Channel 9 - Discussions by glebd

Coffeehouse - Windows 7 Ideas

glebd — Mon, 21 Jan 2008 01:30:33 GMT

I think most of you are missing the point. As I read your suggestions (except maybe for one that suggested rewriting from scratch), the expression "polishing the turd" comes to mind. If Microsoft doesn't rewrite the whole thing and gets rid of all the compatibility crud (perhaps by virtualising it), Windows is doomed. Think about it: Vista code base is 40% bigger that XP code base (an unconfirmed quote from somewhere online) -- there is no way you can turn this to an advantage! There is a good article somewhere on the nets comparing Vista to dBASE IV (hint: compatibility killed it). The problem is, though, that at the moment compatibility is all that Windows has. If people had to choose between, say, Windows-2010-Without-Backward-Compatibility and other OS, they'd probably keep running XP forever (in enterprise) or switch to Mac or Linux (at home). Make your own conclusions.

Coffeehouse - The silence is broken! (IE8)

glebd — Fri, 07 Dec 2007 11:11:25 GMT

I find it strange and even suspicious that Billg doesn't know (or is pretending not to know) what is happening in the team that produces one of the most important MS products. Talk about non-transparency...

Tech Off - Poor performance of .NET Regex

glebd — Thu, 18 Oct 2007 12:30:24 GMT

AdrianJMartin wrote:

one huge difference is that the body of text you are searching will have been pushed into unmanged memory...

Not really, as the text is split into lines first in the C# program before matching each line with regex. So I don't think managed/non-managed memory difference comes into play here.

The only difference in the code was that the initial version used Regex.Match(), and the new version uses PCRE wrapper function to do the same on the same strings using the same regular expression. The surrounding code is absolutely the same. The difference in speed must be in the way regular expressions are implemented in PCRE and .NET.

Tech Off - Poor performance of .NET Regex

glebd — Wed, 17 Oct 2007 10:40:44 GMT

JohnnyAwesome wrote:

Are you able to post any code samples with the difference between implementation of your compiled C# RegEx class and your wrapper call?

Unfortunately no, as all of this code is client-owned. However, I only used simple matching, there is nothing fancy there.

Tech Off - Poor performance of .NET Regex

glebd — Wed, 17 Oct 2007 00:17:08 GMT

JChung2006 wrote:

Did you use a compiled or uncompiled regular expression in your C# version?

Tried both, couldn't notice any significant difference. That's when I pre-compiled the Regex object into a DLL (which didn't help either, quite surprisingly.)

Tech Off - Poor performance of .NET Regex

glebd — Wed, 17 Oct 2007 00:04:12 GMT

Recently in a C# application I was writing I had to process quite a few lines of text using regular expressions (many thousands). Initially I used .NET Regex class for that. The processing of the entire amount of data took a little longer than 4 minutes on a P4 3 GHz machine (Windows XP Professional.) After profiling I saw that most of the time (99%) was spent in Regex.Match() function. I tried to pre-compile the Regex object into a separate DLL to improve speed but it did not produce noticeable difference, the whole amount of data still taking about 4 minutes to process.

Not satisfied with this (we had to go through the whole process quite often and it was very bothersome), I downloaded a Windows port of PCRE C library (Perl Compatible Regular Expressions, http://www.pcre.org/). I then produced a very thin managed wrapper around the plain C API of that library using C++/CLI, and called the wrapper from my C# application, replacing Regex.Match() with the appropriate PCRE function. The regular expression, however inefficient it may have been (I'm no regex expert), stayed the same.

The next time I ran the application with a stopwatch, I couldn't believe my eyes. The whole data processing took... wait for it... 8 SECONDS!

So, using the same regular expression and the same data: .NET Regex: 4 minutes, PCRE wrapped in C++/CLI: 8 seconds.

Can anyone explain this kind of difference? Is PCRE a miracle? Or, more likely, is .NET Regex implementation really THAT bad?

Coffeehouse - So .NET is "Resource" Sourced... what's next?

glebd — Fri, 05 Oct 2007 22:36:23 GMT

ScanIAm wrote:

I'm sure there were some lawyers involved in this decision, but if you limit your developer pool to people who know .Net and haven't ever felt the desire to take a look at how it works, then you might not have the best developers to choose from

Well, you surely would not get those who like to copy-paste their code.

Seriously, are you sure .NET source code is all rosy? I bet they had to clean it up quite a bit before releasing for everyone to see, if the previously leaked Windows 2000 code is any indication (I haven't seen it, just read this review).

From the initial announcement, a snippet of commented .NET source:

...
finally {
    // restore culture
    RestoreCultures(currentThread, prevCulture, prevUICulture);
}
...

