It’s been a great 3 years! All my thanks to my colleagues at Linden Lab and our wonderful customers I wish you the very best for the future and continued success! I am starting-up a company to make Art, Entertainment and unusual things! More on that in a few weeks!
With these words, and a few personal notes to the likes of Jo Yardley, who broke the news to the SL community as a whole, Rod Humble’s departure from Linden Lab entered the public domain.
Rod Humble officially joined the Lab as CEO in early January 2011, although according to BK Linden, he had been logging-in during the closing months of 2010, “exploring and experimenting in-world to familiarise himself with the pluses and minuses of our product and the successes and challenges faced by our Residents”.
Prior to his arrival, and under the much maligned Mark Kingdon, the Lab had been investing in hardware and infrastructure, with Frank (FJ Linden) Ambrose being recruited into the company to head-up the work. This continued through the first year of Humble’s tenure as CEO, paving the way for a series of large-scale overhauls to the platform in an attempt to improve performance, stability, reliability of server / viewer communications and boost the overall user experience.
Much of this work initially announced in 2012 as “Project Shining”. It had been hoped within the Lab that the work would be completed within 12 months; however, so complex has it proven to be that even now, more that 18 months later, elements of core parts of it (viewer-side updates related to interest lists, the mesh-related HTTP work, final SSA updates) have yet to be fully deployed.
Even so, this work has led to significant improvements in the platform, many of which can be built upon (as with the HTTP updates paving the way for HTTP pipelining or the SSA work already generating core improvements to the inventory system’s robustness via the AIS v3 work).
It might be argued that these aren’t really achievements on Humble’s part, but rather things the company should have been doing as a matter of course. True enough; but the fact is, prior to Humble’s arrival, the work wasn’t being done with anything like the focus we’ve seen under his leadership.
A philosophy he brought to the Lab was that of rapid development / deployment cycles, as he indicated at his first (and only, as it turned out) SLCC address in 2011. This saw the server release process overhauled and the three RC channels introduced, making it easier to deploy updates, patches, and fixes to address bugs, issues and exploits.
Humble referred to this as “putting the ‘Lab’ back into Linden Lab”, and in fairness, it didn’t always work as advertised, as with the initial experience tools deployment in June 2012, which resulted in a spate of grid-wide griefing. However, it is fair to say it has generally resulted in less grid-wide disruption and upset.
More recently, this approach has also been applied to the viewer release process, allowing the Lab to focus more sharply on issues arising within the viewer code as a result of changes or integrating new capabilities. This in turn has largely eliminated the risk of issues bringing viewer updates to a complete halt, as happened in the latter part of 2012.
One of the more (to many SL users and observers) controversial aspects of Humble’s tenure was the move to diversify the company’s product brief. When talking to Giant Bomb’s Patrick Klepek in October 2012, he candidly admitted his initial attraction to the post was born from the company being “ready-made to do a whole bunch of other products, which I wanted to do.” He’d also forewarned SL users than the company would be diversifying its product brief during his 2011 SLCC address.
Many objected to this on the grounds it was “taking away” time and effort which might be focused on Second Life while others felt that it was a misappropriation of “their” money, or that it signalled “the end” of SL. In terms of the latter, the reality was, and remains, far from the case. In fact, if it can be done wisely, diversification might even, over time, help SL by removing the huge pressure placed upon it as the company’s sole means of generating revenue.
The problem is that the direction that has been taken by the Lab thus far doesn’t appear to be the most productive revenue-wise, at least in part. The apps market is both saturated and highly competitive (and even now, two of the products in that sector have yet to arrive on Android). Similarly, it might be argued that Desura could be more valuable as a marketable asset than as a long-term investment), and dio appears to be going nowhere. All of which leaves Patterns, which in fairness does appear to be carving a niche for itself, and has yet to be officially launched. It will be interesting to see what, if any, appetite the Lab has for continuing with these efforts now that Humble has departed.
There have been missteps along the way, to be sure. Humble’s tenure has been marked by a series of ongoing and quite major issues with the SL Marketplace which the company appeared to be completely unable to bring under control. These prompted me to wonder if “putting the ‘Lab’ back into Linden Lab” might actually work in all cases. Worse, they led to a clear and continued erosion in customer trust where the Marketplace was concerned and quite possibly damaged Humble’s own reputation. Despite promises of “upping the tempo” with communications and updates, all merchants saw was the commerce team reduce communications to the bare minimum, and refused to hold in-world meetings which might otherwise have improved relationships.
