Icesus from native 32 bit to 64 bit View next topic
View previous topic
Post new topicReply to topic
Author Message
cyto
Site Admin

Posts: 20

View user's profile
PostPosted: Mon Mar 18, 2019 18:44 pm Reply with quoteBack to top

grinne Cyto, if you were to write new icesus, what would you use?
cyto grinne - can you specify - do you mean driver/tech, content maybe?
grinne driver/tech
cyto ok, grinne. if we want to entertain the topic further, let me gather an answer for driver tech.
cyto there would be basically three viable options on the table (everything else is IMO just crap): 1) FluffOS 2/3.x - a continuation to mudos (on which icesus is currently running). this is a native 64-bit driver. pros: mudos->fluffos port would be relatively easy. cons: this would probably be my last option because it is not really maintained right now and there have been complaints of some big problems with the 3.x driver that are unresolved (like, they changed from C to C++ and are now suffering the consequences). 2.x releases are of much better quality.
cyto 2) ldmud - this would be the safest choice of them all. this is afaik the same driver on top of which e.g. hugely popular batmud is running on. pros: good, decent, nice finish, nice polish. very stable future. cons: porting icelib content will require more work as driver on the bottom will change, so baby steps, baby steps.
cyto 3) DGD - option of the bold. the choice of the 'alt' people. it puts a lot of effort into persistence of things, so this is the driver that would not require reboots, has state dumps so mud is basically treated as one big VM that you can snapshot etc. lot of new, daring, spiffy features. another big pro is that it can be licenced commercially, which either options 1 and 2 cant do. so basically you could hire a team of devs to built your mud, if you had the money (which you can raise). so base potential is enormous. cons: bold choice means just that, a LOT of things to port, for example what would happen if reboots went away NOW? a lot of things would be broken (like eq mob repops, etc.)
cyto so there you go, these would be the three choices on the table regarding mud driver tech.
grinne so still going with 'old' tech then? everything lpc-based iirc
cyto yes. 'old tech' in the sense that 'mud is a mud'. for example, i would not involve anything that even remotely resembles attempts to make things 3D. Smile
bandar nice to have a wiz active and sharing his thoughts around... sincerest thanks...
grinne mud is a mud of course, but there are drivers nowadays that can be coded in e.g. python, which would mean there is bigger pool of potential coders
cyto bandar, yeah - i'm trying to sort of share with you also some thoughts about the 'big picture' -- so to let you know that there definitely is one, but we won't be making choices hastily and especially without asking.
cyto there is a mud driver called 'evennia' which does exactly that, it can be coded using python and it links seamlessly to the driver, also sql database of some sort. the problem with it is it is not LP style. it is a MUX (more role playing oriented school of thought in mudding). evennia is quite mature, but i'm afraid it is not suitable due to it's MUX nature.
grinne yeah, i know about evennia
cyto but as i said - i'm just throwing the bone to you guys, feel free to contribute, even at level of thought. i'll be happy to challenge and throw some ideas back in return.
cyto then there is of course, the option 4: we remain "status quo" as it is. we trust in the holy spirit, keep our mud in a neat VM bubble and forget about 'far future'. pros: the game is ready. it really doesn't need this fancy stuff, no need for major conversions, steady as she goes. cons: laws of probability state, with anything that is compiled - the Day of SEGFAULT is coming, e.g. in year 2029, after reboot, mud just won't go up again. we get a mystic number and have absolutely no idea what to do because the driver has been dead for a dozen years.
cyto i'm joking again. Smile
grinne are you though :0
crai Do not worry. The Russians will come earlier than that!
cyto you're probably right.
ezenya i didnt even follow the conversation, but i still stand by the building a mirror icesus with new drivers and all that, and migrate gradually everything from here to there
grinne i'm looking at the dgd driver, looks great
cyto so if we manage to do some sort of 'body swerve' now and prove that our lib is even remotely mobile, we might be spared from the inevitable 'big crunch' in the future, that is the Day of the SEGFAULT. }Smile
crai you post on website that some mages were inventing a new spell and blew the world up!
cyto but thanks to modern VM technology the day of segfault is probably a hundred years away - just as they can still use floppy disks in pentagon. so don't sell your eq just yet. Smile
khade :Khade waves around a sign with The End Is Near printed on it
khade just call it retro
cyto could work, could work.
linran I thought 2029 is the year we should get first neuro link sets running
cyto so that's what i wanted to say, a little heads up of how things could look like in our horizon.
cyto
Site Admin

Posts: 20

View user's profile
PostPosted: Mon Apr 22, 2019 15:22 pm Reply with quoteBack to top

The choice for new driver has been made: FluffOS 2.

Image
Picture: Icelib 1.0 launching successfully on FluffOS 2.

As seen in above picture, if all goes as planned, no branch nor fork of Icesus is needed for migration to 64 bit - the icelib is good "as is". This would be optimal.

More later.
cyto
Site Admin

Posts: 20

View user's profile
PostPosted: Thu Apr 25, 2019 19:11 pm Reply with quoteBack to top

Significant advances on the 64-bit front.

No "foreign body refusal reaction" as FluffOS2 was merged with Icelib -> The operation now has ~75% chance of success.

Image
Picture: Lonely at the Ere of the "upside down world".

Image
Picture: Steady as she goes.
cyto
Site Admin

Posts: 20

View user's profile
PostPosted: Wed May 08, 2019 17:21 pm Reply with quoteBack to top

http://naga.icesus.org/ is the new home for 64-bit Icesus.

This new "shard" of icesus is not open for public so you can't login yet.
cyto
Site Admin

Posts: 20

View user's profile
PostPosted: Sat May 11, 2019 15:03 pm Reply with quoteBack to top

Image
cyto
Site Admin

Posts: 20

View user's profile
PostPosted: Tue May 14, 2019 15:04 pm Reply with quoteBack to top

Icesus64 is currently in alpha testing stage. If all goes well, beta testing phase will open in June 2019.

Beta testing phase means all players are welcome but nothing (exp/eq) will be saved. PDF can be earned for spotting juicy bugs in 64.

Towards end of July 2019, we go from beta testing phase to live (version 1.0). Once we hit 1.0, the current 32-bit server will go down and only Icesus64 remains.
Display posts from previous:      
Post new topicReply to topic
View next topic
View previous topic

Navigation


Powered by phpBB | All times are GMT