I still don't understand why Microsoft doesn't print human readable error code instead of just "0xc0000005".
To figure out what the actual message is, you first have to figure out if you are seeing a HRESULT or a NTSTATUS[1]. In this case the leading 0xC is an invalid HRESULT, so it's for sure an NTSTATUS. Then you just look up the code in the bottom two bytes [2].
In this case it's just error code 5: "STATUS_ACCESS_VIOLATION", which is a memory access error, which typically means it's a bug where the program accidentally is trying to read outside of its memory space.
So, for sure an application error and not something that the user did or some sort of system issue.
There are also "facility" codes which can give you more information what where the issue is happening, but this code is just using the default.
I suppose this is better than when they would print the decimal value instead of the hex. "WTF is error code 3221225477?"
A zillion years ago I was able to see some of the internal communications around the development of OS/2. One topic was the way that non-system floppies would (fail to) boot when left in the PC. What the user would see was something like (I forget the exact numbers)
SYS!012345
SYS!876543
...and they were supposed to go look that up in some reference book to see what it meant.
The question was why they weren't outputting something human readable like "Unable to boot from non-system floppy". And the answer to that question was that, as an international company, they couldn't expect everyone to understand English.
So why not output multiple different languages? Well, the code to output the message has to fit in the floppy's boot sector, so the space is very tight.
And so there was discussion about whether it makes sense to give up having only 99% of users understand it, and instead have 0% understanding.
I think the killer argument was the person who suggested outputting a message something like
______
| |__| |
| () | X
|______|
which would just barely fit, and I think is almost universally understandable. Of course, the inscrutable message eventually stuck anyway.
This kind of pictograms always reminds me of Apple. When a Mac can't boot because of corrupt boot files, it shows a "prohibit" sign.
I suppose the reasoning was the same—not all users can understand English. But in this day and age a pictogram is in fact harder to search for because people describe it differently.
A company that valued good design would have went with the floppy image. It's a very Microsoft thing to go with the inscrutable error.
On the other hand, this is the OS/2 project, which Microsoft did as a political hedge instead of a true desire to invest in that particular future. I don't fault them for not overthinking their decisions.
> which typically means it's a bug where the program accidentally is trying to read outside of its memory space.
I tracked this down with process explorer the other day. What I determined is that outlook is trying to load it's DLLs but the permissions on the FS prevent the DLLs from being read by the user.
I reset the permissions for all subfolders in C:\Program Data\Microsoft Common and c:\Program Files\Microsoft\Office and it started working again for me.
I would've expected a DLL load error to be ACCESS_DENIED instead of ACCESS_VIOLATION because the path for the DLL goes through a system call to open the file.
ACCESS_VIOLATION would be more like dereferencing a null pointer, but it's been a number of years since I had to troubleshoot this far in on Windows.
I suppose it's possible they didn't check for the error in the open file syscall then tried to use the uninitialized file handle for the DLL which caused the segfault.
Same for their OS, it seems to ignore power users completely. Just a few months ago I was going to connect to an SMB share by using \\nas\Downloads and it just wouldn't work, started a network troubleshooter that couldn't troubleshoot anything.
If it had just told me that the name could not be resolved I would have figured out that I needed the full domain suffix because ipconfig wasn't showing it.
This issue goes way back in Microsofts history, always getting on my nerves as a power user. But at the same time probably the best OS for any other type of user.
> In this case it's just error code 5: "STATUS_ACCESS_VIOLATION", which is a memory access error, which typically means it's a bug where the program accidentally is trying to read outside of its memory space.
What would be a (more) user-friendly error message for this though, given the quite technical nature of the error?
For printing the codes? The programmer in me says something like:
"Error: 0xC0000005 - STATUS_ACCESS_VIOLATION - (Error, Microsoft defined, unreserved)
Error: This is an internal error that prevented normal operation of the application. Please contact support at X."
Which conveys all of the information in the code and tells the user what they should do and that it wasn't their fault. Then it's a different story to say which code would be the most useful for the user.
