In the tallbacks there were some negative things about lazarus. Most were from commercial developpers that can afford to buy kylix and have to deal with time to market.

And there was also some confusing that lazarus was started after kylix was announced. Maybe it's better next time to include a small line like: lazarus has been founded in spring 1999. then it's clear that it was started before kylix was announced for the first time.

For distributions? Yes. But distributing GPL'ed programs only will make money for a few people. (the real big projects that can sell support), and they are not enough to recover Kylix developping-costs, let alone allow Borland to make a profit.

The professional programmers of any size will have to buy Kylix, and they will have to distribute their programs, and charge for it.

And a lot of them will have no problems (serverside apps mostly run on relatively unpatched RH servers), but for the rest (that will write "consumer" apps, or clients) will have to be more flexible.

I think also when the first larger projects get deployment problems to other distributions or modified (upgraded by hand) standard distributions.

Codepunk wrote

Michael did it appear on /. I did not see it?

No. I didn't make a specific entry there. I tried on other time and they just ignored my post so I didn't bother this time. I just did a press release which is picked up by most of the news sites. That gets it out fairly well. I am also going to add a release notice to freshmeat as well. I'll let the last rush die down a little first. :-)

Distributing applications on Linux will save "many" people money. The idea for companies is that they can develop their applications on Linux boxes and save purchasing Windows NT/95/98 OS and Client Access Licenses. Not to mention the baby sitting these systems need.

I agree that people like myself and other developers may NOT be able to recover the develpment costs unless you end up getting business from customers by telling them that you can reduce their support costs because they will be using Linux. Granted, Linux will be used in most cases on the server side and not the client, but this is what needs to be pushed. If you write an application for a company that will be running this app on a PC with little to nothing else on it then you could have them use Linux and develop in Kylix.

In most cases however, this will not yet work. Most companies are simply too attached to the Windows environment.

Shane

For distributions? Yes. But distributing GPL'ed programs only will make money for a few people. (the real big projects that can sell support), and they are not enough to recover Kylix developping-costs, let alone allow Borland to make a profit.

The professional programmers of any size will have to buy Kylix, and they will have to distribute their programs, and charge for it.

And a lot of them will have no problems (serverside apps mostly run on relatively unpatched RH servers), but for the rest (that will write "consumer" apps, or clients) will have to be more flexible.

_________________________________________________________________ To unsubscribe: mail with
"unsubscribe" as the Subject
archives at http://www.miraclec.com/list_archives/lazarus

I _seriously_ doubt that. Try deploying an app that needs to access 2 or 3 databases, and you'll see what I mean. Total hell, especially now with all the windows version floating around these days. From Win95A,Win95B Win95B+USB support WinNT 4.0 (SP 1-6)(Workstation/Server) Win2000 to Win ME. Our helpdesk here have become experts in determining exactly what version of windows is running even when the other side of the phone is totally moronic and can't even select a file in the explorer.

Just to say that I don't think the problems will be worse, but there will be enough problems to warrant a decent setup tool on linux. RPM by itself doesn't qualify in this case, IMHO.

Michael.

I think this is no reason. One of my main arguments is: do you know when Borland gives up Kylix, then your lost. It could be happen very fast, for example Corel comes in mind.

FYI, the FCL was started in Summer '98 by Michael V. C. and me. FPC for Linux was created in Summer '96 by Michael V. C. thus FPC/Lazarus for Linux was much earlier started than Kylix.

Moreover, the buggyness of Kylix atm, will spawn some fix-versions and soon new versions, which will increase different sets of libs, or libs with slight incompabilities with eachother.

The total of these things will create something to which the windows DLL-hell will be a non-issue.

Michael Van Canneyt wrote

Just to say that I don't think the problems will be worse, but there will be enough problems to warrant a decent setup tool on linux. RPM by itself doesn't qualify in this case, IMHO.

But one thing comes to mind: As the CLX is dual-licensed under GPL, it should be possible for distributions to deliver [RPM|deb|whatever...] packages?
Only the Kylix IDE itself will be a problem, as it cannot be rebuilt from sources by everybody. But GPL'ed applications could be...

- Sebastian

mooseman wrote

how well did your server handle things?

It was doing OK until this morning. When I got up and checked all of the various servers had crashed. The FTP was the main one to crash. It just got over loaded.

the pipe must have been busy at certain times.

I really notice it when I am going out to get something myself. On top of that my ISP is having SERIOUS router problems. There is a real bad delay and hold up coming in through one of UUNets routers and it is making it bad for everyone on my ISP. This can also be causing some people problems with getting stuff from the site. Hopefully they can get some more pipes into the ISP in the next couple of weeks.

Peter Vreman wrote

In the tallbacks there were some negative things about lazarus. Most were from commercial developpers that can afford to buy kylix and have to deal with time to market.

Yeah I read those too. Kinda of discouraging. Yeah well see them squawk when they are asked to port their product to OS/2 or FreeBSD or Solaris, etc.

And there was also some confusing that lazarus was started after kylix was announced. Maybe it's better next time to include a small line like: lazarus has been founded in spring 1999. then it's clear that it was started before kylix was announced for the first time.

I was going to defend it but then saw that several other people came up to the plate and defended our start date. However, I will most definitly mention when it started next time.

Here is the web server log for Lazarus and the ftp log for my whole host.

http://www.lazarus.freepascal.org/usage/usage_200102.html http://www.miraclec.com/usage/ftp/usage_200102.html

Michael did it appear on /. I did not see it?

Cliff

wow. a bit of an increase. ;-)
lots of front page hits. couple of first level down hits too. interesting country breakdown....

how well did your server handle things? the pipe must have been busy at certain times.

jim.

you wrote

In the tallbacks there were some negative things about lazarus. Most were from commercial developpers that can afford to buy kylix and have to deal with time to market.

Yes, that was amazing. Some people seemed to think that Borland will stop developing products for Linux if somebody tries to write a free alternative.

There is not a single productive programming environment for Linux yet and Lazarus was the first real effort to create such an environment. What comes out of Kylix is yet to be seen, the price is a lot higher than predicted, even for us Borland users.

Markku

Yes, this is one of the things I wrote in my article about Kylix. There is no setup tool � la 'installshield' or 'wise'; given the 20+ libraries that are needed to run a Kylix application, this will give some nice setup problems. Especially since they require specially patched linux loaders and glibc versions.

Michael.