Similarly, some projects were perhaps pushed through either too quickly or without real regard for how well they might be employed. Mesh was perhaps prematurely consigned to the “job done” basket, particularly given the loud and repeated calls for a deformation capability which were spectacularly ignored (and are only now being addressed, after much angst and upset in the interim, all of which could have been avoided). Pathfinding has failed to live up to the Lab’s expectations and still appears to be something that could have been pushed down the road a little so that other work could carried out which might have left people more interested in given it a go.
Broad-based communications between the Lab and the user base had been in decline prior to Humble’s arrival, prompting Tateru Nino to dub the expression “The Silence of the Lab”. Initially, it appeared as if this might be reversed under Humble’s tenure, given that he seemed to be everywhere, Tweeting and chatting via the feeds, taking a hand-on approach to issues and updates, etc. The irony here is that, by making himself so accessible, he inadventently gave the impression this would always be the case, when obviously the demands of his role as CEO would dictate otherwise.
The failure here is not so much that he stopped communicating in this way but rather that the Lab didn’t see the need to put in place a mechanism to ensure outward, broad-based communications would continue off the back of his accessibility through social media. Instead, communications were allowed to decline further. Worse, they also became unpredictable, randomly bouncing from the occasional blog post to e-mail notifications, to short-lived “newsletters”, to forum posts, back to e-mail notices, and so on. Had the Lab sought to ensure more consistent and informative communications of matters of import to users, then issues such as the confusion over tax documentation may have been avoided, while matters relating to the August 2013 ToS changes might have resulted in less angst and anger had the Lab been more pro-active in explaining things.
There has been criticism that Humble didn’t “do more” to attract new users to SL. Ironically, when the Lab looked to open new potential avenues through which to attract new users, they were equally lambasted. However, one can understand why there may not have been more work put into the whole “new user experience”, because the focus has been in ensuring the platform is robust and attractive enough in order to retain any new users the Lab might aim to try and attract in the future.
That said, the sign-up process was overhauled with impressive results (which cannot be entirely dismissed as people creating alts), and the Lab did start experimenting with options for getting people more rapidly actively engaged in the platform, most recently with the A/B testing of the new “Social” and “Learning” islands. What we now need to see is a solid push in this direction, with a cohesive strategy behind it that can demonstrably lead to a greater influx of people into SL who “stick” as engaged users. This is certainly something that was on Humble’s roadmap for the platform given his own statements on the matter of “getting back” the 30 million or so who have tried and left SL, so there is no reason to assume it won’t survive his departure.
For my part, I cannot help but feel that taken on balance, and in terms of Second Life, the Lab’s Humble Years have been largely positive. That’s not to say there haven’t been mistakes or that there aren’t issues which have yet to be addressed; far from it. Rather, it is a reflection that in terms of the platform, it is far more capable and robust and doubtless easier to manage and maintain with his departure than it was on his arrival.
Efforts elsewhere with the company may not have been so positive – but where diversification and new products are concerned, success isn’t going to happen overnight or even within the first couple of years. Hence why it will be interesting to see what happens in this area with Humble’s departure.
As to the future, we have yet to hear from LL and their plans. Will the recruit from outside or promote from within? I would tend to say they’ll opt for the former. However, were they to promote from within, then I’d most likely point to Malcolm Dunne or Don Laabs as possible candidates. Dunne is the current CFO, and that position, coupled with his experience would seem to make him the more likely candidate of the two. However, Laabs is the Director of Product for SL, and he has also been a CTO in the past. Taken together, these might make him an attractive proposition for the board, as he potentially has the necessary understanding of SL and the wider experience for managing LL as a whole.
In the meantime, people will continue to debate the pro and cons of Rod Humble’s time at Linden Lab. For my part, I’d like to take the opportunity to thank him for his time at the Lab, and for the time he took with me in a number of off-the-record exchanges, and to wish him all the best with his new venture, about which I certainly hope we’ll be hearing more in the near future.
