Thursday 28 July 2005

That’s why it’s Called Beta

I WOKE UP this morning to find countless emails and IMs pouring into my accounts asking me about the IE 7 beta.

Some developers are expressing relief at seeing some of the bug fixes and improvements, but of course as I’ve been expressing all along, this is a process with which we have to be patient. Expecting full bug fixes and implementation in any beta software is ridiculous, as is expecting that WaSP / Microsoft Task Force can perform retroactive miracles.

IE7 is in beta. Not only that, but it’s early days yet. So it’s a little bit premature to start complaining that things don’t work. I mean, why have a beta, much less one that’s made it out first to developers and press if not precisely to get their feedback pronto? Brian Goldfarb, Product Manager for the Web Tools Team and Microsoft’s liaison to WaSP pointed this out in a conversation we had today while trying to address developer concerns as they’ve been pouring in.

“The whole point of doing a developer beta is to identify potential rendering breakages and changes and resolve them before we hand out IE7 to the broader marketplace. We are working actively to identify any issues with actual rendering problems and resolving those. This beta is one part of that mission.”

Browser Sniffing Smells Funny

One of the primary complaints coming in from developers has to do with browser detection. In an article published in PC Magazine yesterday, writer Cade Metz makes the following statement:

“At present, IE7 has a problem rendering some Web pages. According to Microsoft, this is caused by the sites, which need to update their detection code for IE7.”

This seems to be the paragraph that’s upset most developers. Metz could have helped keep concerns more effectively corralled by being less vague about what he meant, or by seeking out details from WaSP or related organizations who can best address this concern.

Specifically, any new browser version, whether it comes out of Microsoft, a new startup or a well respected search engine is going to have to be accounted for in any scripts that use versioning as part of the script. Fundamentally, this issue has nothing to do with IE, or Microsoft. What’s more, standards advocates have argued against the use of browser-specific detection as long as standards advocates have existed.

Dori Smith, long time WaSP member and WaSP DOM Scripting Task Force co-lead responds:

“It’s long been known that scripters should write their code to check for objects, not particular browser versions. The DOM Scripting TF is working on documenting best practices for scripting, and this is one of the elementary building blocks of good coding style. Scripters whose code follows best practices don’t have problems when new versions of browsers are released – and scripters know that new versions will always come out in the future.”

Jeremy Keith, who co-leads the DOM Scripting Task Force along with Dori, added these thoughts:

“As the number of different browsers being used increases, browser-
sniffing scripts become more and more complex. They need to test for
all possible combinations of vendor and version number in order to
ensure that they work cross-platform. This is a Sisyphean task that
can result in extremely convoluted and messy code. Many browser-sniffing scripts test for an exact match on a browser’s version number. If a new version is released, like IE7, these scripts will need to be updated.

“Thankfully, the practice of browser sniffing is being replaced with
the simpler and more robust technique of object detection. Instead of
testing for a browser name and/or name, simply test whether a
specific method or property is supported. Then your tests will
continue to work in IE7, IE8 . . . ad infinitum.”

Both Smith’s and Keith’s words point to some critical concerns. To begin with, there’s the not-so-small fact that in a standards compliant environment, browser sniffing of this nature shouldn’t even be necessary. Web accessibility consultant and researcher Joe Clark points this out in his recent article on the subject, IE 7: The Saga Begins.

Another point is that the need to improve scripting practices on the developer side of the fence side is imperative. To fix that really is up to us as individual developers as we strengthen and hone our skills based on growing experience and awareness.

WaSP and Microsoft in a Tree

Scripting issues aside, let’s examine the role of WaSP and Microsoft IE, an issue that many are understandably concerned about. I can’t help but once again ask all readers to take a deep breath with me. WaSP’s direct involvement with Microsoft is brand new. Had we been in months ago, maybe IE7 beta would be different. Maybe we’d be further down the road to success. Maybe.

No matter, we weren’t there, and that was due to a lot of factors. So we got down to it and changed that. Now we’re here, and we all want to do what’s right. But anyone who thinks this is going to happen overnight is not being realistic.

As a fellow WaSP Microsoft Task Force member bluntly pointed out to me as I was trying to strategize how to respond to upset developers, WaSP should never act as Microsoft’s public relations department. And he’s absolutely right. WaSP isn’t here to forgive Microsoft for past practices.

However, as the relationship person here, I can only do my honest best to communicate both sides of what is clearly a complex concern. I can only work to assure you that I, and everyone within this Task Force is extremely motivated to make sure we keep things positive, honest, and respectful so we can continue to work together and hopefully, once and for all, achieve the goals we didn’t succeed at before.

