Coding Sanity

Nitwits who by their actions try to show companies and the public that the GPL is run by a bunch of underpants-on-the-head kind of commie pinko weirdos.

Here's a hint Harald, and what the hell, let's include RMS in this too, most people involved in OSS or semi-OSS stuff are in it to help the community and not to push some radical political agenda. When you give them no choice but to sign up for your latest crusade, then quite a few will drop out.

Free as in do-what-I-tell-you-or-else free

After my previous bad experiences with the Operating System That Was Shipped in Beta, you might be surprised to find out that I recently installed it on my laptop. Interestingly enough it is far more stable and faster on my laptop than my desktop. My personal theory is that my laptop does not have an nVidia graphics card, nor a Creative sound card. Both are companies whose Vista support was late and uninspiring. Anyway, the number one, most irritating bug that I encounter regularly is file copy. Yep, you've heard me right, Vista cannot adequately perform operations on it's filesystem. I wanted to delete a directory holding backups, so I selected the folder, and pressed Shift-Del, and then had to wait for 2 hours whilst Vista carefully went and, from what I can tell, lovingly worked out the thumbnails for each file it was going to delete before actually doing the deletion. Oh, wait, it could also have been checking them for DRM.

Copying is so pathetically slow for anything large that in some cases, it's actually easier and faster and less frustrating for me to boot up a VPC containing Windows XP, share the appropriate folders on each Vista machine and use the XP to move the files from the one machine to the other.

Microsoft have been made well aware of this bug, but have not yet addressed them. In fact these bugs were raise during the Beta program IIRC, and still, Microsoft have not yet deigned to fix them. Copying files, deleting files..., high-tech shit.

I was inspired to blog this, because this morning I decided I'd like to watch a movie in bed. I powered up my laptop, and selected a movie from my desktop PC, and tried to copy it across. It started "Calculating time remaining", and had 0 bytes/second, and just sat there. I pressed Cancel, and the Cancel button just disabled. I waited and waited, and still that useless, buggy as all hell, copy dialog just did nada. So, I force restarted my laptop and tried again. When the exact same nothing happened, I decided to blog about it. So here I am, still staring at the 0 bytes/second, and still "Calculating time remaining". Here, Vista, a hint, if you're consistently at 0 bytes per second due to some unfathomable bug in your non-release-ready code then the time remaining is infinity.

Microsoft assured us that Vista was production ready, and it is actually a lot better than I feared it would be, but it is not release quality code. It appears that consumers are well aware of this, which is why Microsoft's financial results showed little boost from Vista sales. Microsoft have been engaging with consumers attempting to convince them to purchase Vista and not wait until SP1 comes out. This is a common tactic used by consumers to ensure tha most of the bugs are resolved. Such a tactic will mean far less frustrations for consumers, instead of acting as broad-scale beta testers, they will be hopefully actually be getting the code that should have shipped. I'd like to add that I don't think they should upgade existing PC's but only put Vista on new PC's due to the plethora of driver problems experienced by user of older machines.

Good Points

Is Vista worth the upgrade? So, far, just using my laptop as an example, yes. It's snappier, more responsive, much prettier. I find myself using the one-button off/suspend feature a lot and not bothering to properly shut down like I used to under XP, there's just no need. The search start menu feature has made my life a lot easier. Before, I'd go through my menus, setting up shortcut keys everywhere. Now there's no call for that. Mobility is also better, but I still wish they'd store IE settings like the proxy server against a named network, so I didn't have to keep fiddling that whenever I move from A to B. So, I am quite confident that Vista will be a huge success once they fix all the irritating little problems.

Speaking of which, I wish I could tell it to stop whining that I blocked some starup programs. I did it very deliberately, and it was a well-thought-out idea. Every 30 minutes or so, it pops up a bloody tooltip telling me that it did what I told it to do, and that it followed my instructions when I booted my computer 15 hours ago. I KNOW dammit!

Normal Richards has a brilliant idea for making exception handlers a little neater. His example is for Java, but the syntax would work fine for C#.

Well, well, well, it appears that we will need a separate topic for this. My eNaTIS Performance post was very popular, with a lot of people expressing their dissatisfaction with the system. As before, I'd like to try and keep this to facts and figures, however in this case that may be difficult. There have been numerous rumblings about the security on this high-tech boodoggle, starting with [from my previous post]:

The Beeld has a report about the security issues with eNaTIS system. Very worrying was that one did "not need a password to log on as an ... administrator", that the "documents ... are not secured", and "can be circulated [without any protection]".

The eNaTIS website for a while had it's administration area unprotected, and now the site itself has been hacked. Apparently this was during the course of Wednesday. By the evening of that day, according to News24, the entire site had been defaced. The eNaTIS website downplays this incident claiming:

This was apparently due to someone leaving a comment on a page of a section of the eNaTIS public web site (this site).

Ummm, correct me if I'm wrong, but wouldn't this be the tried and true cross-site scripting attack? You know, the one that nobody gets anymore because it's so well known and so easily protected against? Sorta like buffer overruns, such an attack is largely a blast from the past, encountered only by those with very little understanding of security.