Inara Pey: Living in a Modemworld 
I have nothing personal against or particularly for Rod and I wish him well in future endeavours as I would anyone else. But I can’t laud him as a wonder worker at the helm of SL, The following is what I gauge success on and they have been there for the last 6 years;
teleport and actually arrive at the destination and be clothed with attachments,
not be so lagged I can’t move or render,
not be logged out unless I wish to be,
use search and actually find what I’m after,
not being rubber banded at sim crossings,
if told there will be gifts for premium members actually get gifts,
have timely and proactive response to infractions of the TOS,
there are more but I won’t list them………
As a user I would like changes where I can see positive results not the “under the hood we did this to the servers” stuff. I want to see and notice the change. Paving the road but not filling in the potholes doesn’t impress me.
It’s the user experience that needs an overhaul, concentrate on that and the other stuff follows. Build games in games once the user experience is cleaned up.
So you see this is what governs my assessment and so far it hasn’t been met.
Maybe next week…….
Hugs Dee
LikeLike
It’s the “under the hood” stuff that frequently improves the user experience far more than the visible stuff. Just because it’s not an “in your face” improvement doesn’t lessen the cumulative positive impact it can have on us all (the HTTP work being a major case in point). It’s probably fair to say that visible stuff can all too often be cosmetic, delivering a short-term benefit without addressing the underpinning issues (so you might say it can often be “filling the potholes” 🙂 ).
Region crossings have been a mixed bag of ups and downs in terms of rubber banding and other issues, but while there are still issues (and likely always will be issues of some description given the nature of the beast) I tend to think the overall direction has been towards reducing the amount of pain experienced.
It would seem as if some of the issues you are experiencing might be down to factors outside of LL’s control – although this is hard to assess based on the description here. Rendering / teleport / lag / disconnect issues for example, could be down to network issues, possibly router issues – have you tried the HTTP release candidate to see if things are more stable (particularly connection-wise)? And while it is a controversial thing to say, lag does tend to be more endemic at the user’s end of things than at LL’s, and can simply be caused by incorrect settings in the viewer (such as setting network bandwidth too high or over-tweaking a debug setting, for example). There are also other things within the platform but outside of LL’s control which can impact things like rendering and generate viewer-side lag – such as an over-abundant use of extremely large (1024×1024) textures, even for small surface areas.
As to premium gifts, that’s a new one for me. My personal experience is that I’ve never failed in being able to hop over to one of the gift vendors and pick things up, either at the time of release or afterwards, so i really can’t comment, other than to offer to go to a vendor and kick it on your behalf 🙂 .
Search is a pain, both in-world and on the Marketplace.
ToS infractions & AR responses are hard to judge; we don’t actually know what happens the majority of the time as everything is behind one big curtain. We do get to hear of “high profile” cases where nothing seems to happen, but I’m not sure that’s representative of the process as a whole. That said, more informative communications would likely help, rather than ARs seeming to fall into a black hole.
LikeLike
We will not see eye to eye on this for many a reason and I am not trying to sway anyones thought for or against anything, I am stating what has an influence on me and that is results I can see, so far I haven’t been WOW’d.
Where I work we had an IT guy who did all kinds of things unseen to users and when properly checked out nothing had been finished, but who knew. Not saying that is the case here but…. The last user effect implemented by a CEO that I can remember having a great user impact was good old Mark and the Homestead fiasco.
Premium being received was not the problem it was the frequency at which they have been issued. Start monthly,,every 2nd month….quarterly and well six months give or take to the last one. Remember that was suppose to bring in new blood but sort of died in the effort.
So I’m just not as impressed as others seem to be and to me my reasons are just.
Cheers – Dee
LikeLike
Inara,
I think you might be a bit biased because you admittedly had personal interactions with Rod that 99% did not and you might feel that he TRIED his best. But what I find so frustrating is when bloggers like you and others that are much closer to the Lindens and that attend many of the LL weekly meetings are saying how the Grid is so much better now than before Rodvik arrived.
Maybe you and these other bloggers do not spend enough time at real SL Grid events where there are larger gathers…. art gallery showings, parties, special fund raising events, the countless daily live music event, and busy recorded music clubs. If you were at many of these event as myself and countless others have been at that, you would see that the grid is NOT as improved and lagless and stable as you believe it is because of what maybe LL has assured you it is. IT IS NOT!