Coffeehouse - 7.4.3.....holy cr@p Apple!!!

glebd — Fri, 05 Oct 2007 22:23:15 GMT

CannotResolveSymbol wrote:

glebd wrote:

Harlequin wrote:
ANOTHER iTunes update. ANOTHER 55MB. Does Apple not know how to apple patches or something?

I take it you are equally annoyed by Microsoft's "patch Tuesdays"?

The point here is that Apple doesn't patch. They send down the whole installer and reinstall the product, all 55 MB of it.

That's rather large for a point release.

If you release a product and then keep patching it instead of updating as a whole, you end up with an unmanageable mess. It is much easier in the long run to release a complete update than keep supporting various patch combinations and make sure they all work together. Hint, hint...

Coffeehouse - 7.4.3.....holy cr@p Apple!!!

glebd — Fri, 05 Oct 2007 21:25:53 GMT

Harlequin wrote:

ANOTHER iTunes update. ANOTHER 55MB. Does Apple not know how to apple patches or something?

I take it you are equally annoyed by Microsoft's "patch Tuesdays"?

Coffeehouse - So .NET is "Resource" Sourced... what's next?

glebd — Fri, 05 Oct 2007 14:23:38 GMT

thumbtacks2 wrote:

It would be a silly if Mono doesn't hire you because of "seeing the source code". That's their own choice, though, and I can see the legal issues they are trying to avoid.

Remember SCO (there was a company named SCO once, and it was even listed)? They suddenly decided that Linux contained code infringing on their ©.

If a code snippet appears in Mono that is similar to .NET code and there are Mono contributors who had seen the .NET code, MS potentially may want to shut down Mono or do something else along the lines.

Now, given SCO current status, they may want to think twice about that, but still the possibility remains. I guess that's why Mono guys are being careful about this.

Coffeehouse - So .NET is "Resource" Sourced... what's next?

glebd — Fri, 05 Oct 2007 12:35:40 GMT

AndyC wrote:

glebd wrote:

Interviewer: "Have you ever used (seen, stepped into during debugging) the .NET shared source code?"
Developer: "Yes, of course. This helps the development, you know."
Interviewer: "Sorry, we cannot employ you. Your mind is tainted."

That's already far, far more likely with GPL code than it ever will be with the .NET shared source, unless you're planning to go work for Novell on mono.

I think you're wrong here. First, there is not a lot of GPL-licensed .NET code out there that can taint a developer's mind, and if there is, you can avoid it easily. But that is going to change because the entire .NET runtime will become "contagious" in the sense GPL code is now. Except that you can still use GPL code in GPL projects, whereas you cannot use any of the code you see inside .NET runtime for anything. Oh, and you will be hard-pressed to avoid looking at it — Visual Studio will gladly download it for you and automatically step into it, should you inadventently press "Step Into" key during debugging.

Coffeehouse - So .NET is "Resource" Sourced... what's next?

glebd — Fri, 05 Oct 2007 12:18:43 GMT

Open-sourced? Yes. "Shared-sourced"? I don't think you would want that. God forbid you to write something similar after looking at the "shared" code that is still protected by MS ©.

As to your question on what's next... I expect soon C# developers will start hearing a certain question during their job interviews:

Interviewer: "Have you ever used (seen, stepped into during debugging) the .NET shared source code?"
Developer: "Yes, of course. This helps the development, you know."
Interviewer: "Sorry, we cannot employ you. Your mind is tainted."

Coffeehouse - What do you want to know about Vista?

glebd — Fri, 07 Sep 2007 08:52:01 GMT

Bas wrote:

glebd wrote:

Shouldn't you have organised this before starting the thread?

~~How is one going to organize the appointments with the right people if one doesn't know what questions to ask who yet?~~

Well, if you are going to gather questions about something, you could at least check if it is possible to get answers. I don't know how Microsoft works inside, but I imagine there are team leaders and people who know more than the rest about particular technologies. If you are playing a host for a Q&A forum, there is only so much you can delegate. Most of the work falls on you -- researching questions, finding the right people to answer each question or even a part of each question, and posting answers. Knowing the right people to get answers from, and knowing that they'd be willing to spend their time on this, is the organisation bit I had in mind. Otherwise you are only going to create more frustration in the ranks of your most loyal users.

Bas wrote:

Also, why do I keep falling for your petty hit 'n run attacks?

Um, because they make sense?

Coffeehouse - What do you want to know about Vista?

glebd — Tue, 04 Sep 2007 10:42:02 GMT

Charles wrote:

glebd wrote:

Charles wrote:
Still trying to organize this. Yeah, it's pretty sad that it's taking so long. What can I say?

Shouldn't you have organised this before starting the thread?