So What Do We Do?

WaSP’s continued effort to work with rather than against Microsoft at a very frustrating time in history means that we all have to have patience, and we have to ask everyone to have patience with us in kind. This isn’t easy for anyone, not the Microsoft developers, not WaSP as an organization and of course not the working Web designer and developer.

WaSP liaison to the WaSP / Microsoft TF DL Byron points this out in practical terms:

“It’s much easier to criticize Microsoft than actually engage them. The constructive thing to do is to respond to the beta team and lay out your concerns.”

Microsoft already knows which existing IE 6.0 bugs need fixing, and what needs to be implemented to come up to full compliance. What you can and should do if you are on any beta for Microsoft is use the avenues available to you to identify real rendering issues and bugs and submit them. If you’re not involved in betas, drop by the IE blogs and let the developers there know in practical, respectful terms your constructive criticisms.

Filed under:   general
Posted by:   Molly | 15:21 | Comments (49)

Comments (49)

  1. I’m with ya, breathing deeply, slowly, patiently. And getting a nice cup of tea to sit out the ongoing battle from all the other people. :-) Have some wine, Molly!

  2. Pingback: Brian Goldfarb's Blog

  3. Judging from the mezzoblue review, this beta should really be considered an alpha if MS is planning on adding significantly more standards support.

    A beta should be largely feature-complete. The purpose of the beta is to ferret out the bugs in the features. Are we going to have another beta to examine the bugs again when even more features are added. And you are asking for patience?

    We’ve been patient enough. There are already other browsers available from other vendors that are much more capable than IE 7.0. They’ve been available for years.

    Microsoft may have won the deployment war, but they lost the browser battle. IE 7.0 will compete with Netscape 8.0 for the Why Bother? Award.

    Get Firefox, indeed.

  4. Pingback: Rimantas’ murmurs » Beta bag of bugs?

  5. Pingback: Elliott Back

  6. Jon Wiley,

    I think you have the concept of a beta wrong, although you are almost right.

    At least in Microsoft, they release 3 betas, each beta represents a milestne of features that are cmplete but not bug fixed. Beta 1 implies on some features (partialy) complete, beta 2 represents many features complete, and beta 3 implies almost every feature is complete and bug chasing can begin.

    Then MS goes into RC (Release Candidates) where all features are done, and remainder bugs (often unknown at the time of RC release) are being fettered out.

    Beta 1 of ANY Microsoft product is often 18+ months from being finished. To even expect a near final product at beta 1 is just plain silly.

  7. Molly to be frank this is not accurate and misleading. IE7 has some serious issues that sometimes breaks that browsers outright and does not break IE6, regardless of object detection or sniffing for the specific user agent string.

    Its one thing to say well use the correct methods and it will be well, its quite another to actually get your hands dirty and test if indeed that is the case.

  8. Your comments on using object detection instead of user agent string detection are true and have been echoed by others. Obviously, we would all like it if we never had to use them. That all browsers render things the same is a highly utopian dream (as you no doubt fully understand!).

    Where the frustration comes in as a developer, is that there’s no browser detection capabilities in CSS. Sure, browsers should ignore stuff it doesn’t understand but it doesn’t ignore stuff it simply gets wrong.

  9. I got to play with the IE7 alpha…and let me tell you…. There’s a reason they call it alpha as well!

  10. > So it’s a little bit premature to start complaining that things don’t work.

    Wouldn’t you say though, that there’s a *huge* difference between a beta that has virtually no improvements to its rendering engine, and a beta that has at least some significant improvements, however buggy they may be?

    In four years, Microsoft have managed two CSS bug fixes (not even new properties!), one PNG bug fix, no HTML improvements whatsoever and no DOM improvements whatsoever.

    Is it really likely that they are going to implement anything significant before the final release of Internet Explorer 7? Surely the first beta means they are well over halfway to wherever they think they are going, doesn’t it? So why should we expect anything significantly more?

    > Microsoft already knows which existing IE 6.0 bugs need fixing, and what needs to be implemented to come up to full compliance. What you can and should do if you are on any beta for Microsoft is use the avenues available to you to identify real rendering issues and bugs and submit them.

    But as you yourself just pointed out, Microsoft already knows what the real rendering issues and bugs are.

    If we mention them, we get accused of harping on about stuff Microsoft already knows about.

    If we give reminders to show it’s still important to us, we get accused of being vague and not giving specifics.