I go to 3 to 6 live events each night. I see sims with 30 60 70+ and even 100 avatars very often. I see mesh commonly not rezzing on avatars, venue builds taking forever to rez if at all. I see crashes often at venues. I see griefers using Mesh as a weapon. The Server side baking has done NOTHING to improve the grid performance.
And please don’t respond with “maybe your pc or network or router or viewer is not right”. I have a top end powerful PC, 25mbps internet, tried many of the viewers. The lag and instability are coming from SL not from my side. And countless of residents at these events are all bitching about the same stuff. Try showing up at these events as frequent as many of us are and you might not be as impressed at Rodvik’s successes.
And then there is the on-going issue of LL not fixing long time lasting bugs like broken attachments on notices, avatars disappearing when they jump on a pose ball (which showed up after SSB), not increasing max groups, etc. etc.
You failed to mention how Rodvik was responsible for gagging the Public Jira as part of his jihad on closing most avenues of communication between LL and their customers. A perfect example was the way Rodvik and staff handled the current TOS content creator IP highjacking.
The fact that LL refused to even formally announce his departure and that he all but slithered quietly away from LL by his own personal announcement has to raise eyebrows as to why.
conclusion for me… I initially thought Rodvik showed promise for LL. Now that hes is leaving, I for one can comfortably say I am glad to see his departure.
LikeLike
This comes back to what I said initially. What we as av’s encounter travelling the grid is what will set our minds as to whether or not we cheers, curse or go ho hum on the labs progress over a set period of time. Not what we don’t see. Whoever is at the helm during that time frame wears the garland of roses or is pelted with tomatoes, that comes with the territory.
We don’t have to agree in this case but it is interesting to note that not everyone is cheering for Rod or throwing tomatoes but I am going “ho hum”.
Still waiting to be shocked by positive change on the grid and when my experience becomes better I will cheer for sure.
Time to move on to the next topic 🙂
LikeLike
“I think you might be a bit biased ”
Bias is evident in everything we do. Human nature. Had my bias been towards the predisposition that Rod failed, failed, failed, you’d doubtless be here praising all I’d said, because then my bias would be more closely aligned to your own.
But that said, this article was taken as an attempt to step back is look at things from a) what has been going on at the Lab for 3+ years and b) how most people I encounter in my in-world travels, and who comment on this blog and attend user group meetings, go about their own time in-world; and as a reaction to what has been happening through the forums / venues I attend and the activities I’m involved in.
You may disagree; that’s you prerogative as dictated by your own bias 😉 .
LikeLike
The first two problems are often caused by factors the Lab can’t control.
LikeLike
I can’t remain quiet any longer. This pesky issue about losing attachments during TP is NOT, I repeat NOT a matter over which LL has no control. The issue is related to the association of asset files with accounts as they move from server to server. There is plenty of evidence that with the server now performing many more tasks per unit time than it used to do, the re-association of files is becoming less reliable. I have encountered this issue on a number of occasions, sometimes several times daily and it has nothing to do with either my network connection quality of my computer’s operation. The quality of hand-offs between severs has noticeably worsened in past months, making simple region-crossings on “foot” variable and with vehicles nigh-on impossible. Yes, there were permissions issues which were (eventually) addressed, but the fundamental issue of the nature of the communication server-to-server is as bad as ever. Server-to-server communication is now only adequate if the servers are in the same rack at a datacentre. If they are at different data-centres….hold on to your hat, ‘cos it is likely to be blown off!
LikeLike
I would also add that it is not simply a matter of the Lab not being able to control some issues, it is a matter of them not being WILLING to control or address them….Oh and it should have been “or”, not “of” my computer….
LikeLike
Ayesha, I does seem to me that a lot of those particular problems seem to coincide with the introduction of Server Side Assets (SSA). In fact I tend to see the particle cloud more often on login now while waiting for my av to rez. I think investment in computers that can handle the current demand or re-assigning the shares is the problem. In fact it really always has been in my assessment.
With all the new add ons/capabilities is anyone looking under the hood at hardware?
sigh
LikeLike
“Server-side Assets”? Do you mean Server-side Appearance? If so, that perticular mechanism shouldn’t have any impact on attached assets. It applies only to avatar system layers, just as the old viewer-driven baking process didn’t actually touch attached assets.
LikeLike
My comment was broader than loss of attachments – and was actually meant to refer to teleports in general. I’m certainly not for a moment suggesting that that issue in particular is something the Lab can’t control.