No. Questions first. Then answers... I may end up just killing this. Having a hard time finding anyone to take this on. Pretty sad in my opinion.

C

Put it on a public wiki. If we are not getting all the answers, this would at least show all the unanswered questions about Vista. A forum post is not the best form for this, IMHO.

Coffeehouse - What do you want to know about Vista?

glebd — Sun, 02 Sep 2007 14:27:50 GMT

Charles wrote:

Still trying to organize this. Yeah, it's pretty sad that it's taking so long. What can I say?

Shouldn't you have organised this before starting the thread?

Coffeehouse - Interesting Vista Music vs. Network article

glebd — Mon, 27 Aug 2007 01:37:02 GMT

Microsoft explains (Translation: we don't know WTF is going on, but most likely it's all this vast amount of DRM code gobbling up CPU, and we cannot really admit this, so we need to come up with some more FUD.) Pathetic.

Tech Off - Version control documentation

glebd — Tue, 24 Jul 2007 20:25:30 GMT

ScanIAm wrote:

VSS works just fine for a small team

I hear this a lot. No it does not, and it has nothing to do with the size of a team. VSS is completely broken, its integration with IDE is broken too and worst of all, it teaches you all the wrong ways of working with a source control system. If a person who doesn't know what a SCM is starts off using VSS, she is worse off than learning a proper system from scratch.

Subversion is not panacea, it's true, there is Perforce which is very good, IIRC Microsoft used to use its fork called SourceVault internally (not VSS obviously), now presumably switched to TFS which I hear repeats VSS mistakes (the "destroy" command―people, an SCM is supposed to _keep_ the source and allow you to rebuild it going back in time, not destroy it!) There are also distributed SCMs like git and mercurial. And there are heavyweights like ClearCase if you have tons of money and dedicated support staff. These all are proper SCMs.

Search Google (or Live, for that matter) for "SourceSafe bugs" and you'll see what I mean.

Tech Off - Version control documentation

glebd — Mon, 23 Jul 2007 09:38:29 GMT

Larsenal wrote:

blowdart wrote:

Larsenal wrote:
We have a small team of 3 working on this project. We're getting to the point where we need to keep better track of who changed what and when.

We also need to track tweaks for specific clients.

We're all for something simple. Any recommendations?

Simple? Sourcesafe; heh.
Free: Subversion, with Tortoise or Ankh
Paid for: Vault

I should have been more specific. At this point we don't intend to have full version control in terms of source code management. We simply want to keep track of bug fixes, improvements and modifications.

So you want to do things a version control software normally does, like keeping history of source modifications and tracking tweaks for clients (in other words, branching), only by hand? Why would you want to do that? Use Subversion, it'll save you lots of grief in the long run. Do not use SourceSafe, it's not a real source code control system and is worse than doing this by hand (IMHO).

Coffeehouse - Adobe is clueless sometimes.

glebd — Tue, 17 Jul 2007 10:48:02 GMT

blowdart wrote:

And as I was in firefox it downloaded a wrapped MSI.

Which doesn't update the ActiveX plugin for IE.

Why exactly do you expect a Firefox-specific installer to update your MSIE components as well? This is not a Windows system-level update, right? Seems to me it's not Adobe who is being clueless.

Coffeehouse - Think for yourselves...

glebd — Sun, 15 Jul 2007 19:36:57 GMT

Sabot wrote:

Take for example the iPhone or even the iPod, what killer feature makes them head and shoulders above the competition?

Well, for me it's about good design, taste and functionality. And functionality doesn't necessarily mean cramping every conceivable feature into a piece of hardware or software. This is something Microsoft unfortunately doesn't get. (MS Office 2007, I suspect, is the first version of Office a proper designer had worked on.)

Coffeehouse - What does your team use for SCM and bug tracking?

glebd — Fri, 29 Jun 2007 10:16:31 GMT

I use Bugzilla v3 and I think it's not bad at all.

Coffeehouse - Why I probably won't get an iPhone.

glebd — Thu, 28 Jun 2007 00:40:31 GMT

Ray6 wrote:

glebd wrote:

Why would you need that? You have multi-touch dialling and an excellent contact list. I never use voice dialling myself.

Now this is the real kicker. No voice dialling? Loads of people use voice dialling for making calls while they're driving. Odd thing to miss out. But if they didn't miss out some of the most obvious features, how would they pursuade you to buy another one in two years time?

When driving, you should not use a mobile phone, period. You should pay attention to the road. There are too many morons creating life-threatening situations because they think they can drive and use their phones at the same time. Here's hope that at least iPhones won't cause road accidents.

