Struggling to attend Borland's Developer Studio Roadshow
David runs late
I was over half an hour late for Borland's Developer Studio Roadshow, despite leaving plenty of time (I thought) for the trip across London - signal failure outside Slough; more signal failures on the Underground at Baker St. This is getting so normal that I start to suspect process failure. A process which might regard symptoms as more important than causes, which might institutionalise underinvestment resulting in little preventative maintenance - and (which really hurts) a process that doesn't penalise poor service because I still have to pay almost £100 for the return trip even if the trains fail to meet their SLA, I have a miserable overcrowded trip and arrive late. Me, titter and bwisted? This is on-topic, honest, because addressing process improvement rather than individual symptoms of failure is what the new (slimmed down) Borland is all about.
But there's another Borland, run by developers (like David Intersimone, known as DavidI), and for developers. This is all about neat languages and cutting better code more efficiently than the competition. This Borland is still alive and well, according to David I and Jason Vokes and they do talk about DevCo (this may be a placeholder name; as I said, my train was late, so I may have missed some caveats).
Yes. Borland is selling it off, but (it claims) it's more of a spin-off, with VC investment, than a sale of the crown jewels to Borland’s rivals in the IDE space.
Call me naive, but I believe all this (Tim Anderson is rather more cynical here, and there are some interesting comments in his blog too). The enthusiasm from the European Developer Studio 2006 Roadshow presenters was palpable; and the audience wasn't hostile (David I promised a continuing life for Delphi, which goes down well in Europe), although it seemed a little wary of the future.
There was even enough of a confused message to remind us that this was the real Borland speaking. The IDE is now Borland Developer Studio, but you can still buy Delphi 2006, C++ Builder 2006, or C# Builder 2006, which all come in three (or two) flavours, so Delphi 2006 Architect edition, for example, includes C++ and C#. Although there isn't a C++ Architect Edition, because C++ Builder is for Win32 (there were a lot of Win 32 developers in the audience) and the others also target .Net.
E&OE - you work it out. The products look excellent, BTW - I just don't see why Borland doesn't simply market Developer Studio with optional language support (an opinion echoed by a questioner in the Roadshow audience).
David I is on the DevCo leadership team and the various roadmaps are part of what's being sold - so it looks like the Borland IDE will continue in the direction it is going (although there's always some risk with a sale). He "expects to excite investors with new products" on the roadmap.
Eclipse is still important (BTW David I claims that Eclipse releases come in June because a lot of the IBM developers on Eclipse are Europeans and want to go on holiday [Americans don’t take holidays] in July - awwww, sweet). David I didn’t announce anything, but the team is also working hard on AJAX; and Ruby support seems to be the gleam in the Intersimone eye...
So, PLOB (Plain Old Borland) seems set fair for the future and even Delphi is far from dead (although you'll probably find C++ or C# easier to sell to your paymasters). But isn't there a danger that you just get stranded in a comfortable old niche with this stuff (remember that Developer Studio has excellent, programmer-friendly, CORBA support – I like CORBA and it still has its uses - with VisiBroker; although CORBA's no longer exactly fashionable)?
Well, no, I don’t think so. Not if you hear Jason Vokes waxing enthusiastic about ECO; State Machines; support for round-trip modelling, and so on. This is all welcome - you don't have to use it but it's good to have it available (I think) for when you discover that programming at higher abstraction levels has benefits. For a start, metrics, models and audits will help you sell your perfectly crafted code to your paymasters (who probably wouldn't recognise well-crafted code if they fell over it).
But here's a point. All this advanced abstraction and code analysis stuff appears to come out of Together and other technologies which are part of the New Borland. So, when DevCo is spun off, where does it get this advanced stuff from? ®
Good to get a comparative view from Robert Pointon. Yes, while I found DavidI moderately inspiring, I'm not sure that I'd find selling Delphi to my paymasters (if I was still an active developer) terribly comfortable...
I think Developer Studio would be an easier sell - but, as Robert says, the MS alternatives are awfully cheap (mind you, the initial purchase price is probably much less than the lifetime spoport cost, including the programs you write, for any development tool).
As for travelling from the North, I sympathise - I love Yorkshire and the Lakes and they're not easy to get to, relying on public transport as I do (coaches might be an option). However, I'm sure I heard DavidI or Jason talking about DevCo going outside London RSN...
I also attended the Road show and listened to David I tell us how good Devco would be for developers. I sincerely hope he is correct as I have been using Delphi form Ver.1 onward. Six years ago I started coding Delphi full time. So I have a large investment in Delphi, and the company I work for has an even greater investment.
I retire in five years so Delphi will probably see me out, maintaining old code. But can I honestly advise the company to start any new projects in Delphi? Can I advise my Grandson to learn Delphi for his collage work. When the Personal New User edition costs £857.00. Or should I tell him to lean Microsoft C# cost £00.00. Microsoft software will probably be supported by the collage using software purchased at educational prices. As he will leave Uni owing a five figure sum, and needing a job, Microsoft win hands down. Simply at this introductory level C# is cheaper and better supported.
Devco will have to address these issues if they wish to grow the Delphi community. In this country Delphi was adopted by a lot of Clipper heads because of packages that allowed the reading of ntx index’s. When we got Delphi we found an excellent product with great help file and many well written books. Where is the reason for new users to buy Delphi 2006?
We now have refractoring, a programmable editor and the history feature. Anybody remember Vi, Brief (now owned by Borland?), MultiEdit. ect, ect. We should have had this years ago. Some new stuff for formatting pages in the IDE. Welcome but not earth shattering. ECO, as David Norfolk says this works closely with Together. The future of this is uncertain. It seems to me it will be good for smallish new projects, but could get to be very complicated for larger stuff developed over several years. Will we have specialist Eco/Together people in larger companies? Eco would benefit from a book dedicated to it. Maybe Devco could sponsor one.
The current situation reminds me of the death of Dbase4. A couple of bad rushed releases, and Foxpro cleaned up. Let’s hope Devco can up its game do some top class marketing. Get some bright young developers who will give us things we don’t yet know we want, and get Delphi back to the top of the tree.
This rant sounds a lot more pessimistic than I thought it would when I started. I hope this time next year after the release of Highlander I see a brighter future for Delphi.
And while I am at it, can we have a road show up north. If David Norfolk thinks he has travel problems he should try traveling from North Dalton.