Part of the issue here is that you seen the problem as endemic to the grid, but many (myself included), simply do not encounter it. And in my case, it’s pretty fair to say I’m travelling far-and-wide across the grid on an almost daily basis. Further, a frequent companion I have and who has many more attachments than I, hasn’t to my knowledge encountered the issue in the 18 months we’ve been travelling together.
This is not to say that just because I don’t have the issue it’s not happening – but it does potentially suggest that if the majority of SL users are in the same boat as me, and aren’t encountering issues, it’s entirely possible that continuing problems aren’t (unfortunately) that big a blip on LL’s radar at this point in time.
In that respect, it would be interesting to be able to see how many JIRA have been raised on the matter and how frequently, and the steps needed to conclusively repro the issue. As it is, I really cannot remember the problem even being raised at simulator or server beta UG meetings, much less being repeatedly raised.
LikeLike
Part of the problem, Inara, is the repeated insistence of Lindens that ANY assessment of SL function be made using their default viewer. Now this attachment issue is suffered by users of both the Linden Viewer and Firestorm users alike, or at least I am told that it is being encountered by LL Viewer users, by those users.
Any BUG report I raise is either discounted until I report while using the LL viewer or is closed as a duplicate. I am tired of wasting my type. I do believe that the attachment is exacerbated the more attachments that your avatar uses, and I use a lot I freely admit, though these days the bulk of them are not scripted.
The reason, in my opinion that this issue is never raised at UG meetings is that most UG attendees do not adorn their avatars like Christmas trees!
It will be interesting to see how the integration issues of the fitted-Mesh viewer and Firestorm proceed, I have a feeling that this could be a rocky road.
LikeLike
Your comments are factually wrong. Let’s begin, shall we? For starters, JIRAs are accepted regardless of the viewer you are using. Second, I have raised issues in UG meetings – and some were considered, actually – and I have gone on record for exclusively using Firestorm. Third, I have yet to see anyone “lose” attachments. What I have seen and experienced, and it still happens, is rez fail for attachments, but this is on its way to getting fixed. And finally, there are quite a bit UG attendees that have avatars chock-full of attachments (prim-based, sculpties, mesh, everything).
LikeLike
Also, I confirm what Inara has said; I have considerably more attachments (some of them scripted) than Her, and I’ve never lost a single one of them in our journeys, regardless of where we teleport to, and regardless of how many regions we cross in a row. However, if you really are facing this issue, do by all means file a JIRA and detail what you experience, what is happening, so that the engineers at the Lab will be able to look into it.
LikeLike
In fairness, I can understand why LL request that their own viewer be used to assess matters. They’re not in a position to test every viewer to ascertain whether a particular issue lies within the core viewer code or within the UI code (which may have been modified by a TPV). Sometimes that does mean that they appear to be pedantic where a issue appears to be entirely server-side, but it does make for a consistent approach – and it’s not as if their front-line support staff can be expected to know the nuances of very single viewer.
As to the UG meetings … you’ll be surprised how many attachments some attending them have. But that aside, even if those at the meetings are lightly loaded, they do get to hear about things, and have problems raised by others with them on the basis that they do attend meetings. As such, it’s not unreasonable to anticipate such a widespread concern having been raised, and repeatedly raised…
LikeLike
OK, I hold my hand up as regards UG meeting attendees, I was speaking in ignorance there, it would seem.
But, Mona, I have had the response from Lindens regarding viewer use on several occasions, my data is simply not acceptable, I am told, unless it is obtained using the Linden default viewer. That IS a fact and it is not up for debate. I can understand why to some extent but given the breadth of the Firestorm user-base in SL it is not very helpful.
Actually ignorance of the some of the most popular viewers in SL is a shortcoming in Linden frontline staff which they should address (in my opinion).
LikeLike
In the SL Bug Tracker (the JIRA) there are numerous submissions where bugs, exploits etc., are being reported by people who use TPVs, not only LL’s official viewer. While it is true that LL does not maintain or support TPVs (which may not always incorporate the latest version of LL’s official viewer code and, thus, may not take advantage of recent bugfixes and improvements), more than 90% of your favourite TPV’s code is LL code – and it doesn’t matter if we’re talking about Firestorm, Cool VL, Singularity, Exodus or whatever. The remaining 10% (usually less) is minor tweaks and embellishments that aim to give a more appealing UI and make certain options more easily accessible to the user. Bug reports are accepted, regardless of viewer used.
