed outputs to fit them. They all had customer and sales records, and one also had inventory stuff in it.
I was impressed by what one could do on an 8-bit CPU, 64K of memory and one or two floppy disks.
Anecdote: years later, in a later actual job, I walked past an office whose name I recognized as one of my 3. Went in and chatted a bit - they used the dBase-II stuff for quite a few years before moving on to something based on MSDOS (or Windows?) and a hard drive.
I cut my database teeth on dBase III Plus and the Clipper compiler in the late 80s.
I couldn't understand why my PC's hard disk was constantly filling, despite my test data being relatively small.
After much consternation, I eventually found that each string allocated a minimum of 1KB, irrespective of its actual value. Which was kind of inconvenient, given the PC allocated to run the system I was writing only had a 20MB hard disk.
Well.....dBaseIII code I wrote in the 1980's is still running today at Linux Mansions:
(1) Picture database (90,000 images) compiled with Harbour and running on Fedora 44
(2) Books and Inventory databases running on Fedora 44/WINE using M$ Visual FoxPro v6.0
Recommendation: Harbour 3.0 (see https://harbour.github.io/) compiles 1980's dBase code just fine!
Admittedly, I did not have much time to learn SQL, so Borland's Paradox was easier to get on with, and then I discovered PAL (Paradox Application Language) which allowed me to make things smarter (using the BRIEF editor with a Paradox add-on).
I didn't have much money for IT in those days, but that was solved by being a Borland beta tester so I had access to just about anything they made (which in those days came in fat boxes with disks and thick manuals), and I once used a Paradox and a bit of Turbo Pascal code to turn two manweeks of work every month into about 15 minutes of computer effort, most of which was spent downloading a file from a vax over a 1200 baud terminal line.
That was after ignoring management who had this attitude that programming was only done by the gods in some other department who got uppity when you took their work and made it (a) work better and (b) have a better UI. Kinda fun to pull the rug from underneath that attitude but yes, I didn't stay there long. I did a few more things after that but I discovered coding wasn't my thing - I'm merely someone who picks up enough skills to fix a problem, I went more towards large scale networking.
It's still fun to solve problems, though :).
I worked for a company which when it started in 1992 was using dBase III + files for its financial data, with Clipper scripts to enter data. When I started in 1994 the head of the data department was very keen to upgrade to Oracle but the boss was "Why would we do that? dBase works fine, always has, always will" and so we were still using the dBase III+ file format in 2016.
My job in the mid-90s was to replace the old Clipper code with new C++ code, and having worked out that a dBase III+ file was basically a text file with a binary header, it was a doddle to knock out some functions for reading and writing data from/to the files, with the hope that one day we WOULD move the data to an SQL database and I'd only have to change the read/write routines - hopefully?
As it stood that never happened, and after the banking crash the chance of getting the boss to shell out for new servers and temp staff to parallel-run the data for a few months just wasn't going to happen as we slid towards liquidation. I'd worked out how to transfer it all to MySQL by then though the insane levels of inconsistency in file structures was a nightmare.
I was the last man standing who knew how any of that stuff worked, but by the late 90s had moved over to being responsible for all our Linux systems.
The first job where I was paid to write PC software was to reverse engineer a mainframe application to run on a PC using dBase III+. It was a database for storing and analyzing QC data for a manufacturing firm. They hired me to do a number of projects, and this was the first one to get ready while they were setting up the factory.
The mainframe company was EDS, and they had a system which did the same job, but they wanted something like a quarter of a million in one time charges to add us to the system plus an eye watering monthly fee. The company that I worked for weren't going to pay that sort of money so they told me what it needed to do and I had it working in a couple of months, having never seen dBase before. That saved them a very big bundle of money which they were able to spend on QC lab equipment instead (which they then had me set up and program).
We later ported it to Clipper, which I don't seem to recall being a major problem (although this was a while ago). I do recall that the speed improvements of the compiler were not all that great in practice because the real bottleneck was the hard drive throughput, not the CPU.
DBase however was very widely used at the time to create business applications of all sorts.
My impression was that what eventually caused it to fade away was that as more languages and more databases became available at reasonable prices, there was less and less of a market for an interpreter and database as a single integrated system, and they were left with supporting a legacy market and relatively few new applications.
One of first IT things for which I was paid was knocking up something in dBase III on a NEC APC III running CP/M-86 with two 8" floppies.
I recall the dBase environment wasn't too bad and forms based stuff was fairly straightforward. I don't imagine anyone would claim dBase was a DBMS… more a records storage and retrieval system.
Soon after I was building a RDBMS application in C and the Unify 4.2 RDBMS on a Unix platform… quite a contrast; two extremes. Decades later only C and Unix are still kicking but arguably in rather different forms.
For good or ill, the world moves on, even if the ill frequently seems to gazump the good but move on it must.