    So what’s the alternative? Shut up? Then Microsoft will just say that standards weren’t as important as other things.

    What is the appropriate way to say “the CSS, HTTP, HTML and DOM issues are a real pain and you know exactly what bugs you have and I want you to fix them” without a) telling them what they already know and b) feeling like you are talking to a brick wall? And, often, c) get people complaining that you are doing it on a weblog post that isn’t about those things, because the developers never make posts concerning standards?

  11. In case people are looking to browser sniff IE7 and bugging you about that, I released some code to do that.

    The IE blog, is a bit misleading as the wording for detecting IE7 currently reads as if you can only detect it on longhorn. It detects fine on XP too, even without all the winfx stuff.

  12. Pingback: » Archief » Eerste beta IE7 onder luid gejuich ontvangen

  13. My concern is not just that IE7B1 is little more the a poor interface makeover for IE6 with 2 major CSS bug fixes, it’s that Microsoft has so far withheld, and will likely continue to withhold, any and all information (with the exception of a few vague hints about “improved CSS2 support”) regarding exactly exactly will and will not be supported in the final product. It’s little wonder that, since we’ve been kept in the dark for so long about these details and given this as our first real hint about what to expect, this release has been met with outrage.

    There’s also been little information about the development cycle (e.g. a roadmap?!) from which we can determine the progress of the overall product. Combine that with the confusion caused by the comparisons of an IE Beta with Firefox/Opera Betas with regards to feature completeness, the knee-jerk reaction from many people seems quite justified. Although if Travis Owens’ comment above is correct, there may actually be faint glimmer of light at the end of the tunnel to look forward to.

    On the other hand, while I’m disappointed that the vast majority of us who aren’t MSDN subscribers generally can’t get access, it’s not a major concern to me; as being able to try it out for myself for a couple of days would be nothing compared with actually having some information about what features it does, does not, will and will not have, and approximately when we can expect to see the results.

    IMHO, Microsoft needs to release this information; they need to release a lot of it and they need to release it ASAP. It’s the only way, in my opinion, that Microsoft/the IE Team can earn any respect and the patience you claim they deserve from us.

  14. Travis,

    No he doesn’t have it wrong. That is the industry accepted (read: everyone but MS apparently) understanding of alpha, beta et al.

    I am just all rosy that Microsoft operates the way you layed out, but brass tacks is that yet again they are doing things their way with no regard for the rest of the world.

    This release is an Alpha to the rest of the world, period.

    Good article Molly, I do not envy you this task.

  15. Pingback: TechBlog

  16. Does’nt MS follow the industry standard : there is the alpha releases and after that the beta releases that are “feature freezed” ? …

  17. Standards are not part of the IE product pipeline – at least not for an IE7 release. I’m sure the developers are keen on getting this right, hence The MS Task Force’s involvement, but at the end of the day what real say do the developers have at a company like MS? I would wager not much.

    Remember that IE 7 cropped up as a surpise after MS had told us there would be no updates until the next OS. What were the reasons for this change of heart? Firefox, correct? So MS’s priorities are tabbed browsing, RSS support and a version number bump to give it some publicity. Standards were never involved.

    It’s going to take a bit more time I think. As the MS Task Force will tell you, things are moving forward, but no where near as much as they or any of us hoped unfortunately.

  18. WaSP could influence Microsoft, but there is no backbone within WaSP to do it. Microsoft flattered you with a bit of attention and all of a sudden your attitude is to wait and give them time. How long do we have to wait? It took 5 years for Microsoft to get from IE 6 to IE 7. Do we wait another 5 years and hope that IE 8 will be standards compliant?

    IE 6 did not get 90% of the markup share because it was the consumer’s choice. IE’s market shared was obtained through the abuse of monopoly power.

    Microsoft is smart – by appearing to be open to WaSP’s input, they’ve succeeded in sapping WaSP’s independence (…keep your enemies closer) and you are all so flattered to be “working with” Microsoft that you can’t see this.

    WaSP needs to drop its relationship with Microsoft right away. They’re not listening to WaSP or anyone else. Their beta is not a balloon they’re sending up to see how folks like what they’ve done, nor to encourage feedback. A beta is a beta – that is the development is FINISHED. They’re only looking for bugs.

    So this is what WaSP needs to do: 1) find its backbone again and 2) lead a LARGE campaign to get people to stop using IE. Successful boycotts are the only leverage Microsoft will understand.

    In the past, WaSP urged people to stop using Netscape 4 because it was not standards complaint. If you could do this for NS 4, why can’t you do this for IE? Oh right, they’re your friends now! You’ve fallen for the oldest trick in the book.

