Borland Linux Poll: Take Two
A fair number of you probably voted in the recent poll about Borland exploring Linux development. As the other article touched on, they are looking to hire people, and are running a poll to test the waters. However, the prior poll's server got slashdotted, and they missed a lot of people's comments. If you didn't make it at first, head over again and make your voice heard.
The fact that their server got slashdotted should tell them something about the demand. Heres to C++ Builder and Delphi on Linux!
Actually, they want everyone to come over and vote again, they're not combining the 2 polls results (at least, according to the e-mail seen on Linux Today).
I received a cold call from somebody at Inprise
(and why are we still trying to call it borland?)
She seemed rather perplexed and annoyed that I
was only interested in products for Linux and Digital Unix, and was very not interested in talking about anything else. Could I have had
an effect?
-fb Everything not expressly forbidden is now mandatory.
I answered, "Open Source Makes no difference to me" on the questions about the importance of GPL. I hope they port their products and sell them and make money on the Linux platform. There are a lot of programmers out there like me that don't care about GPL or "Free" software, but use it because its a cool environment and it provides a level playing field that isn't inundated and overwhelmingly dominated by Microsoft.
This was my favorite tool in Windows, I hope they port it to Linux (although a complete rewrite would make it a lot nicer, ports with Winelib are ugly and awkward).
...some rational people are voting, too.
How do you think the playing field got so level, if not for Free Software?
---
Put Hemos through English 101!
Linux MAPI Server!
http://www.openone.com/software/MailOne/
(Exchange Migration HOWTO coming soon)
How many millions and millions of resumes do companies who so much as imply the tiniest, slightest, smallest, miniscule chance of a job opening get hit with when they appear on Slashdot? Or is it billions of resumes. You could see Borland's mail server explode from 3000 miles away.
How would it explode isn't there a bandwidth limitation. That should keep the box from exploding verry easily.
The death of one man is a tragedy; the death of a million is a statistic --Joseph Stalin
At this rate, it isn't going to take long for companies to realize that hiring Linux developers is easy, and people will buy your product in hordes. If Borland plays their cards right, they could very well end up with an incredible market share rebound.
------------------------------
1. Didn't /. run a story about how Borland was getting rid of OWL and become another MFC compiler? Doesn't that mean that Borland/Inprise is going in too many directions at once? As much as I want Delphi and Jbuilder, I want it to be *quality*. I don't want the port to suffer because too many of the staff were diddling with MFC garbage.
2. The IDE must let me use whatever editor I want. Once the rest of the world realizes this their products will sell (in the Linux sphere). Likewise, the compiler suite must mesh with whatever I have, ie useful command line tools.
3. It must work out of the box, and not be crippled of useful functionality. Right now most Windows IDE's are slanted toward Windows functionality, obviously. If all I got was a toolbar and some graphics libraries, well, no thank you. I want things on that CD that will do for me, a Linux developer, what MFC does for the Windows guys.
Wait. Maybe I should rephrase that.
I want it to really improve on the coding, I want it to really do some of the work for me, I want it to help me develop better software. Thats a tall order as I see it.
ZOMG I WOULD LOVE TO KNOW ABOUT YOUR FEELINGS ON MACINTOSH VERSUS WINDOWS, VI VERSUS EMACS, AND HOW YOU'RE NOT A DORK
I used to use Turbo C to write DOS apps all the time. It was quick and had a spiffy IDE. Of course, that's the only Borland product I've ever used, I dunno if they still make 'em like that.
If they decide to do anything in the Turbo C range (i.e. cheap, but it does what I want it to) for Linux, I'll certainly buy a copy.
F0 07 C7 C8
They have re-renamed the development division back to Borland. In honor of the great Al Borland of Home Improvement. (Well, maybe not that last part :) )
Posted by OGL:
Any word on when they're going to publish the results of this? Some of the questions were really interesting (for example, what % of Linux developers do GUI work, and what % of those use GNOME or KDE?). But as for Borland porting their products...I don't know. I sure as hell would never use Delphi, and an IDE that uses the GNU tools seems like it would be pretty easy just to write (I did it myself in Windows for DJGPP, in my pre-Linux days).
