Phoenix goes mesh

Yesterday, the much-anticipated release of Phoenix was made. Version 1.6.0 1591 brings with it the ability to render mesh objects.

This means that the majority of users in SL are able to see mesh objects rendered correctly in-world, if not import them. However, the release announcement from Jessica Lyon is liable to make difficult reading for some:

“We stated some time ago our active development commitment is now focused on the Firestorm viewer and that continues today. We still feel strongly that the end of V1 functionality is an inevitability, so it is more important to develop an alternative viewer for our users they will enjoy for when that time comes than to spend our efforts on a dying viewer and then leave our users with no alternative once it’s gone. However, we also promised we would try to keep the phoenix viewer alive for you until it is no longer feasible to do so. As you can see, we are not walking away from that promise, but it is important to understand that Phoenix is no longer our top priority. When necessary we will continue to keep it up to date with advances/fixes from other third-party viewers and provide them the credit they deserve for that work. But ‘we’ are no longer actively developing Phoenix on our own steam.

“Any future releases of Phoenix will be sparse and only if needed. I will not commit to saying this is the last Phoenix Viewer Release, but I will also not commit to saying it isn’t the last either. I will say… this is one of our last. As time passes we will determine if another release is absolutely necessary and/or sensible and make a decision then on whether another update is mandatory in order to keep our promise to you.”

While it may not be a popular move, one can hardly blame Jessica and the team for taking this position: maintaining an aging code base which itself is built on something LL no longer maintain (Snowglobe) is liable to become harder and harder as time goes on, and for a Viewer to remain functional and relevant, it needs to keep pace with the evolution of the grid and as the Phoenix / Firestorm project has made the step of producing a V2/V3 hybrid, it makes sense for them to focus on that work in order to do so, rather than splitting efforts (and doubling the workload) to try and maintain two sets of code.

As well as mesh rendering, this release also brings with it:

  • The Firestorm 3.2 log-in / splash screen options
  • Contact Sets
  • Removal of the Google chat translation API options from Preferences
  • A host of “small” fixes and changes

A signficant element not updated was that of RLVa – it was decided that Kitty’s time and focus is better spent on the numerous projects with which she is already fully engaged: her own Viewer (Catznip, reviewed here), working on bringing the spell checker to Viewer 3.x, her continuing support of RLVa for other V2/V3 TPVs, and so on). In the release blog, Jessica suggest that those wishing to update to the latest RLVa implementations should give either Firestorm or Catznip a try.

In the meantime, and if you haven’t already, you can grab Phoenix 1.6.0 1591 directly, or go to the Phoenix home page and use the Quick Download links.

14 thoughts on “Phoenix goes mesh

  1. Well i downloaded Cool viewer, that has already latest Rlv!
    Soon Henry will port multiple lawers.
    More and more i see in world uisers with singularity and Cool viewer!
    So not only Phoenix team will not manage to make firestorm used for less then a minorty, but will also loose all the rest to other tpv’s.
    So LL the way they are acting now, lol!

    Like

    1. Currently, and according to figures passed on to me, the top 5 viewers in terms of overall usage are:

      1) Phoenix
      2) Firestorm
      3) Viewer 3
      4) Singularity
      5) Viewer 1.23

      Where people will swing is hard to say – but the fact that “Viewer 1.23” is in the top five was a major eye-opener to me. Whether the figure (which came to me indirectly from the meetings between LL and the TPV developers) actually refers to the de facto Viewer 1.23, or a combination of 1.23 + Snowglobe + Snowglobe-based TPVs (such as Henri’s Cool VL and Boy Lane’s Rainbow Viewer, which all perhaps report as “Viewer 1.23”) – or some combination thereof – I have no idea, but rather suspect it to be the case.

      What is interesting is that an amazing number of people, once settled on a Viewer are remarkably reluctant to change. Rianbow Viewer went for around a year without significant updates – but people kept on using it. This being the case, I do wonder how many will simply stay with Phoenix for as long as they possibly can – and that as such, we’ll still be seeing it 6 months hence, whether or not there are further updates.

      Like

      1. I forgot to mention: the reason cited for relatively high 1.23 usage is the group of people who feel uneasy about using a third party viewer but don’t want viewer 2/3 UI.