They could have the error code, name and a small explanation about what the error means. Eg.: The program tried to access an invalid memory address and had to be terminated. That is likely a bug in <program name>
I wonder if large companies get more of a pass with outages then small ones? Bias towards authority in a way.
If a service has a big enough base an outage could almost seem like a pass for everyone. “Oo couldn’t do that this morning GITHUB was down.” Whereas if my small GitLab server was down bosses could be “well why aren’t we on GitHub like everyone else, it’s up now?”
When google calendar was down a bit ago I tried looking at random people responses on twitter and many were “whew I get out of some meetings!”
Perhaps there is some accountability for large scale outages like these but it really feels like consensus is often “shit happens” which is totally reasonable just seems like it could hurt big companies a lot less.
I think an interesting counterpoint to my observation is Firefox claimed they lost a lot of users because Google services just broke in odd ways only on Firefox so users moved to Chrome.[1]
My counter to that would be, even though Firefox was and is a big company it's easy to change. A lot of services like Outlook, Chrome, or Gmail or just downright colossal at this point. Changing from Gmail isn't easy because of all the other services and logins associated with it.
Not to mention, Gsuite, GitHub, Outlook, and such are all very corporate. Higher ups make the decision as to when to use these often enough, so you can't just say "this has kind of failed us a couple times lets try something else".
So all in all, I still stand by my point, entrenched big companies probably get more of a pass.
> Changing from Gmail isn't easy because of all the other services and logins associated with it.
You can keep the Gmail account and start using a new one somewhere else. You start registering to new services with your new account and gradually migrate old services to it if they let you. You end up with gmail for social logins (which hopefully you can migrate too) and little else.
I do have a Gmail account. I use it to login in Google when I work for a customer which use the Google Cloud, when I have to upload videos to YouTube (I logout after that) and for Google Play.
> You start registering to new services with your new account and gradually migrate old services
Try doing this with with thousands of employees with multiple devices and varying computer skills, at the same time, while keeping a business running. Don't forget to factor in multiple 2FA systems, that you also have more service accounts than you probably think, and that your spam filters will change behaviors (for that added bit of uncertainty).
Congratulations, you've identified a market need. If someone builds a tool to do all this (if the market isn't saturated with them already), companies will pay for it. There's plenty of tools to convert to/from Microsoft Active Directory deployments or varying quality, I don't see why Google should be any different.
If a company married Google or any other similar service provider I expect that the migration in case of divorce is going to be expensive. For a single person not so much.
I agree, but I think there's also a bit more to it.
Thinking specifically of cloud infrastructure, if you run your systems on infrastructure that is also used by your largest partners and customers, then your outages will tend to coincide, so (a) nobody cares that you're down because they're down too, and (b) even if they did care, they're busy enough worrying about their own systems that they have less time to come at you.
An example would be a company specializing in Shopify "apps" running their systems on GCP because that's where Shopify runs... if GCP goes down then Shopify itself is also likely down as well, and you'll get more mileage out of the "sorry, GCP is down" justification (if it's even needed).
Yeah that’s a good point, it also shows up in weird ways. Like when Facebook auth went down a slew of unrelated apps did too because they relied on it.
When I was younger I really wanted one thing to do everything, when chat was integrated into gmail is was like “oh boy, one less app!” And then similar feelings when it connected to SMS... now it’s just “great everything is getting more connected and homogenous and I seem to have less and less options”
I think with a smaller product you're trying to fight for market share and retaining a client - Microsoft is so large and with so much momentum and dug so deeply into their customers that this sort of thing doesn't cause them to want to find something else. If it hurts to change, it's easy to excuse something like this.
I think it's more "momentum". With a smaller company that affects a small number of users, the possibility of cutting your losses is quite feasible. On the other hand, with Outlook you've got your whole org using it, your servers running on it, your customers integrated with it, custom rules setup - a cut and run is just not simple.