-W.W.
Didn't a while ago Microsoft had invested $100M to its parent company, Inprise? [1] Wouldn't such investment prevents Borland to support Linux, or forcing Borland to compromise Linux supports? With marketshare in stake, M$ might wrote some hideous conditions that will allow the company to take away the invested resource (or worse) if Inprise ever dared to support Linux or anything Bill perceives as a competitive threat.
----
[1] http://www.news.com/News/Item/0,4,37501,00.html
The $125M that Inprise got from Microsoft (plus patent cross-licensing agreement) was completely no strings attached. Inprise was close to going under, but they had a last trick up their sleeve, a software patent for GUI builders.
The patent covers automagically being able to work with the gui designer view or the source code view and having the changes be reflected immediately in the other view.
It doesn't matter what you think about software patents, what matters is that MS Visual C++ and Visual Basic were in violation of the patent. Instead of fighting it out in the courts, MS just paid out, what for them was a trivial sum.
So, yes Inprise got a bunch of money from Microsoft, but it is so completely strings-free that we needn't worry about it having any impact on Borland tools for Linux.
I've been a Delphi user since the 1.0 version. However, all of my development work is now in Linux. I've had difficulty adjusting to building GUI applications on this newer (and much better) OS.
I've noticed the comments many people have made about ObjectPascal as a language. Pascal, and particularly Borland's implementation in Delphi, leaves a lot to be desired from an OO perspective. I suggest that Borland take another fork in the road a bit and develop an IDE that has Python as its underlying language. Admittedly, Python is a scripting language, but is is running very well on Linux and Windows. Furthermore, using Python would relieve Borland of porting ObjectPascal immediately. Moreover, Python is easy to learn and it works well in both Linux and Windows -- making the transition phase from Windows to Linux that much easier.
I also suggest that the things that most novice and "corporate power users" liked about Delphi was that the IDE relieved the developer from a lot of tedious overhead issues that are necessary for making programs run. All Borland needs to do is apply their tried and true formula (i.e., making it easy to write good programs quickly) and simply port that over to Linux. No matter what language they use, if Borland puts a familiar façade in front of all those Windows developers, hide the gory details (heresy, I know) and make that functionality work in Linux, Borland will make money and the Linux world will be better for it.
Simply creating an IDE for python would be less work than porting ObjectPascal (with all of its concessions to the Windows OS). Thus, Borland could make its corporate splash onto the Linux scene quickly and take the pressure off of getting a quick Delphi port out the door. Hopefully, Borland would take the opportunity to clean up ObjectPascal and make it a truly object oriented language (particularly in the area of object collections, TCollection sucks).
Hey, maybe I just don't understand....
What would it take to make an Open IDE, with pluggable compilers? Then, pick your editor, pick your language and pick your source code control. I program in multiple languages and this would make life a whole lot easier!
Hey, if this already exists, someone please let me know where I can get it.
RHIDE is also available for Linux!
I found a URL for y'all:
http://www.tu-chemnit z.de/~sho/rho/rhide-1.4/rhide-linux.html
I think it's one of those apps that's in a perpetual beta stage [kinda like ICQ for winbloze...]
I used RHIDE it when I was still looking through Windows. I was happy when I discovered I could use it in the Open [source that is]. It came with my commercial copy of the SuSE[5.3 or 6.0 or 6.1- I canna remember] distro...
HTH,
Steelehead
-- 100% MS-Free as of 4-4-1999, 11:47:38 PST. "The lapdance is always better when the stripper is cryin'" Free Kevin,
Or do it for Perl too...
--
An esoteric scratched itch:
Homeworld Map Maker Tool
Why did the first developer survey fail?
Were they using CGI or some appserver?
Did they throw more hardware at it?
Just curious.
I use JBuilder 3 (Enterprise or Client-Server) at work and for a $2.5 grand product there are an awful lot of bugs, including some that are just plain lazy and sloppy programming (remember Borland C++ 5???). Even Micro$oft manages to start making their software useable by version 3 (I'm sure that the useable J++ would be more popular if it just stuck to the Java spec). JBuilder drives me insane... thank goodness for Emacs! Before porting to Linux, may I suggest fixing the bugs before introducing tonnes more.