I worked at an engineering company in late 80's/early 90's where we had an Apple IIGS with a z80 card running cp/m. It had dBase II (or was it III?) on it, and I wrote a bunch of scripts to do basic job tracking. We later got a PC and a friend gave me a copy of Clipper Summer 87 so I ported the code from the Apple II over.
Once I realised how much more I could do with Clipper I ended writing a complete MRP system that the company continued to use long after I'd left.
I later bought Clipper 5 and wrote a tool calibration recording system in it, which I managed to sell a couple of copies of.
Good times.
I answered a job post for a student programmer under a grant program. The job was to write a data driven menu module system for an existing database. I had only heard of dBase prior. I was confident I could pick it up, the grant manager, not so much. I proposed a one week test assignment.
The existing approach was dead simple on the system side and overly complicated for the end user to navigate. Most users would take notes of where they had been, thus producing a tree-like map. Since the menu had users select by number, they also had branch number sequences noted to get to frequently used data.
In that week, I learned enough dBase to duplicate that design by Wednesday, by Friday I changed it to arrow key highlight selection and implemented a path status display. With the addition of a DOS mouse driver mapped to arrow keys, enter, and back; it could be driven by mouse. The grant manager supplied the user, notepad and pen, and the pc with the database. I brought the sample menu module, mouse, mouse driver. At the end of the demonstration the user had an unusually blank notepad, the grant manager was impressed, and I had a student job. Numerous additional modules had to be completed. On the back-end I had also implemented truer "modulization" so each module was easier to implement than the last.
Probably one of the best DOS developement softwares there has been.
Very robust, pretty fast, very customisable.
Ran great in Client Server mode with Advantage XBase Server.
Still using Visual Objects for real world applications today. I find it a lovely language and dev environment to use.
We have sites with 50 or more users, hammering the system, yet VO keeps rrolling on, the data is huge, but reliable.
We get a bad index about 3 or 4 times a year over all of our customers.
One piece of software kept Clipper and VO relevent even today.
Advantage XBase Server (now Advantage Database Server).
I preferer the NLM version, 4x quicker on same hardware to windoze service.
I kicked around dBase II and III, but quickly switched to R:BASE. Coincidentally it was sold under licence by Microsoft, and I believe some of their programmers helped write MS Access. R:BASE worked multiuser straight out of the box and used standard SQL - the menu, forms, reports and language were all very powerful. Even the documentation had appendices with what didn’t work the way it was meant to, with suitable work-arounds.
We still used dBase, until (the disaster of) dBaseIV. For us it wouldn’t work if the network stack was loaded, so we had the idiocy of deciding whether to boot off the hard drive, and not being able to use the database, or boot from floppy for the database, log out, reboot, move it to where it was needed, and reboot again.
A story I heard after the Ashton-Tate, Sybase, Microsoft, SQL Server product was that Ashton-Tate sued the xBase distributors for copyright infringement - then found out that its basis from JPL meant that it was funded by Congress, and therefore not copyrightable - wiping out not only the court cases, but much of the value of the company.
I was grateful for them both, particularly R:BASE, as it was a stepping stone into writing small shrink-wrap systems that I could sell more cheaply than the client-server stuff I was doing before (It was lucrative, but a small number of sales - the smaller systems were easier to sell and kept the lights on).
Once your expensive proprietary app's file format and programming language escape into the wild and become partially standardized, that can make it hard to keep making money from it.
The truth or falsehood of your statement depends on the app and the sellers' revenue model. id Software made, literally, millions of dollars from shareware Doom and Quake. The data formats for those programs were not kept secret. Ordinary (non- id Software) people can create maps and in-game objects which can be used by those games' engines. The source code for the game engines themselves have been released. People have created new, enhanced Doom- and Quake- compatible game engines. At the same time, id Software didn't charge eighty-gazillion dollars apiece for their games.
It seem that the announcement of the death of Dbase may have been premature.
To quote from a recent update from Dbase LLC :
"We're thrilled to share some great news — the dBase LLC servers and websites are officially back up and running. Our team has been working nonstop, and we're excited to say that the core systems you rely on are fully operational again.
Here's what's live right now:
• dBASE.com (main site)
• The dBASE Online Store
• Our full email system"
Sounds more like they were MIA rather than DOS. Anyway ...... :"We're back baby!"
Tate set up a new company called Ashton-Tate – there was no Ashton, but he later bought a parrot, named it Ashton and made it the mascot.
I would bet actual money that was one of those "get our name to the front of the phone book / top of the vendor list" type decisions that so many entrepreneurs made throughout the 20th Century.
It always amazes me when "real" companies use the same sort of logic as the local body shop that goes by "AAA Towing & Repair". And it amazes me even more that the reason they employ those tactics is because they have traditionally proved genuinely effective. You can never go wrong betting on human laziness, and a tendency to pick the first option they're given that seems reasonably acceptable.