Honestly though, I really don't get why rollback isn't more common for these larger pieces of software. Just some logic to detect multiple crashes, check if a rollback would break anything (i.e. dependencies) and then rollback. If you remove so much power from the average user in your OS, you ought to really have some amazing automated procedure for these scenarios.
I thought they also do limited roll-out as well? I think I remember the Windows anniversary bricked a bunch of machines in the limited roll-out and they just pushed it out anyway?
The complete lack of control as to when an update is performed is astounding - I know there are some settings to stop this, but apparently there are cases where that setting reverts back to auto or you need to switch it back on in order to manually update. I remember the last straw for MS and myself was accidentally clicking update just before a meeting whilst on very low battery, with no way to back out and no charger on hand. It didn't end well.
These days if something is "Windows only", it goes in a VM along with other software I don't trust.
I don't really know of any small services I gave up on because of outages more or less than I would for any other service regardless of company size....
The complexity of a product probably makes folks less likely to 'walk away' but I'm not convinced they do more or less for any other given company based on size alone.
MS definitely has customers like hospitals and the military that will be very upset and frantic in any outage.
At least in the days of yore MS distinguished itself with support for that kind of client, famously sending service technicians in helicopters out to remote installations.
It would be interesting to hear, in the new era that they're in, if that kind of thing continues.
One angle to look at this from is that if Slack or GitHub goes down, there's a good chance your competitors are being held up too. It's like an armistice on (your corner of) capitalism. Whereas if it's just your system, you've been put at a slight disadvantage.
Also, unless it's happening every week there's probably an element of, "well it doesn't make me regret the overall choice to use this service, so there's no sense being upset about it".
The other opposite angle to look at it from is that if slack or github goes down but you aren't using them then you have a slight advantage over your competitors.
Most of the customers are too dumb to really care, and if they did, what would they do
It will take you 3 years to unwind your Microsoft business. Few people with that level of rage over poor Windows/Office software can maintain it that long. Normally, companies yell at the IT Staff, unless it's a really bad problem, then they yell at the TAM or account manager, who takes some abuse, wears an appropriate sad/chastened face, and talks about the Microsoft release ring framework, and that most problems are your fault because you don't follow a similar model.
In my experience recently with O365, these issues usually are an update issue related to some regression in Windows 10 combined with an Office issue, or changes to Microsoft's authentication infrastructure. Microsoft has a unique understanding and methodology for making things good enough to persist -- they don't need to sell the product anymore, just make it not suck too much.
The first issue is because no human can keep a product working that has like 50 different active releases while the company prioritizes pushing out slop. The second issue is usually tied to a product release or revision, combined with new infrastructure. (Especially if it's in new IP space)
Well for many end users at large companies they won't have the authority to do so and will have to wait for their support group to do so, assuming they let the update through.
Some have a victim pool, er, test pool, to help protect from such as this
No worries at all - they didn't even mention that the article had been amended, so most folks finding this HN thread now will likely feel just as you did. I appreciate your comment, as it alerted me to the fact that there is now an all-in-one page with a good overview of the issue and workarounds that I can share with others.
It's worth noting that this appears solely to be a failure with Office 365/Click-to-Run-based installations. If you buy your software instead of subscribing to it, and service Office with traditional Windows Update mechanisms... everything is fine today.
This is what happens when you rip out well-tested IT infrastructure that has a reasonable update distribution process and entrust the cloud to update your software for you.
CtR is this containerized install-from-web platform that Microsoft came up with for .NET applications. Yeah, it's actually a lot of years old. But generally speaking, if you get an Office disc or ISO file (say, from Microsoft's Volume Licensing Center), and install it, you get a traditional Office install that gets serviced via KB downloads from Windows Update. If you get Office via a web download from office.com, you get CtR, which self-updates itself sort of like modern web browsers.
A huge difference for business is that Windows Update allows you to approve/deny updates and cache them on a local server with WSUS, whereas CtR leaves each machine to go talk with Microsoft amongst itself on what to install and mostly cuts your IT staff out of the loop.