On a side-note, isn't JBuilder partly implemented in Java... wouldn't that make porting easier?
Borland is obsolete. Their C compiler was used initially because that's what Minix was compiled with. But that lasted only until Linux was good enough to support a real compiler.
Two problems. The CGI for the first poll was written without any kind of real traffic in mind. It did a lot of unnecessary DB processing to generate the HTML and also did some post processing of the survey submissions. It basically was not efficient at all. It worked fine if only a couple of people were using it at a time, but throw a few hundred at it at once and things got pretty messy. Over the weekend it was rewritten to be an order of magnitude more efficient.
Second, the hardware it was running on was sort of a "multipurpose" server that gets used for miscellaneous projects. It was not part of the official I.S. system. The server happened to be running about 70 other processes for others as well. - no elbow room
We ended up getting new hardware and fixing the survey software, but in the mean time we had rehosted the survey with, Infopoll, an outside company that specializes in web surveys to keep the survey running. It was working well so we stayed with Infopoll rather than switch back to our original "now-working properly" system.
That's my story and I'm sticking to it.
-Michael
They come with RedHat. Isn't that what you're looking for?
I learned to program C with a pirated copy of Turbo C 2.01.
(I was 13, I had no money.) One of the first things that I bought once I had a job was Turbo C 2.01. The manuals alone were worth the price (like $75).
I agree with the comment. I'd be happy to have something the quality even that IDE (1988), and more importantly, that debugger. I can't stand gdb.
You can argue with me all you want regarding gdb, but, for me, it's the biggest reason I still develop for Windows. I don't care for most of MS's products, but their debugger is great, even if it has a couple of features that drive me up the wall.
Go Borland! I can't wait to see Turbo/Borland C++ for Linux. But make it good. Borland C++ 4.x was one of the worst pieces of software ever created. TC++/BC++ 3.1 rocked!
I bought at the time Borland C++ 3.10 with Application Frameworks even though its cost was $500. These are two and a half my monthly salaries now. This was in fact almost the only software that I ever bought. You can quite easily guess why.
The coolest thing in it, apart from the fact that I had printed documentation, was that its libraries were available in source. I am a self-made programmer, and not only that but I also never had a chance to work with people from which to learn and this was my first encounter with professionally written source code. I'll never forget those days.
Let's forget the price now. I wouldn't buy their software again because with Linux I have access even to the compiler's source code (thanks to GNU of course). I like companies like Borland and Corel and I wish them luck in beating the Redmond giant but their chances are slim. I believe that the days of mass appeal proprietary software are numbered. One exception of course is games. No matter how many free games there are, the demand for more won't decrease. They are like books or movies: the good ones only increase the appetite for more.
I hope they do port Delphi and C++ Builder over. But lets not forget one very important point.
Aside from JBuilder, Delphi and Borland C++ are Intel only compilers. If they are going to port to Linux they should port to the other prevailing CPU architectures (Alpha and PowerPC). Because of their limited finances they may not. If they don't port, at best they are offering only a partial solution. (Or not one at all).
On the other hand....They may if someone from Compaq and Motorola is paying attention and is willing to finance the project.
For those of you who are excited about Borland tools on Linux, add the request to port to other CPU's in any available comment fields in the poll.
Thanks for the candid explanation.
We too have had trouble with a CGI/Entera solution. Too slow and inefficient.
Why did you not use your Appserver product for the poll?
Please no more useless features - just fix the dozens of crash bugs and assorted problems with JBuilder.
It's a shame that Supercede never succeeded
(and was bought out by some no-name company and left for dead). It still remains the fastest and most stable Java development environment to date. It *never* crashed. You can't say that of its competitors. Marketing and being first to market is all that matters these days, I guess. Proof: Symantec Visual Cafe's success - yuk.
enquiring minds want to know
Not to belittle this post, but you made $1.25 an hour?