        Like

  2. What makes me mad, is not the fact that the Phoenix firestorm team will not keep developing Phoenix, is that they don’t see in world the reality, as you say most uses phoenix, and once one learns how to use a viewer, will hardly move, unless it started with v2!
    Phoenix still has some unique features that despite not having mesh made it still the most used.
    The best radar, the fact that you can uncheck highlight selected prim when on edit mode (So useful to work on small pieces like i do).
    Short cuts for bringing up the inbuild ao, to disable auto tlp and so on that makes it still unique!
    Just wander around a racing sim, to see how many are using other then phoenix!
    And the regular user, that logs in every day, does not want to loose time downloading something and wasting time!

    Like

  3. It’s strange to see someone, especially one of the TPV-agents feed more into v1, and then sit there on the verge of insulting it, despite a majority of users still using v1-based viewers.

    I know LL likes to ignore v1 like a troubled child, but they know they can’t remove its access or usability – They usually end up having some of their Torley or other Linden vlog/blog be how to make the new release viewer look like v1 >> Why LL pretends they can’t bring back the v1-style is beyond me. Not the v1 code-base, but the actual visual HUD.

    Like

  4. I have to add my two Penn’orth here. As a former user of V1RLV, Cool, Rainbow, (yes Emerald and Emergence too) and as a long-time Phoenix user, I have to say that the decision as of now to stay with a V1 codebase makes no sense at all.
    I am not a natural experimenter, and, as Inara well knows, I took a lot of persuading to migrate to Firestorm, but as of the official release I did. If one follows the installation recommendations on the Phoenix site it is a doddle to set up. Three of my friends, using very different computers are also now on Firestorm, and all report it to be a significant improvement over Phoenix, both in it’s ease of use, and in how the new render engine shows SL.

    The new Mesh facility was one I pooh-poohed a while back. I was wrong. It has some way to go before it is better than prims, but it is certainly better than I thought it would be. With Karl Stiefvater’s parametric deformer, it will be better still.

    I do not denigrate Henri Beauchamp’s determination and skill in keeping v1.23 alive, nor Boy Lane’s work with v1.22 or Siana Gearz’ team modifying Snowglobe over at Singularity. but in the end none will run Mesh and the new server systems as well as a viewer which uses the V2 codebase.

    Judging from the number of questions that looked bug- rather than incompetence- based on the Phoenix site today regarding the new Phoenix, I’d say it is distinctly buggy.
    The simple fact remains that V1 based viewers are very unlikely to get a lot better, whereas V2/3 based viewers like Firestorm will rapidly get a lot better, even from the high level of performance already achieved.

    I am a regular daily resident of SL, and if Firestorm didn’t work for me, I would not use it.

    Like

    1. Glad to hear Firestorm is working out for you :), and I couldn’t agree more with your overall statements vis-a-vis SL and Viewer progress.

      In fairness to the V1 devs, where their work will continue to have value is in the OpenSim domain, which has mesh in an SL-compatible form, but where V2/V3-based viewers largely struggle. As such, I also applaud those devs willing to put the time and effort into keeping the code going as it provides greater accessibility for those wishing to try OS. Henri has also done marvels in brining so much capability to those who are unable to upgrade to more recent, high performance systems.

      Like

      1. No, well I am not competent to comment on other Grids since I haven’t used them since 2009.

        If such viewers are used for other grids, then their validity is undisputed, and I have nothing to say in that area.

        @foneco zuzu:
        I use a GTX580 and I have no issues. You are not specific so I cannot be detailed in my response.

        Like

      2. Uhm, why exactly should V1 viewer be better suited for OpenSim? I use LL Beta 3.x as well as Firestorm with the Hypergrid, adding meshes to my Sim and so on. It may be part of the cultur to go as GPL as one can and thus prefering Imprudence (and its import / export features) but other than that?

        Like

        1. For the reason I stated: many report the V2/V3 Viewers + TPV struggle on OS grids. Not all OS grids by any means, but the issue exists. I’ve found the same myself (using Firestorm, Dolphin, and V3.0.x through V3.2.x) on the several OS grids I tend to visit, whereas the likes of Imprudence, Cool VL, Rainbow have given a far smoother experience.

          Work is underway within several V3 TPVs to stablise performance on OS grids, but some continue to caveat their use on grids beyond SL.

          Like

Comments are closed.