I can spend a lot of cash up front on Exchange and Office, then deal with constantly patching them. Or I can spend only per seat/per month and the patching is automatic and the service is likely more reliable. After 3 years the costs are almost the same. Its not like my private environment is so much more stable than an MS datacenter. And the software has the same amount of bugs.
I've seen about ten major Office 365 outages since my last major Exchange upgrade? I don't know how beliefs about cloud service uptime being good persist unless your IT staff were really, really bad at their jobs. My house is more stable than a MS datacenter. And by and large, I actually like Microsoft.
I would not call this a major outage. The service still worked, just not from a single client app, many apps still worked as did the web interface. My house is more stable too but my office is not. I've had UPS systems fail. I've had ISP outages. The overall up time is better, in the case of Exchange, than I could make it at the office - especially considering the 24/7 nature of email today. We need fewer IT people because we offload some of the work. I don't want to be Rambo.
I would say I've found "cloud engineer misconfigures something that knocks the whole service offline" happens more frequently than "my building has a hardware, power, or ISP failure". To some degree, I feel like cloud services have, after overengineering their redundancies for such common failures... introduced a far worse failure via their extremely sophisticated update cadence and orchestration systems.
I'm not a huge fan of cloud in general. Office 365 provides a lot of value for our organization. In addition to Office and Exchange we get a bunch of other things like Sharepoint, Forms, Flow, etc.
I had terrible problems with hosted Exchange from Intermedia. I've had pretty good luck with Exchange on Office 365.
Office 365 support is absolute garbage, they really want to punish you for contacting them, especially so if you've found a bug.
Microsoft switched to no longer providing updated ISOs for Office 2019 and actively pushes Office 365. Plus their package management and update tools for enterprises are riddled with their own issues that makes managing hotfixes and updates a major PIA.
Haha what? No, never. This just doesn't happen in the software world. Besides, Microsoft has some terrible SLAs. Even with Azure it is almost impossible to get a refund for outages that are entirely Microsoft's fault.
What is going on with Outlook for Windows? For years, it froze its entire UI whenever there was some network lag on the IMAP server, and the recent improvements for Outlook strangely focus only on their Mac, web, and mobile versions. Does the Windows version still have a future? It seems to be stuck in some maintenance limbo.
I think it is probably a very large program but does have some legacy cruft they probably can't fix easily. Either because backwards compatibility for COM or due to that if you rewrite something to be async the callers also need to be async and callee need to be non-blocking[0].
However, Outlook is very much actively developed and myself being a Premium user am receiving new features now and then, like UI refreshes etc.
It still freezes (as in says “not responding) for ten seconds when I type the first character in the search/filter box above the email list. How does something like that to pass through testing?
Something that scales with mailbox size, network latency or similar is happening on the UI thread in Microsoft Outlook. This boggles my mind every time it happens, which is every day.
Something tells me Outlook is a huge pile of legacy C++ that no one dares rewrite any major part of even in the face of issues like this.
If you look at the release notes[1], Outlook for Windows is getting a lot of love too. The majority of updates are performance and bug fixes, but the UI issue you mentioned highlights the need for this type of TLC in the Windows client. That said, the frequency you receive these will be pretty dependent on which release cycle your install is set up for, as it can range from monthly to semi-annually.
The Mac, web, and mobile versions are getting a whole lot of feature releases in comparison to the Windows client, but virtually all of them are intended to bring those other clients to feature parity with the Windows client. I've interpreted this as more of Microsoft treating these alternate platforms as first-class citizens finally, rather than as neglect of the Windows clients.
This also isn't limited to just the Outlook client - it's a trend across the entire Office suite. For example, Power Query[2] is a really neat and powerful ETL system hiding within Excel (as well as used in PowerBI, and originating in SQL Server Analysis Services). It's been available for years in Excel for Windows, but only recently became usable on Excel for Mac thanks to a major refactor[3] to port it to .NET Core and strip out Windows dependencies. Eventually this will reach feature parity with and replace the legacy Power Query code in the Windows Excel client, then all net new development will be cross-platform and available on both clients simultaneously.