Coffeehouse - Why I probably won't get an iPhone.

glebd — Wed, 27 Jun 2007 11:14:13 GMT

stevo_ wrote:

^ Your MMS point is stupid, do you even understand the difference between SMS and MMS? It's not a flop at all, someone already stated that it's a big thing in the EU.

Here in the UK the concept of a mobile phone has designed around people taking quick snaps and using MMS to send them around to each other.

OK, let me rephrase it--nobody I know uses MMS, and I live in the UK. The concept is there, yes. But while on average 1 million MMS messages were sent daily in the uk in 2006, the daily SMS count was around 126 million (source). MMS has fallen short of the industry expectations.

Coffeehouse - Why I probably won't get an iPhone.

glebd — Wed, 27 Jun 2007 10:29:46 GMT

> The mobile version of OS X or whatever it is the iPhone runs takes up 700MB of the device's capacity. Damn son!

And you care about this why exactly? I'd rather have a good OS on my phone than several more crappy music tracks.

> There's no way to cut, copy, or paste text! WHOA! Big, big mistake.

On my Nokia Series 60 3rd Edition phone for the entire year of daily usage I can count probably three occassions I've copied and pasted text. It was extremely cumbersome. To make this work when there is no stylus, using your fingers for text selection would be close to impossible. I could totally live without this feature.

> No A2DP support. That, friends, is such a huge bummer right there.

Not a big deal really. People using bluetooth headsets look like cybermen anyway. I've even seen some people wearing BT headsets on the underground--how silly is that?

> Sorry, music can't be used as a ringtone -- even if it's just a raw MP3. No additional ringtones will be sold at launch.

Thank God! No stupid ringtones. This is a huge feature. Good call Apple.

> Pogue again confirms document file reading -- but not editing -- for PDF, Word, and Excel (only).

Who in their right mind would *edit* documents or spreadsheets on a *phone* is beyond me.

> Adobe Flash support is officially out. It's just not in the browser. Neither is there any other kind of embedded video support. Sorry everybody, that's that.

I can see why. This must be all about reliability. They even have a timeout for JavaScript on pages, something like 5s of run time. Nothing you open from the Web must stop the phone from functioning. This is also the most likely reason for the lack of native SDK.

> It will take snaps, but won't record video. How can Apple love YouTube as much as it does and not realize cellphone-shot movies make up a sizeable chunk of the crazy crap you find on there?

I can live without crappy video footage from phone cameras.

> Oh, and no MMS.

MMS is a flop, no one uses that. It was a nice idea, but SMS is so much simpler, qucker and cheaper.

> And sorry, no voice dialing, either.

Why would you need that? You have multi-touch dialling and an excellent contact list. I never use voice dialling myself.

> Contact groups can't be emailed as contact lists.

You mean there is no way to bulk-mail from your phone or share your bulk-mail list with anyone? Give me just one use case for this.

> Apple sez between 300-400 charges the iPhone will lose battery capacity -- you'll send it in and get the cell replaced for a fee. Meh. We knew this would be the case, but still, meh.

Let me see. Even if you charge it daily, it's a year before capacity is lost, and I don't think it'll just stop charging overnight when you plug it in for the 401th time. It's no different from other phones, except you can't change battery yourself. Not very nice, I agree, but the iPod owners out there don't seem to mind.

> Apple can (and supposedly will) be rolling out periodic updates -- no surprise there.

I'm sure many of the drawbacks cites in the reviews will be addressed.

> Battery life is, somehow, almost as mind-blowingly good as Apple claims for calls, music, and movies.

And that's a reason for you not to get an iPhone?

> As we suspected, users are prompted with lists of WiFi networksifyou're not nearby a trusted hotspot. We've seen this on other phones, and we're afraid this would get friggin annoying.

It's like saying, "I haven't seen this feature, but based on other phone implementations I'm sure it'll suck too." Do you see how silly this sounds?

> It's said to be very scratch resistant. The facade both front and rear apparently just doesn't pick up marring like regular iPods do.

So that's a plus or a minus? You seem to have confused your points with the title of your post.

> Voice quality is said to be good -- not great.

This is so subjective that I would not pay attention to it unless they said it's total crap and completely incomprehensible, which they didn't.

It amazes me how people hurry up and bash a product that hasn't been released yet, and some even start selling what they call "iPhone killers". Look how many iPod killers are out there. They all are crap, some better that others but none come even close to iPod.

A Windows Mobile device as an iPhone killer? Gimme a break.

Coffeehouse - So, we're having a baby

glebd — Tue, 12 Jun 2007 10:12:54 GMT

Congratulations! But don't get too fancy with the name, or you kid will hate you later. You may also read the last chapter of Levitt's Freakanomics (http://www.amazon.co.uk/Freakonomics-Economist-Explores-Hidden-Everything/dp/0141019018)--it's about baby names