Mailing List Archive: 49091 messages
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search
 

ANN: New Rugby release + website

 [1/3] from: m::koopmans2::chello::nl at: 8-Feb-2002 22:42


Hi All, There is a new Rugby release + website , although minimalistic ;-) I have added lots of code, and finally removed even more. What is Rugby now? What it was, but only with http transport, that uses compression. Improved marshaling, security bugs fixed. I invite you all to use it and test it, if you find bugs, report them to me. I'll fix it asap. Within a week or so I will send a standalone version of RebXR to Andreas, as XML-RPC has disappeared as well. In the end I tossed out a lot of goodies that I said would be in the next version, which I called XPi. Why not? As it turns out you can do everything (well, most of it) based on the current core. So the advice well taken was 'keep the core simple'. If you think you get less, you are *right*, but less is more. A lot of redundant code has disappeared, and the thing is much leaner now. For programmers that tends to be a good thing. Start using it! That's what I'll do the next few months.... Another reason not to put too much effort in some features is that REBOL is still young and evolving. I expect the next releases of REBOL to incorporate changes in the area of networking, serializing values etc. I think I added short but complete documentation as well. And the license is BSD, as always. You can find it all on http://www.rebolforces.com/~erebol/ http://www.vrijheid.net/ Break the chains, make the change Maarten

 [2/3] from: petr:krenzelok:trz:cz at: 9-Feb-2002 3:22


Maarten Koopmans wrote:
>Hi All, >There is a new Rugby release + website , although minimalistic ;-)
<<quoted lines omitted: 8>>
>In the end I tossed out a lot of goodies that I said would be in the next >version, which I called XPi.
So you don't call current version XPi?
> Why not? As it turns out you can do everything >(well, most of it) based on the current core. So the advice well taken was >'keep the core simple'. If you think you get less, you are *right*, but >less is more. A lot of redundant code has disappeared, and the thing is much >leaner now. For programmers that tends to be a good thing. > >Start using it! That's what I'll do the next few months.... >
I like the doc coming with Rugby distro. Very clean, understandable, short and good for starting to use rugby. btw: I would make Rugby even easier by using: sexec secure-serve get-rugby-service/secure-code secure-result-available? using: rexec/secure serve/secure get-rugby-service/secure result-available?/secure It would be even cleaner, easier to remember, powerfull, - rebolish. - just one refinement to remember, and the same function names to use ... Quite interesting part is: NOTE: In REBOL/COMMAND 2.0 rsa-generate-key does sometimes not function. As a side effect you cannot use secure transport with Command 2.0 In that case, ask Rebol Technologies for a different build. Is that confirmed /Command 2.0 has any such bug?
>Another reason not to put too much effort in some features is that REBOL is >still young and evolving. I expect the next releases of REBOL to incorporate >changes in the area of networking, serializing values etc. >
That's probably OK. Has current rewrite any speed difference in comparison to "old" Rugby? Also - what is, and what is not implemented? Profiler? Functions for adding, removing Rugby functions, exec-code? Adding a module system, module exchange and migration , etc? Thanks, -pekr-

 [3/3] from: petr:krenzelok:trz:cz at: 9-Feb-2002 4:05


Few other things: I found inconsistent to use: serve/with .... tcp://:9005 and then - rexec/with .... http://localhost:9005 tcp vs http - serve/with should use http://:9005 too, to look more consistent ... What do you think? As for performance - I am a little bit disappointed. I really don't know what is wrong with W9x stack in cz Win distros, but 100 echos now take exactly one minute, while old rugby consumed 44 secs. The time is consistent - nearly 30% slowdown is a little bit too much to pay for compressed http (if it is the reason of a slowdown) - or is slowdown caused by architecture change? What is your speed experience when comparing XPi to Rugby 4.3? Thanks, -pekr-

Notes
  • Quoted lines have been omitted from some messages.
    View the message alone to see the lines that have been omitted