All things considered, I view the resource allocation across the platforms as a positive. Presuming the feature development happening for the non-Windows clients are as similarly cross-platform as the Power Query effort, it'll eventually lead to cleaned up Windows clients with decades of legacy cruft refactored out, feature parity and consistency across platforms, and substantially greater velocity as those resources currently split between platforms eventually consolidate and focus on a unified, cross-platform, and cleaned up codebase. And if the refactors continue to prioritize .NET Core, it also bodes well for an official Linux client eventually.
If you're on Office 365, the Outlook web client is great (https://outlook.office.com). In fact, I haven't used the full client since 2015, except for rare cases.
Or, you can enjoy the quiet while email is down :)
Would have to disagree. With any ad/tracking protection or non-chromium based browsers, the web client is not very good. For MacOS, it constantly requires reloads in order to get different folders, emails, or other basic functionality (Calendar, Categories, etc) to work correctly. The text editor is completely random in what it decides to do with formatting, links automatically have a gigantic preview link inserted into the text body, the UI is very much so different than the Outlook Client UI and does not port over any of the Outlook Client items (this is more a design decision in general from Microsoft on what is client side vs what is server side...)
The Web Client in its current form is functional, in that I can write and receive emails, but the experience is greatly diminished. Don't mistake me for liking the Outlook client (I am a very modest email user, no more than 400 some per day + a Tetris-style calendar), but this load crashes the client on a daily basis, on start up it takes a good 30 seconds before the client becomes responsive (with multiple local profile rebuilds), and sometimes just crashing randomly when reading certain HTML formatted emails.
Outlook is an exercise in frustration in general that unfortunately my entire company has built its foundation upon. The core ideas are great, but the execution on all fronts is awful. I'm actually almost disappointed when I open the mobile app and it __hasn't completely changed some UI/UX element__ as daily changes are the norm for me.
(all this being said, I'm shocked that Gmail somehow took Outlooks unusualness as a challenge and designed a worse UI/UX experience...)
Hmm; Maybe I'm not a power user of Office 365 Outlook, but I'm on openSUSE Tumbleweed, running Firefox and that's all I use for work e-mail. I think I have Chrome and Chromium on here, but I haven't used them in months.
At this point, if the OS wasn't an issue, I wouldn't use the full client, ever (for all of the reasons others have provided). The only gremlin I've encountered, routinely, is a situation where the rich editor doesn't see a "space" until I've typed a character after it. It's mildly distracting at worst, and far less of a problem than typing and having the thing just sit there like it'll get to my input as soon as the spirit moves it.
On that note, really, all of Microsoft's web products have been pleasant to use. They've become good enough that I rarely end up in the LibreOffice equivalents. The "Teams" client for Linux (in Preview; I run the Insiders version) integrates well -- clicking links from my calendar opens meetings. Some things are a little off, but are a few settings away from being ideal for me. It's lacking some of the features of its more developed Windows client, but it isn't as bad as the mac Lync client was back when I last had an opportunity to experience Microsoft's conferencing platform on non-Windows.
... apparently you have to know that your (personal) Office 365 account is no good here? I thought maybe Microsoft cleaned up their confusing mess of logins a few years back. Perhaps not!
> You can't sign in here with a personal account. Use your work or school account instead.
I would prefer to use the website, but I never got desktop reminders to work properly (i.e stay on screen and not hidden behind other windows). That's on macOS. Calendar and tasks without reminders is pretty much useless for me.
This is probably a good time to ask: Are any of you receiving your checks from Microsoft for all the beta (and, apparently alpha) testing that you're doing? Mine aren't showing up. I'm not sure who to contact about it. /s
In the calls that I've gotten my tiny little Customer base has had something like 25 people "down" today. At least ActiveSync on phones and OWA were still working. I know some people were miffed that full-blown Outlook was unusable because it's a major part of their workflow.
How did this get past QA? It's simply mind-blowing...