The fate of the report, on the other hand, is another matter. If it can’t be reproduced, the report will be closed. If insufficient data is given, the report will be closed. If the problem reported is caused by something other than the viewer or the server, it will be closed. If fixing the problem could cause bigger problems, the report will be closed. If the problem is because of outdated code in your TPV and has been fixed in a newer version of the official LL viewer, the report will be closed. And so on. Reports are not closed because of the viewer used. And I know, because, in several UG meetings, I’ve raised a few issues (most of them feature requests), and Nyx, Oz and Simon have all asked me to prepare some JIRAs for them, so that they can see what they can do; and they know (because I’ve gone on record for this) that I use Firestorm and not the official viewer. I still haven’t gotten around to writing these JIRAs yet, because RL has been tremendously hectic these past few months, but they didn’t say “you use a TPV, so you can’t file a JIRA with us.” Quite the contrary, they actively encouraged me to file JIRAs, because they want to improve the platform.
At any rate, I think you really should write down and detail the behaviour you have observed, including your system’s details, where it happens, how it happens, under which circumstances it happens, file a JIRA, and hop over to Tuesday’s Server/Script meeting to give Simon, Kelly and/or Baker a heads-up regarding the issue, pointing them to the JIRA. At the very least, they’ll look into it.
LikeLike
Mona,
ever since Rodvik gagged the Public Jira, the Jira has become almost an ineffective useless tool. Since no one can see each other’s jira entry and add facts and experiences and confirmation to a reported issue, Rodvik pretty much Gutted the Jira. Since he did, I have not used or even enter the LL Jira system. This is one of the many Rodvik failings… one of the biggest ones. This Jira Gagging was even quietly hated by LL Engineers themselves as a few Lindens told me they were not in agreement with Rodvik gagging the Jira and it would majorly hurt their ability to identify and solve grid problems.
So… the Jira is no longer an effective grid problem solving tool for LL… thanks Rodvik.
LikeLike
It’s easy to claim LL should do a lot of things – harder in practice for them to do so.
Just which TPVs should they try to support? What criteria should be used to determine which they support and which they don’t? What about those they opt not to support? Why should their users be left out? There will be upset and claims of favouritism, FIC and so forth. How much time should support staff devote to supporting viewers. How much time should they devote to keeping up with UI changes and additional functionality (and two different UIs) within viewers?
If you’re unhappy about filing an LL JIRA, have you filed a Firestorm JIRA? The FS team work in close cooperation with LL (as do all the “core” TPVs), and they regularly share JIRA information. If this is a widespread issue, filing with Firestorm may help bring it to LL’s attention as it is verified and referred back.
LikeLike
Furthermore, in addition to what Inara has pointed out, I would like to note that the “TP” in “TPV” means “Third-Party” for a reason and it means that the development team behind your TPV of choice (Firestorm in this case) is responsible for its development and support and that they are always the first persons where you should go for support. The people behind Firestorm, as well as the people behind every major TPV, are in very close cooperation with LL to ensure things work as intended. So, I highly recommend starting with a Firestorm JIRA, talking with Firestorm’s people and, if the behaviour you describe can be reproduced and attributed to something that has to do with viewer code, then they can work it out, either by themselves or in collaboration with LL.
LikeLike
OK, OK. With a polite note to all, let’s not have this deteriorate into intractable opposing views and statements. Suggestions have been made which don’t need belabouring.
LikeLike
OK, this is my final comment here, it is getting too contentious.
1) Yes I have filed Firestorm JIRAS about matters prior to raising LL BUG reports, though on this specific occasion (attachment issue) I have not done so yet.
2) I speak frequently with the Firestorm folk on a variety of issues.
3) Mona, I never said that BUG reports I filed were ignored, simply that they were closed unless I reported the behaviour in a Linden Viewer. In retrospect some of the issues I have attempted to understand have been related to RLV(a) and this present one might be also, I am investigating this.
4) You are probably right about the TPV issue and LL, Inara, though Firestorm are already accused of “cosying up” to LL by some ne’er do wells, so it is very likely a case of “Heads I win, tails you lose”. Ho hum.
LikeLike