  19. Thanks for a great post, Molly.

  20. Why all the fuss about IE7. There are already great browsers available now. There is no reason to expect IE 7 will be nothing more than a new incarnation of the piece of garbage that the current IE is. IMHO, Microsoft is only interested in market domination, not good software.

  21. Pingback: IE7 Reactions | K-Squared Ramblings

  22. Pingback: IE 7 « dave j lowe

  23. You clearly don’t know what ‘beta’ means.

  24. Pingback: IE7 is IE6 combined with Firefox and Safari @

  25. You state, “So it’s a little bit premature to start complaining that things don’t work. I mean, why have a beta, much less one that’s made it out first to developers and press if not precisely to get their feedback pronto?”

    However in this case complaining is the feedback. If there are glaring holes in what the browser should provide, then now is the time to point those out.

  26. Whether it’s Beta, Alpha or Gamma is completely uninteresting. It’s just different interpretations of definitions.

    Standards compliancy is really the big issue. That is what standards are for. A FireWire device plugs into your computer and it runs. There are no five different interpretations of what it should do. It transfers data according to a standard.

    All browsers should adhere to the same standards. Those standards are set. They are clear.
    More are on the way, and that is a bit scary. As already we are seeing a little implementation of CSS3, while it is not finalized, while only iCab (to the best of my knowledge) is the only distro available which passes Acid2, and thus the only one that does CSS2 right.

    The war for the customer should be fought on speed, features, looks, interface-design.

  27. Huh?

    “Microsoft already knows which existing IE 6.0 bugs need fixing, and what needs to be implemented to come up to full compliance. What you can and should do if you are on any beta for Microsoft is use the avenues available to you to identify real rendering issues and bugs and submit them.”

    Microsoft already knows about all the bugs, so instead of complaining to Microsoft about the bugs, we need to let Microsoft know about the bugs?


  28. Isn’t ALL Microsoft software still in beta :)

  29. Molly, next time you have a meeting with Microsoft, I challenge you to ask them directly if they plan to support application/xhtml+xml MIME type or the ABBR tag or ACID 2 test, etc. in the final release of IE 7. If you don’t get a simple Yes/No answer, then they don’t respect you.

    If you are telling us to be patient and wait for the final release of IE 7 and Microsoft does not deliver on these missing features, then you will loose our respect.

  30. To me, beta means a bunch of unpolished features, not a near complete lack of new features.

  31. This is a bunch of *insert 4 letter word here*!!

    Sorry but this is Microsoft we are talking about. They can get it right and they should. Why should we cut them a break?

    Okay lets look at Firefox – yes it was in BETA for one hell of a long time. But as we all know the early BETA versions of Firefox were fantastic!

    From the reviews I’ve read Microsoft have continued to build ONTOP of a dead platform (IE6). Why are they doing this? For a while they said there WOULD be no standalone browser, that the new RENDER engine would be built INTO Longhorn (Vista). My reading now makes me think they have NOT done this they have taken the IE6 render engine started putting MORE bandages on it to FIX it up – from the pressure of the community and WaSP.

    This is NOT the approach they should take. They’ve had 4+ years (6 if you go back to the start of IE 6). They have had all that time to rebuild EVERYTHING.

    I want to know what the HELL they have done with ALL that time? Apple have built Safari from ground up (okay it’s built ontop of…but yea) in that time. Why can’t Microsoft take a leaf out of that book?

    YES it’s a beta, but BETA’s are NOT usually there to continue ADDING new features in. Beta generally are 90% finished applications – it should be for testing implementation of these features.

    This is just rubbish … Molly I’m sorry but WaSP involvement with Microsoft really excited me – but I’m begining to feel let down … and reading your comments and other posts I think the whole community is feeling let down by the 1 body (WaSP) who could have turned this mess around!!

  32. Molly,

    This page

    coming right out the archives of still fails to be rendered correctly in MSIE 6. So what has been doing during all these years if it was not to provide a clear testcase/demopage of the bug which should have been fixed in 1997!

    This other page,

    coming from the archives of is not rednered correctly in MSIE 6. How much more should we wait?

    Molly, MSIE 8 final release will be released in 20012, possibly in 2013. You can mark my words.

    Gérard Talbot

  33. Even if Microsoft would fix half of the documented, reported bugs and implement just half of the missing, incomplete support for CSS2, DOM 2, and other expressed/formulated specific requests in the final release of MSIE 7, there would be major problems for web developers.

