[Raion]

So besides Nekoware II, my own projects that are under wraps, I wanted to get some community feedback on what exactly do you think is important as a focus?

I would hope everyone taking the poll would also post below elaborating on why that choice was made. 

Anyways, explanation of options:

1. Improving the IRIX experience

This focuses on areas of usability such as the IRIX tar wrapper project. Things that would affect the look, feel or usability of the base system. This would probably fall under IRIX Community Edition stuff if it gets picked up. 

2. Opening up the IRIX codebase.

By this one, I mean finding, modifying or recreating portions of the IRIX operating system as free, open source software. HPE already isn't interested in releasing the code under a FOSS license of any kind, so any opening of the code would need to be either grabbed from other OSes, modified from other projects, or made from scratch. 

3. Hardware/driver support.

As mentioned before on many topics, the IRIX USB and Firewire stacks are not very good. Other improvements like filesystem drivers, backporting XFS changes to IRIX, or other drivers would fall under here. 

4. LLVM/Clang. 

I see this one mentioned a lot, so if we have advocates of LLVM, please stand up. 

5. A new browser

I've spoken at length about this before, but if I left it off the poll y'all would have my heads. 

6. Tools or utilities to ease development

7. Better documentation

8. Other

[commodorejohn]

* Replacement of "focuses" with "foci?"

[Raion]

It seems both are correct ways of saying it. I know Latin, but I tend to use English plural forms...

[Raion]

I appreciate the votes guys; if I could get some feedback on __What__ exactly you want or have ideas about, that'd be helpful!

[hamei]

* Replacement of "focuses" with "foci?"

Locuses of concentra ... aw, shoot

cf Raion's questions ...

1) Is okay where needed, but most of the "Irix Experience" is fine, imo. It's Irix. If you want it to be Loonix or OS/2, then run Loonix or OS/2. Or Solaris or whatever. The fun of it is, it is what it is.

Of course stuff that does make life easier is nice but I wouldn't go overboard on that.

2) Too much work. If you want it to be "Irix but different" then just start from another OS, one that is already faster and easier to deal with.

3) Yes for sure. A few basic things we all take for granted now that SGI did badly, would make life much easier.  Finding a SCSI reader for an SD card is well-nigh impossible. My firewire one works good but those are rare as hen's teeth. A USB driver for Wacom tablets ? Wow.

4) Clang would be nice I guess but MIPSPro works fine, for me most of what you need a newer compiler for shouldn't be done. GTK3, for example. 

btw, 'mentioned a lot' doesn't mean anyone would use it even if you got it done. Do you have any idea how hard I tried to get people to just download and test software that people had built for them ? Getting three people to test a 1 meg simple program was well-nigh impossible. Don't kid yoruself about how many people would leap at the chance to port or develop for Irix. The number is about three, and they are already doing it.

5) Too big. It *might* be interesting to take the last one that diegel got working and strip all the shit out of it. Fireflop does a ton of behind the scenes worthless garbage that has to slow down thee elderly cpu's. Removing that crap might make the flop a bit quicker.

6) Already have Rapidapp, probably not going to come up with anything better. Enticing Shade of Blue, joerg, jp here, j-j, dexter1 and others to answer questions is likely to have more real-world effect.

7) Already have TONS of documentaion. RTFM ...

The stuff that was learned the hard way over many years on unowhere, what can I say ? Nothing good ....

8) My prescription   -- more useful software and more people excited to use it. More Irinikuses (Iriniki ?) Unless you are strictly a museum, the fun of an old computer is using it to do something. The chopper guy is a good example for us all.

For the people who like programming, I personally think stuff like backporting the Gimp to Motif would be great  

Motif is kind of a bitch, I played with it a little and it can drive you crazy but it IS the toolkit for Irix.  So Motifyfying useful programs is cool. (It's not impossible, either. One guy did a Motif-to-GTK conversion of Grace in a couple months, by himself. That's the wrong way but demos the possibility.)

Failing that, gtk1 isn't as awful. Backing out gtk2 and 3 to 1 might be easier than going to Motif in some cases ?

Wider use of programs like Eclipse would be good to encourage, but if a guy is already a Gimper, a Motif Gimp would be spiff.

MPlayer is sooooo useful. Axatax is currently The Man (and prior to him the five or six guys who mipsed the code, super.) That kind of project is wonderful. Somethign useful that runs well on SGI boxes. Nice if it runs on an Indigo2 or lesser but even if it's just good for the newer stuff, still an asset.

A little utility to use cdrecord, similar to XCDRoast but well-thought-out, that'd be useful. Maybe just a quick-n-dirty graphical app to create a command line for that purpose.

Lots of little daily stuff that people can use .... that'd be my goal. Small projects that can actually get done and used. Stuff that means instead of being a box under the desk, people use these computers.

[jan-jaap]

Motif is kind of a bitch, I played with it a little and it can drive you crazy but it IS the toolkit for Irix. 

Plain Motif is tedious, but you can use RapidApp or BXpro to point-and-click your interface together. Then you just need to implement the callbacks to do the "real" work, and not worry about resize policies, layouts, keeping the right window on top etc.

[mgtremaine]

I think first step is to "Improve the IRIX experience" , IE update the really clunky old parts as some people are working on. Neko II or whatever we can get as package manager would be great.

Once that is sorted I would think that drivers and codebases, and compilers and etc will be a lot easier. We just really got clear out some of the cobwebs on these installs and have a more solid "modern" foundation to build from. (I have a feeling I'll be doing a fresh install before the end of the year, I have not even started to try and remove python 2.4 and openssl).

-Mike

[Raion]

I think first step is to "Improve the IRIX experience" , IE update the really clunky old parts as some people are working on. Neko II or whatever we can get as package manager would be great.

Once that is sorted I would think that drivers and codebases, and compilers and etc will be a lot easier. We just really got clear out some of the cobwebs on these installs and have a more solid "modern" foundation to build from.  (I have a feeling I'll be doing a fresh install before the end of the year, I have not even started to try and remove python 2.4 and openssl).

-Mike

I sincerely appreciate your feedback. This is in addition to Nekoware II, but I absolutely agree with and see your point. The main goal is to keep IRIX, IRIX, but also to allow for better user experiences. That tar wrapper is a nice step, as is the new inst bounty (IF SOMEONE WOULD TAKE IT UP LOL) but we'll see.

[hamei]

Plain Motif is tedious, but you can use RapidApp or BXpro to point-and-click your interface together. Then you just need to implement the callbacks to do the "real" work, and not worry about resize policies, layouts, keeping the right window on top etc.

RapidApp seems okay if you are building your own program ... I was trying to modify an older one; trying to figure out what the heck was going on was a headache. But kind of fun when it worked

The one I'd love to see Motiffed is Fontforge. The interface is absolutely ghastly, and it's not even a "well, it's already gtk so why not just get that to work" kind of thing. It would be a good project, and nice to have on Irix. Artistic and stuff :-)