I always thought I really disliked Outlook ( because of so many issues) until I worked for this well known startup and started using Gmail. Man, how I miss all the Outlook issues... at least it was good for reading/writing email.
Disliking something can be quite relative I guess...
What's wrong with it? Just curious, I use Outlook for work and my main pet peeve is that
i) local search isn't great and
ii) in general, the application UI is very 'delicate' -- e.g., I have mail rules that move mailing lists to a folder. I browse through them from time to time but housekeeping those is a nightmare-- Outlook consistently freezes while deleting larger numbers of email.
I am constantly amazed at the number of bugs and seemingly obvious UI issues pervasive throughout microsoft products. It feels like their product philosophy is "jack of all trades, master of none" - they try to support so many uses cases that there isn't a cohesive best practice anywhere.
I feel like this has only really been the case since about 2015, so instead of ascribing it to a "jack of all trades" philosophy I believe it can be more readily explained by a "lets hire a bunch of webdevs to work on our desktop OS" philosophy.
I wonder if this is an attack vector. On one hand it isn't because usually the HTTP traffic of Exchange is encrypted, on the other hand it is a crash triggered by a remote condition.
I don't know if the article mentions this because (ironically) Bleeping Computer is now down, but you may be able to fix this by holding Ctrl while opening Outlook in order to start it in Safe Mode. Whenever I'm in a broken state, I do this, load it up fully, close it again, and start it normally. Then things usually work okay.
I had to do that this morning on my work machine, not sure if it was related to this issue.
The article mentions this, but it says that apparently, with this specific bug, you also Have to hold Ctrl every subsequent time you re-open Outlook - it’s not a one-off fix. They suggest going back to the previous release as an alternative.
I had the problem with Outlook; I have to use it for work. While waiting for the help desk to get back to me on Outlook crashing, I used "quick repair" and it cleaned up the problem.
That worked perfect for me also. "%Programfiles%\Common Files\microsoft shared\ClickToRun\officec2rclient.exe" /update user updatetoversion=16.0.12827.20470
After 5 hours of failed attempts and many failed system restore points the updatetoversion link above worked on multiple pcs
Somewhere there is a programmer that tried to fix a bug in 10+ year old legacy code (mso98win32client.dll) who has learned the bugs of legacy are features of today.
Wild - I experienced this first thing in the morning. I'm running Windows 20H2 (19042.388) and Office Beta (v2008, b13102.20002) and assumed it was from the recent updates. Online repair didn't work, full repair didn't either, but a complete uninstall/reboot/reinstall worked and preserved my profile just fine (I thought it was perhaps the profile too but guess not) - Wild!
> there is nothing obvious about what went wrong. It just...closes.
Yeah, that's disconcerting from a support standpoint. Fortunately, I was able to remotely load their event logs and get the status code. DDG pointed right to the above article.
After 5 hours of uninstalling and failed windows system restore points, and a quick 24 google search worked! The below link mentioned really fixed the issue on multiple pcs. Thank you!!!!
"%Programfiles%\Common Files\microsoft shared\ClickToRun\officec2rclient.exe" /update user updatetoversion=16.0.12827.20470
To figure out what the actual message is, you first have to figure out if you are seeing a HRESULT or a NTSTATUS[1]. In this case the leading 0xC is an invalid HRESULT, so it's for sure an NTSTATUS. Then you just look up the code in the bottom two bytes [2].
In this case it's just error code 5: "STATUS_ACCESS_VIOLATION", which is a memory access error, which typically means it's a bug where the program accidentally is trying to read outside of its memory space.
So, for sure an application error and not something that the user did or some sort of system issue.
There are also "facility" codes which can give you more information what where the issue is happening, but this code is just using the default.
I suppose this is better than when they would print the decimal value instead of the hex. "WTF is error code 3221225477?"
[1] https://docs.microsoft.com/en-us/openspecs/windows_protocols...
[2] https://docs.microsoft.com/en-us/openspecs/windows_protocols...