The eNaTIS system and database is still secure and cannot be accessed via this web site

Good, good, of course, the whole idea of eNaTIS was to have it available to the public via a website. So what they effectively appear to be saying here is "we're really happy we haven't got around to implementing the web access part of eNaTIS yet". It's not a proud day for security when only your project delays mean that your core database wasn't exposed.

From Fin24: "It's nothing to write home about," said Transport Department spokesperson Collin Msibi when told that part of the eNatis website had been hacked. Let's see some other prescient/accurate quotes by this man:

• "[Msibi says delays there were probably caused by faults on Telkom lines and not with eNatis system]"
• "[licensing offices should be operational tomorrow following the installation of extra computer capacity for the eNatis system]"

Yeah, not a high strike rate. I'll tell you exactly why it's something to write home about. The ease with which the website was hacked indicates a possible profound lack of security knowledge on the part of the administrators. This indicates that perhaps the auditor-general was right when eNaTIS was castigated for a lack of security. That's the system, not the website that was being talked about as a matter of interest. So, what we're seeing here is a possible pattern of insecurity which is quite worrying. The information coming out of the DoT and Tasima seems to be optimistic in the extreme given the numerous documented failings of this system.

This system is a debacle, recognised by such by all and sundry, except the DoT and Tasima. Exorbitantly expensive, late, unreliable, slow, and insecure, it seems to be a poster child for poor development practices. To cap it all off, us taxpayers who have already paid for this system once, must continue to fund Tasima's efforts to keep it stable through a R30 surcharge on our transactions.

CNet has a story where they discuss a slowdown in the percentage of software developers targeting Windows. The figures come from Evans Data who say:

• 65% of developers are targeting Windows, down from 75% last year, almost a 9% drop.
• They expect a further 2% drop next year.
• Linux in the same period has picked up 34% more developers, up to 12%.

The rest of the article progresses into a fantasyland of pabulum about a WebOS. Reading the Evans release, John Andrews, CEO says:

it's clear that a shift away Windows began about two years ago, and the data show that this migration is now accelerating. Linux has benefited, but we also see corresponding growth in niche operating systems for non-traditional client devices. The landscape is changing.

Evans also clarify that the survey is based on North American developers.

Why?

Well, in the comments of the CNet article, it appears that a lot of blame is being laid at Vista's doorstep, in my opinion quite rightly. The massive delays, huge cost overruns, bad DRM implementation, bugginess and poor driver support in Windows Vista have led to it being perceived as unreliable, frustrating and slow. I can certainly attest that in my personal experience it is all three of those things.

Quite simply, people who never even considered the OS as a choice are now reconsidering their options. In addition, more and more apps are being delivered over the Web, so the targeted platform of the developer is in a very real sense HTML rather than Windows/Linux. Projects like Mono make it easier for developers to bounce from one OS to another, and this trend will only continue and accelerate. I honestly believe we are not seeing the beginnings of the death of Windows so much as the commoditization of operating systems.

What next for Linux?

Well, for starters the Linux guys need to be calm about this, the jump is a one year spike that can probably be attributed more to Microsoft's blunders than any siginificant success for Linux. That said, those extra developers will make it easier for Linux to develop more compelling applications, especially as one hopes a few of them have a clue about UI. Ouch! Okay, that was a cheap shot, Linux UI's are getting much better these days, really. Priorities should still be on making interoperation easy, pushing the commoditization of the operating systems even further along. Hell, Linux is free, so what the hell do they care if prices of OS's drop precipitously? It will still be a long haul, and as always Linux's biggest ally in fighting Microsoft is Microsoft itself. Worst case scenario for Linux: someone manages to revolutionize MS development practices, and the MS Marketing team is fired en masse.

What next for Mac?

Yeah, not so much. Evans were pretty silent on that. Mac will gain due to more targeting of web front-ends, but so will all the other OS's. It'll gain a bit from improvements to Linux, but I think most of the extra developers will be working either on web apps, or will mainly be writing Linux desktop apps, which won't work out especially well for Mac. Result: Mac OS will likely stay a fringe OS, and any OS commoditization will hurt Apple too.

How about Windows?

Well, much as some people will cheer loudly about these stats, this is not even close to any kind of death knell. The plucky underdog has landed a good right cross, that's all. The loss to Linux is anticipated to be far less next year, although even a gradual loss, if steady, can be seriously damaging. If SP1 sorts out the most egregious Vista issues, it might actually turn it into an OS worth getting excited about. Keep in mind that XP only really started getting popular around the SP1 timeframe too, and Vista does appear to be on a better base. I know there has been a lot of introspection about what went wrong with the Vista development, and hopefully something useful and concrete will come out of it. Result: Iffy, given a good bit of management and technical leadership and Microsoft could easily become more dominant in operating systems than ever. Continued missteps could lead to gradual attrition. I honestly believe that this game is Microsoft's to lose.