    1- People do not automatically upgrade to the latest available browser release. Adoption/migration of more recent browser release versions is a slow process and takes years. Studies have shown that a majority of people usually go ahead with a browser upgrade along/at the same time with an operating system upgrade. So one should expect to see MSIE 6 to represent a large chunk of the browser market for the next 5 years or so.

    2- Many windows users will not be able to upgrade to MSIE 7 anyways since MSIE 7 will be possible only for XP users and Windows 2003 users. A considerable minority (20% – 30%) of the windows market will not be using MSIE 7 anyway during the next few years.

    So, regarding web standards support and compliance in browser usage, developers will have to deal with an overall fragmented browser picture in the years to come. That’s at best.

    Now the only remaining question is how well, good, complete, how much will the MSIE 7 final release fix the utterly explained, described, documented CSS/DOM bugs, unsupported DOM 2 attributes, methods, unsupported CSS 2.1 properties? This beta 1 release is sending the message; it suggests me that we should not expect more than a 50% effort from Microsoft.

    So, speaking of patience means in my mind that things will be fixed completely and for good with MSIE 8 … around 2012.

    Gérard Talbot

    The link in one my previous post should have been:

    Developers gripe about IE standards inaction
    October 2003


    Why You Should Dump Internet Explorer
    June 2004
    written by Daniel Miessler, a Microsoft Certified Software Engineer

  34. BTW, Molly, that domain doesn’t seem to be associated with Microsoft at all, as far as the WHOIS information goes (just some Chinese guy). Better link to just to be on the safe side – don’t want it turning goatse on you in a week’s time.

  35. Pingback: Joe's Apt.

  36. Pingback: Rimantas’ murmurs » Beta bag of bugs?

  37. Molly –

    Despite your own recognition of the danger of being co-opted by Microsoft – it was a Venus Flytrap. You are listed as an apologist for Microsoft here->

    Jacques Surveyer

  38. Pingback: Afkalkeren » Standards and CSS in IE

  39. Pingback: Lifehacker

  40. Seems to me complaints now are far more likely to see resolution than complaints after the release. When *is* the right time to commence, O wise mistress?

  41. “fact that in a standards compliant environment, browser sniffing of this nature shouldn’t even be necessary”

    That’s better said than the writer of the site, since a standards-compliant site still needs browser sniffing when object detection doesn’t handle small quirks certain browsers have. Just because the object exists, doesn’t mean that any particular browser follows the CSS standards correctly, etc. Things need to be rearranged for certain browsers to get around bugs. IE, for instance, has many quirks with CSS. It comes down to sometimes having to write to the DOM differently for each of Safari/KHTML browsers, Gecko browsers, IE, and Opera.

  42. To clarify the above, I meant “That’s better said than the writer of ‘IE 7: The Saga Begins'”

  43. Pingback: » WaSP Microsoft Task Force Update: Upcoming Products, XAML, Acid2, SXSW, and IE7 Revealed

  44. Its a shame microsoft don’t appreciate the term, backward compliant!

    It drives me nuts that MS expect developers to keep up with their shabby coding. If they took the time to ensure everything worked prior to launch they wouldn’t be under so much pressure from our friends at firefox..

  45. Thanks for clarifying that netdragon was wondering what you meant.Looks like the abouve two commments Appartments St.Martin and freeware downloads are comment spam Molly I ll send you an email incase you miss this post.Love your blog hate to see it spammed!

  46. reviews I’ve read Microsoft have continued to build ONTOP of a dead platform (IE6). Why are they doing this? For a while they said there WOULD be no standalone browser, that the new RENDER engine would be built INTO Longhorn (Vista). My reading now makes me think they have NOT done this they have taken the IE6 render engine started putting MORE bandages on it to FIX it up – from the pressure of the community and WaSP.

  47. Pingback: WaSP Microsoft Task Force Update: Upcoming Products, XAML, Acid2, SXSW, and IE7 Revealed - The Web Standards Project

  48. Pingback: IE7 Beta 1: The Vultures Fly High • Blog Archive • Joe’s Apt.

  49. This is the latest and hottest ghd styler ever. If you need a ghd hair straighteners, this is a must buyghd hair straighteners,cheap ghd hair straighteners,pink ghd hair straightenersghd straightenersComme vous pouvez le voir, il s’agit d’une paire de chaussures shox classiquepink ghd hair straighteners . Si vous souhaitez poursuivre la mode, nike shox NZ sont votre meilleu…nike tnCette paire de Nike Shox Torch est chaud en maintenant la chaussurestn chaussures

Upcoming Travels